TechniFundamentals

Merry Xmas Edition: Fun With Statistics: Chernoff Faces

2010-12-17T19:30:00.005+11:00

Chernoff Face Features Mapped To ValuEngine Model Variables

The Faces Of The Thirty DJIA Stocks

Let's take a break from High Alpha-Low Beta stocks for this week and have some fun with statistics, with a visualization tool called Chernov Faces. Chernoff Faces help us to visualize multivariate data, by putting a 'face' to it. Here's the explanation from Wikipedia:Chernoff faces, invented by Herman Chernoff, display multivariate data in the shape of a human face. The individual parts, such as eyes, ears, mouth and nose represent values of the variables by their shape, size, placement and orientation. The idea behind using faces is that humans easily recognize faces and notice small changes without difficulty. Chernoff faces handle each variable differently. Because the features of the faces vary in perceived importance, the way in which variables are mapped to the features should be carefully chosen (eye size and eyebrow-slant have been found important). For the drawing of the face of each DJIA stock, I have selected 14 of the ValuEngine model variables and each stock's values of these variables constitute its face. The whole matrix is standardized by scaling each column from 0 to 1 so that face features are relative to the population. The face feature and the model variable that is associated with it is shown in the top Table. We have to remember that for some variables, the smaller the value, the better e.g. valuation % and volatility. Let's see what we can glean from the faces of the DJIA stocks:

1. I think the most unique face belongs to Alcoa: The slanted eyes represent its small market cap. The small face shows its undervaluation.2. Caterpilar has an extremely long nose, and that's due to its high Beta. 3. Microsoft's eyes are extremely far apart because its got good 5-year returns.4. Bank of America's small face shows extreme undervaluation.And so you can compare the faces by yourself. The spreadsheet of the actual values is available from me if anyone wants it. All in all, I would think that (1) they all look quite alike except for weird people like Alcoa and Bank of America. Everybody is smiling because the last three columns that involved features of the mouth were not used. General Electric looks honest and Coca Cola looks like the jovial pastor from the church on Main Street. I don't like the looks of Hewlett Packard, the narrow jaw relative to the wide forehead.

And finally for a Gallery of The Art of the DJIA, like the Contour Plot in the top image, see it on my other Blog at http://www.fu-lu-shou.net/2010/12/stock-market-data-as-art.html Until then, Have A Merry X'mas and a Happy New Year.

HighAlpha-LowBeta Stocks: Week 3

2010-12-11T14:18:00.006+11:00

A beautiful representation of ValuEngine's model variables applied to the thirty DJIA components. Model variables were first standardized by converting to Zscores and Principal Component Analysis was done on the selected 15 variables. What you see is a 3D surface and contour visualization of the coefficients of each Principal Component. For more of DJIA as art, see http://www.fu-lu-shou.net/2010/12/stock-market-data-as-art.html

The list of stocks
Last 12-Month Return % (Momentum)

Beta

EPS Surprise %

M/B Ratio

P/E Ratio

Valuation %

Volatility %

This week we continue with our screen for high Alpha, low Beta stocks [see previous two posts]. If you compare this week's list with the first list, you will notice that many of the stocks were on the first list e.g. Frontier Gold, US Gold, Logmein, 7-Days Group etc. You will also realize that many of the stocks are in the high expectations industries e.g. mining, biotech, Chinese Internet, Oil and Gas. These are what we usually think of as high Beta stocks, but the difference is that these particular stocks have a high Alpha and low Beta-which makes any run-up more sustainable [see their low Beta values in top table]. But ValuEngine's Beta are longer term Betas, and therefore it is possible for these stocks to have short term correction. Frontier Gold was 9.19 when we listed it two weeks ago versus 10.42 now. US Gold was 5.56 versus 7.56 now, Western Refining was 9.25 versus 10.22 now. So whether these stocks continue to run up depends on the market, the perceptions on Gold, Oil and the USD. What we can do is to take a look at the performance of this week's portfolio next week. For now, just take a look at each of the image maps based on some of the model variables. First of all, note that the list of stocks are in two clusters: One in segment S3 and the other in segment S1. Segment S3 include the red-hot stocks like FRG,UXG and WNR which have already made good gains.

The images are self-explanatory. The stocks are plotted against the backdrop of the S&P500 stocks.

Look at where the stocks are positioned, and look at the scale below. In general, these high Alpha stocks have become more expensive [see valuation map], with high P/E ratio, and higher volatility. But their M/B Ratio is low and the EPS Surprise % is very high. Momentum is high, and Beta is higher for the stocks in S3, but lower for the stocks in S2.

My bet is on the lower Beta stocks in S2: PBT,LOGM,SVN,MNRO,OPEN and HTWR. Let's see how they perform next week.

Performance Of Last Week's HighAlpha-LowBeta Stocks

2010-12-04T16:57:00.003+11:00

+ 37 % change of last week's stocks versus S&P500's + 3.9 %

Last week, I accidentally left out the full details of my ValuEngine Institutional screening. Well, here it is: Because there is no definitive way to interpret the value of Jensen's Alpha (other than to say that it must be positive), and because Alpha is derived from Beta, and Beta is about performance relative to the market as represented by an Index- I used the Ranking system in ValuEngine. That is I screened for stocks with Jensen's Alpha Rank > than 90. This means that in relation to ValuEngine's Universe of 4500 stocks, these are the stocks in the top 10 % in terms of Alpha value. Also, the stocks have to have a minimum market cap greater than $0.5 Billion and average daily volume greater than 100000 shares. Then I added the constraint that the Beta should be less than 1. When this yielded too many stocks (about 76) I progressively lowered the Beta value until it was 0.6.
Now, lets take a look at the performance of the selected stocks which was based on their location on the Self-Organizing Map [SOM]. (Refer to last week's post). Well, from the image above you can see that the results have been quite good. Against the S&P's performance of +3.9 %, we have a portfolio average of + 37 %. Mostly due to two big gainers: Frontier Gold and U.S. Gold Corp. Tune in again next when we shall do more of the same and demonstrate that even in a flat market, high Alpha low Beta stocks can outperform.

Stocks WIth High Jensen's Alpha and Low Beta For An Uncertain Market?

2010-11-28T17:03:00.011+11:00

The formula for Jensen's Alpha

The High Alpha/ Low Beta stocks

Beta map Sharpe Ratio Map

P/E Ratio Map

M/B Ratio Map

LAst 12-M Return % Map

Forecast 1-M Return % Map

I just discovered that ValuEngine's Institutional software does have Jensen's Alpha in its output. Jensen's Alpha is basically risk-adjusted Alpha [see formula in top image], and is what hedge fund managers use as measurement of their performance (and to justify their management fee). Jensen's Alpha is derived from the Beta, and therefore inherits all the flaws of a simple measure like Beta. Beta is a meaure of regression/correlation and its value depends on (1) the period used for calculation (2) assumption that it doesn't change much during the period of measurement and is not volatile [bad assumption]. What I wanted to know was, what kind of stocks are high Alpha low Beta and whether they would be suited for an edgy market situation like currently. When the market is strongly trending upward, you ought to pick high Beta stocks and never mind the Alpha, But when the market is uncertain, low Beta high Alpha stocks should do better since hopefully the Alpha component of a stock's characteristics will carry through to make gains in spite of the market.

Thus using ValuEngine, I screened for stocks with (1) minimum market cap > $ 0.5 Billion (2) Average daily volume > 100000 shares (3) Beta<>

The results are interesting. Refer to the images above and looking at the scale below each map you will find the following characteristics of the high Alpha low Beta stocks in the table above.

1. These stocks have a high Sharpe Ratio- these are safe stocks, as measured by their returns/standard deviation over 5 years.

2. They have high last 12-month returns % and in the world of fundamental analysts that's high Momentum!

3. They have high P/E Ratio but low M/B Ratio- i.e going by book value ( and not just earnings) again, these are safe stocks.

What more could an investor wish for in such uncertain times? Now let's go over to the qualitative side and look at the Companies that are on our list. [If you look at any of the SOM, you will see that most of the stocks (their ticker symbols) are clustered very close together- a good sign that they have a high degree of similarity not only in their Alpha, but in all the other fundamental variables of the ValuEngine model]. For this reason, let's leave out WNR Western Refining way out in S1 cluster and ALK Alaska Air also away from the tight cluster.

FRG Frontier Gold: Exploration and Mining of Gold, Silver, Copper and Uranium

LOGM Logmein: Information Technology remote connectivity services for small and medium businesses

NFLX: Netflix. We know what it does, maybe there's some takeover of this business?

NRGY: Inergy L.P: owners of gas pipelines, storage tanks, terminals and other distribution networks.

PBT: Permian Basin Royalty Trust: Royalty rights in mineral properties in the U.S.

QCOR: QuestCor Pharmaceutical: Drugs for Nervous System, inflammation, insomnia

SQNM: Sequenom Inc: Biomedical Genetic Analysis and Molecular Dynamics for humans, agriculture and livestock.

SVN: 7-Days Group ADR: Chinese budget hotel chain with 400 hotels.

SVR: Syniverse Holdings: Wireless voice and data services for telecommunications companies worldwide

UXG: U.S. Gold Corp: Gold mining with properties in USA and Mexico

VHC: Virnetx Holdings: engages in developing and commercializing next generationsoftware and technology solutions for securing real-time communications over the Internet.

Well, wouldn't you agree that these are all sexy Companies? Just remember that the Alpha and Beta values used here are longer term calculations in line with ValuEngine's fundamentals-based models, so don't expect this to be like technical analysis. Perhaps next week we screen again for such stocks and denoise them or look at them with Wavelets like in some of the other posts.

Enigmatic Stocks Of The Singapore Exchange: Capitaland, Hyflux, Jardine

2010-11-27T18:48:00.006+11:00

Continuous Wavelet Transform of Capitaland

Capitaland De-Noised

Residuals (Noise) of Capitaland

Continuous Wavelet Transform of Hyflux

Hyflux De-Noised

Residuals (noise) of Hyflux

The Inexplicable Jardine Cycle & Carriage

Using Wavelets, we take a look at three Blue-Chips of the Singapore Exchange and try to explain why even long term holders may lose patience with certain Blue-Chips. Capitaland and Jardine Cycle & Carriage are components of the 30-stock Straits Times Index of Singapore. Hyflux used to be a component too, I think. The data here is for approximately 5 years [see number of days indicated at the top of each chart]. Before you begin: (1) you may want to read more in general about Wavelets from the other posts on this Blog (2) The Y-axis represents 'scale' on wavelets, the higher the longer term. (3) Patterns that you see are real. They are fractal self-similarities if they extend up the image (3) The higher on the Y-axis, the bigger the picture i.e the low frequency approximations while at the bottom are the high frequency details.

And if you look at the charts, Capitaland and Hyflux have been going nowhere, despite the strong fundamentals. Hyflux is one of the top water technology Companies in Asia, and Capitaland is South-East Asia's largest property developer. The notes on each image are self-explanatory. To summarize: As the years went by, shorter term players began to lose interest in Capitaland and Hyflux, and now there is less and less interest in them. Hyflux does have higher 'turnover' than Capitaland, but these are on the longer term scale. Hyflux is a totally different animal from Capitaland, i.e. its holders are of a different type. The autocorrelation level of Hyflux is very high compared to most other stocks which means that it is more suitable for technical analysis (of the longer term kind, maybe using weekly data). As for Jardine, its Continuous Wavelet Transform shows it to have basically no pattern of any kind. I don't know what to say abgout it. Its most common use as noted by many Singapore traders is as something to throw in a few seconds before the closing bell, to affect the ST Index one way or another. Will Hyflux and Capitaland change their 'personalities'? Not unless they draw more interest from shorter term players.

Explaining and Interpreting Wavelets

2010-11-21T17:59:00.007+11:00

The shape of a typical Wavelet, in this case a Daubechies

Breaking down a signal with Wavelet

Continuous Wavelet Transform (CWT) of iShares FTSE/Xinhua China 25 Index(966 days till November 20 2010)

Hope the above helps in explaining and interpreting Wavelets. Basically, a Mother Wavelet like the Daubechies wavelet and its Scaling function the Father Wavelet is stretched and shifted along the length of the signal, and the correlation between wavelet and signal recorded as coefficients- the larger the higher the correlation with the signal. Starting from a small wavelet, that caters to the higher frequency (details), the wavelets are stretched into bigger wavelets for approximating the lower frequencies at the higher end of the scale. Actually most wavelet transforms are Discrete and have less scaling and shifting and are yet able to approximate the signal for full re-construction approximate the signal. But in our case we have used a Continuous wavelet transform without redundancy , to go every step of the way and get a full pciture because in highly non-stationary financial data, interesting details are lost if Discrete wavelets are used.

Shanghai Index (SSE) More Technically Driven Than Singapore Index (STI)

2010-11-21T02:21:00.006+11:00

CWT of XLF

1. Continuous Wavelet of SSE (Shanghai Composite)[Daubechies 6 to Level 5]

2. SSE De-Noised with Haar wavelet and Adaptive Thresholding to Level 5

3. Residuals of SSE, Histogram, Autocorrelation and Spectrum

4. Continuous wavelet of STI (Straits Times Index)[Daubechies 6 to Level 5]

5. De-Noised STI with Haar wavelet using Adaptive Thresholding to Level 5

6. Residuals of STI, Histogram, Autocorrelation and Spectrum

Contrary to popular opinion, a market that is less-fundamentally driven [by which I mean it is more technically driven] is easier to predict. This can be illustrated with the example of the SSE versus the STI. The SSE is wilder in its swings and speculation plays a greater part. It is also more insular and less affected by international events and other international markets. The STI on the other hand is very open, and like Singapore's economy is easily rocked by international events and other markets. Trying to predict the direction and magnitude of change for tomorrow's close of the STI is nigh impossible. On the other hand, using technical analysis and other non-linear tools, it is possible to for a short period grasp the mathematical properties of the SSE for prediction purpose.Image 1 and image 4 compare a continuous wavelet (CWT) of the SSE and the STI. A CWT can capture the fractal self-simlar patterns of a time series, and looking at the SSE and the STI we see that the patterns are less defined for the STI. As an afterthought and to show you something that is even less predictable than the STI, I put the CWT of the XLF which is the iShares ETF of the US Finance Sector right at the top.

In image 2 and image 4, I used a Haar wavelet to denoise the SSE and STI. The original and de-noised signals can be seen as an overlay. Unlike a moving average, wavelets have no lag. And if we were to for example put a +1 % upper band and a -1% lower band to the de-noised signal, we could use it as a Buy/Sell signal. You would be able to ignore the small moves, ride the big upswings and cut loss on the big downswings. Images 3 and 6 show that the noise removal has been quite effective. The residuals are random, the histogram is Gaussian (normal) the autocorrelations drop nicely from zero, and spectrums indicate that signal energy has been retained for the lower (significant) frequency approximations.

Beta Is A Double-Edged Sword

2010-11-20T13:42:00.009+11:00

1. Clusters

2. Cluster Statistics

3. Cluster Summary

4. Beta

5. Valuations

Beta, which measures the sensitivity of a stock to a market Index, is a double-edged sword. When the market has high upward trend and momentum, high Beta stocks do well. When the market has a strong correction, high Beta stocks will also be strongly affected. As a conservative investor, I prefer to look for stocks with low Beta and high Alpha. Alpha is difficult to define. While Beta is the slope of the linear regression line that has the stock's % change on the X-axis corresponding with the market Index's % change on the Y-axis, Alpha is the point on the Y-axis where the Beta line starts from X=zero. Unfortunately, this definition is not helpful in terms of stating quantitatively what a value of Alpha implies. For example what does a value of 0.8 for Alpha mean, what does Alpha > 1 mean? Just as hedge fund managers define Alpha as the portion of their skills that enable them to outperform the Index, I will now define Alpha as a stock's 'distance' (mathematical, Euclidean Distance) from the Index in terms of overall fundamental characteristics- the greater the distance the better.

A Self-Organizing Map which clusters stocks according to their degree of similarity in overall characteristics, will allow us to look for high Alpha stocks. With the SOM, high Alpha stocks are therefore those which are as dis-similar to the market as possible-in a good way of course. With this in mind, and using ValuEngine's fundamental variables, I first screened for stocks with minimum market cap > US$0.5 billion, and Average Daily Volume > 100000 shares. Then I ranked them by their Beta, and took the 40 stocks with the lowest Beta. # It must be mentioned that the value of Beta depends on the time period that you use for calculating it, and ValuEngine's Beta is a long term Beta. These stocks were plotted on a SOM using the S&P500 as a backdrop. Image 1 shows the clusters generated by the SOM, and the Ticker Symbols of the selected low Beta stocks, and their position on the SOM. Looking at Image 2 and Image 3, it is clear that S1 cluster is the cluster where most of the S&P500 stocks reside. So in our task to select low Beta, high Alpha stocks, we leave out all of the low Beta stocks which are in S1, as we want stocks which will be as different from the Index as possible.

Next, we look at cluster S2. Is it really different from S1? From Image 2 which measures the difference between clusters in terms of standard deviation (the longer the bar the greater the degree of difference), there really is not much difference between S1 and S2. It's only S3 which has a big difference with the Index. Nevertheless, there are some differences between S1 and S2 in some
key model variables. For example, S2 stocks are undervalued while S1 stocks are overvalued, S2 stocks have higher 1-month forecast return %, S2 stocks have lower Sharpe Ratio, lower 12-month return % etc.

Next take a look at Image 4 which shows the Beta as a color-scaled intensity map. And Image 5 is the same type of map for Valuations.

The low Beta areas are Blue/Violet. I have identified and labeled the low Beta stocks which also have low valuations. The list of stocks and their business profile from Yahoo Finance are:

ADM- Beta: 0.23. Archer Daniels Midlands Company: Archer Daniels Midland Company procures, transports, stores, processes, and merchandises agricultural commodities and products in the United States and internationally. It operates in three segments: Oilseeds Processing, Corn Processing, and Agricultural Services.

AVAV-Beta 0.23. AeroVironment Inc: AeroVironment, Inc. designs, develops, produces, and supports unmanned aircraft systems and efficient energy systems for various industries and governmental agencies. It offers small, hand-launched unmanned aircraft systems (UAS) that provide intelligence, surveillance, and reconnaissance, including real-time tactical reconnaissance, tracking, combat assessment, and geographic data to the small tactical unit or individual war fighter. THAT'S HOT !! :)

APOL- Beta 0.09. Apollo Group: Apollo Group, Inc., together with its subsidiaries, provides various educational programs and services at the undergraduate, graduate, and doctoral levels. The company offers associates, bachelors, masters, and doctoral degree programs in arts and sciences, business and management, criminal justice and security, education, human services, health care, psychology, technology, and nursing through its campus locations and learning centers in 39 states and the District of Columbia, and Puerto Rico, as well as through online educational delivery system.

PBCT Beta 0.21 - People's United Finance. People's United Financial, Inc. operates as the bank holding company for People's United Bank that provides commercial banking, retail and small business banking, and wealth management services to individual, corporate, and municipal customers. The company operates in three segments: Commercial Banking, Retail Banking and Small Business, and Wealth Management.

TFSL- Beta 0.28 TFS Financial Corporation operates as the holding company for Third Federal Savings and Loan Association of Cleveland that provides retail consumer banking services in Ohio and Florida. The company offers various deposit accounts, including savings accounts, NOW accounts, certificates of deposit and individual retirement accounts, and other qualified plan accounts

HMY- Beta 0.29 Harmony Gold Mining: Harmony Gold Mining Company Limited engages in underground and surface gold mining. It also involves in related activities, including exploration, processing, and smelting. The company operates a total of 10 underground operations, 1 open cast mine, and 8 processing plants located in the Witwatersrand basin of South Africa, as well as the Green Stone belt. It also holds interests in the development and exploration prospects at Hidden Valley and Wafi in Papua New Guinea. In addition, the company holds interests in the Amanab and the Mount Hagen Projects located in Papua New Guinea.

As you can see from the list of selected stocks, low Beta high Alpha has no sector preference. I thought that Utilities and big cap Pharna stocks would make the list. Anyway, this is a more holistic way for selecting stocks with low Beta and high Alpha.

Technical Analysis: Lag_Free Noise Removers Using Wavelets

2010-11-13T00:12:00.009+11:00

1. My experimental no lag indicator: Buy/Sell when cross zero: Keppel Corp

2.Normal 5-Day Simple Moving Average

3.Continuous Wavelet Transform of Keppel Corp

4. Compression

5. Residuals of Compression

6. Denoised and original signal

7. Independent interval thresholds for change in characteristics

8. Residuals of Denoising

9. Regression Estimate of Main Signal and original signal

10. Residuals of Regression

# Credits and Citations for posts on Wavelets: Much of the pioneering work on denoising using wavelets and the use of thresholds was done by D.L Donoho and I.M Johnstone (USA) and Kerkyacharian and Picard (France). Also thanks to Amara Graps of the South-West Research Institute , Colorado and and Robi Polikar of Rowden University,NJ whose articles made understanding of Wavelets by the layman possible. And finally, thanks to Matlab, whose Wavelet Toolbox is the industry standard, and whose User Guide is better than any text book.

Every technical analysis enthusiast is familiar with the Moving Average and what it does- it smoothes the signal and thus acts as a de-noiser so that you can ignore the noise. But noise always has to be user-defined as what is not noise to a short term investor is noise to the long term investor. Although a useful tool for medium term and longer term traders, a Moving Average is less useful to the short term trader; and choosing the optimal period for an MA is a bit of guesswork, since you can back test all you want, but the market and the stock's characteristics are always changing.

In this post we use Wavelets to denoise a stock signal [the example used here is Keppel Corp- a Blue Chip on the Singapore Exchange]. Wavelets are a more recent technology used in digital signal processing for working with signals that are not regular, and have sharp sudden and transient moves. Fourier transforms with their regular sinusoids cannot handle such signals. Stock market signals fit this category of signals. In wavelet methodology, the signal is also decomposed into constituent wavelets like in Fourier transforn, but different parts of the signal in time are handled by different stretched and shifted versions of the mother wavelet. Thus changes in characteristics of the stock signals can be accomodated, Moreover, wavelets have compact support and are orthogonal. Compact support means the wavelest have a cut-off point unlike regular sine waves which go on forever. Orthogonal means there is no overlap of information being handled by different constituent wavelets. There are three different ways that Wavelets can take away noise from a signal: (1) by normal denoising (2) by compression (3) by regression. Although the algorithms for each is different, the end result is the same: you get a signal that is smoother than the original signal. The test for how good a denoiser is can be shown by the residuals. That is, after denoising, compression or regression, the leftovers, if they are indeed (white) noise will show a random distribution. Lets go through each image from the top:

Image 1 shows a technical indicator I developed using the wavelets methods below. The idea is to Buy when the indicator crosses above zero and sell when it crosses below zero. As you can see from the vertical lines aligned with the stock chart, the indicator is quite efficient. Only problem is how do we determine whether the crossing of zero can be sustained. We will have to use it in conjunction with another indicator.

Inage 2 shows a normal 5-day Simple Moving Average of the stock. 5- day is too short and you won't want to move in and out of the market so often.

Inage 3 shows a Continuous Wavelet Transform of Keppel Corp using a Daubechies 4 to Level 5, and scale 1: 128. All it can tell you is that the stock does not move randomly. There is a large deterministic component, but it is always changing in its parameters. The image shows the fractal self- similarity characteristic of Chaos Theory.

Image 4 shows how compression can give you a smoother less noisy signal. In this case, 99 % of the signal's energy (entropy) was retained while 92 % of the image was padded with zeroes. That is, only 8 % of signals's content was sufficient to derive the compressed signal. But if you look at the residuals in image 5 , the residuals are not as random as the residuals for denoising and regression below. Which means that some useful content was also taken away. Still I think that if we are trading a Blue Chip like Keppel Corp where we can have a greater tolerance for 'noise', this derived signal is the best of the four methods here.

Image 6 shows the signal after denoising, and it hugs closely the original signal and has no lag. This denoising was done with two different thresholds. As shown in image 7, the market got noiser and more volatile during the last 300 trading days or so, so a different threshold for defining what is noise was used. The residuals show that the denoising was quite effective.

Image 9 and 10 shows how regression can also be considered a form of denoising. Regression algorithms are 'fitting' algorithms but the end result is the same. We get a smoother less nosiy derived signal.

I could fine-tune the regular denoising and regression algorithms to hug the main signal less as in the compression algorithm. But it's hard work and still a trial and error thing choosing the right wavelet family, the appropriate number of vanishing moments, level of decomposition etc. Nevertheless my point is that it's time the people who design technical indicators think of using Wavelets to do the work. Wavelets are a better tool to analyse the kind of signals that stock markets generate, taking into account the non-linear adaptive dynamics that characterise stock markets.

Do Speculative Stocks Have Patterns?

2010-10-31T01:04:00.005+11:00

Yangzijiang Shipbuilding
Wal-Mart Stores Inc

Singapore Press Holdings

Golden Agri- Resources

Genting Singapore

If stocks do have some repetition in the nature of their movements, it would still be too elusive for us to 'capture' for prediction. If patterns do exist, they are always changing. Thus, a Moving Average crossover that served you so well in the past will suddenly not work. Even so called natural phenomena like Fibonacci and Elliot Waves cannot cope with these constant changes. Yet, the changes in a stock's time series are not totally random either. Besides being underpinned by fundamentals in the long run, they also have rhythms and cycles caused by autocorrelation, their Beta, and momentum. And like all complex adaptive systems, action creates reaction, and evolution spawns co-evolution. Thus in this dynamic environment where equilibrium is a moving target, we buy and sell, attempting to impose some order on the seeming Chaos, but never quite succeeding. The reason is the great degree of non-linearity which cannot be captured by traditional statistics or any other parametric, linear approach. However, Wavelets- a technology that greatly enhances the efficiency for de-noising, compressing and analyzing 1-dimensional and 2-dimensional signals can give us an inkling of the character of a stock. For more on Wavelets see:http://www.fu-lu-shou.net/2010/10/stock-market-data-as-art.html Here we use a 1-dimensional continuous Daubechies 5 wavelet for analysing and comparing the 'predictability' of stocks. The patterns that you see in the images above are patterns of fractal self-similarity, which is something from Chaos Theory and much related to the work of Benoit Mandelebrot the mathematician who recently passed away. Books that tell the story of the great Quantitative Analysis traders who ran the proprietary trading desks of Goldman Sachs or started hedge funds show that many of them consulted Benoit Mandlebrot on aspects of Chaos Theory that could be applied to short term trading.

Intuitively, when there are more distinct patterns, we can say that the stock is more predictable [at least during the period when the patterns are distinct]. Mathematically, to delve into a more quantitative analysis of predictability involves looking at the wavelets different levels, the Details amd Approximations and the coefficients.* When a signal is decomposed by stretching parts of the Mother Waevelet, to fit a particular length of signals, a 'tree' with two branches and several levels is created: Approximations for low frequencies of the signal and Details for high frequencies parts of the signal. Analysis of wavelets requires deep domain knowledge which a Wavelet novice like me do not possess. It takes years to learn to pick the right family of wavelets to apply, and it is still an art to choose the level of decomposition, and then to interpret the results. Maybe for image compression or classfication Wavelets requires less heuristics, but definitely for use of Wavelets in financial data, a person with domain knowledege of both Wavelets as well as financial markets is required.

In the images above, of some popular stocks on the Singapore Exchange, all I can conclude is that the movement is not entirely random, there is method even in the madness of a stock like Genting, and we can say that some stocks are more predictable than others. For a great contrast, look at the 'regularity' of a stock like the USA's Wal-Mart and compare it with Genting where some areas are without any patterns at all. A Blue Chip like Singapore Press Holdings bears some resemblance to Wal Mart in behavior. Golden Agri and Yangzijiang look frightening. The fine details at the bottom of both show very short buying and selling 'cycles' as in contra trades.

Forget Healthcare, Tech. Go For The Heavy Stuff

2010-10-30T19:08:00.008+11:00

A. Oops forgot the P/E map

B. Frequency Distribution: 5-year average annual return of 3000 US stocks

1. Stocks in Selected areas of the SOM for highest 1-month forecast return %

2. Darkened nodes marked by squares are the selected areas

3. Market/Book is low

4. Valuation is low too 5. However volatility is on the high side

March 2009 was when the market was at its lowest, and of course since then it has rebounded to a remarkable extent. But if we go by the 5 year average annual return % of of 3000 U.S. stocks, the picture is not so rosy. As at now, many stocks still post a negative return [see top image]. In the frequency distribution above, the area associated with the number of stocks with returns less than 3.4 % average 5 year return is much greater than the area after that. In fact the distribution has a heavier, fatter tail on the negative side of the X-axis. Since the immediate outlook for the market is so murky, I focus on the short term. Using ValuEngine's 1-month forecast return % model, I screened for all US stocks with market cap greater then $100000 and and Average daily volume greater than 100000. The screen generated a list of 3000 stocks. These stocks are then placed on Viscovery's SOM, and using the mouse, the areas of highest 1-month forecast return were selected. The stocks in the darkened selected areas [marked by squares] are in image 1. Their location on the SOM remains the same when we look at them from the perspective of the other model variables. They score high in terms of having low M/B ratio and low valuation. But are more volatile. [see the other maps]. Low P/E too, see top image.

But the most information can be gleaned from image 1 showing the list of stocks. Take a look at the sectors that these stocks belong to. Are there any Tech , Healthcare or Finance stocks? [well there are two Tech stocks]. I am intrigued by the over-representation of the number of Transportation, Energy and Public Utilities stocks, considering that they form a smaller % of the US stock population. Take note too, of the smaller market cap, reasonable valuation and Beta. When public utilities and transportation stocks sport a higher than normal Beta, it means something.

A Sneak Preview of 19 China ADRs Trading On SGX Tomorrow

2010-10-21T13:21:00.006+11:00

1. The 19 China ADRs trading on the Singapore Exchange tomorrow
2. Returns comparison: Ctrip, China Mobile, PetroChina and Baidu

3. ValuEngine's portfolio optimization for China ADRs 4. Portfolio optimization engine's parameters and constraints

On October 22, nineteen China ADRs currently listed on NYSE and NASDAQ can be traded on the Singapore Exchange [SGX]. Their total market cap of approximately US$600 million is equal to the market cap of the whole SGX. Will they significantly boost the liquidity of SGX, and more importantly will retail investors take to these ADRs? In this sneak preview, using ValuEngine Institutional software, we highlight some aspects of these ADRs. [Individual reports on these ADRs are available on ValuEngine's web site]

1. Image 1 shows the list of 19 ADRs. They cover a range of sectors from Internet companies like Baidu and Netease to Utilities (Huan Eng Power), Energy (PetroChina, Yanzhou Coal), Biotech (Mindray), Alternative Energy (Suntech, Trini) to Telecoms (China Telecom, China Mobile etc) to Services (Ctrip, China Eastern Airline). Price-wise, Asian investors may not be used to the high price [see market price as at yesterday]. Especially the traders with a shorter term investment horizon looking for a quick turnover. They may be a little disappointed as these big cap China ADRs' behavior are not much different from the average S&P500 constituent stock. [ see previous post on Sept 18 ]But for the longer term fundamentals-based investor, they are now able to tap into China's enormous growth potential. Being listed on U.S. Exchanges, they are also subject to the regulatory environment there and therefore have some edge in quality over other China stocks.

2. Image 2 is a comparison of four stocks: Ctrip a travel and tourism site, China Mobile the cellphone telecom, PetroChina and Baidu the well known Search Engine. I did not choose the older more mature ADRs like Yanzhou coal and Huan Eng Power as their returns were well below these newer industries. You can see how each stock behaved before and after the financial crisis and judge for yourself, where the future lies for China ADRs.

3. In image 3, I tried to do portfolio optimization of the 19 ADRs to determine my % capital allocation if I were to create a portfolio of these ADRs as investment. ValuEngine was not able to optimize all nineteen ADRs as for one reason or another, some of the ADRs lacked some data required for quantitatve analysis of their valuation, forecast, and risk assessment or were automatically filtered out by my specifed optimization constraints. Therefore the results show a portfolio of only ten of these ADRs. The ValuEngine models gave greater weighting % to Changyou the interactive games developer, Shanda the multi-player online games site, Ctrip the travel site, and surprisingly, an old-fashioned industry like Huaneng Power. The valuation column shows the +/- % valuation, and the valuation rank column ranks the ADRs in comparison with 4500 oher U.S. stocks. (from 1 to 100, the higher the rank the more undervalued). You also have the forecast for 1-month as well as 1 year, and you will notice the very high P/Es that these ADRS are trading at- except for Changyou and Trini Solar. I wouldn't pay too much attention to the forecast as these ADRs are more dynamic than typical US stocks, and their characteristics change at a faster pace. Image 4 below shows that such a porftolio would have a forecast annual return of 18.67 % based on the current situation. The Beta of the portfolio is not too bad at 1.34, but the standard deviation (volatility) is high at 29 standard deviations.

4. Image 4 shows my constraints for this portfolio optimization exercise: I restricted each stock to a maximum of 20 % of portfolio capital, with no stocks from the same sector constituting >50 % of the portfolio in $. I also deleted any stock that is > 20 overvalued. The initial capital is US$1 million. The portfolio was optimized based on 1 month forecast. * This is important to note: Of course I can optimize based on 1 year or even 3 months forecast. But this would be essentially useless. If there's one thing I learned all these years: Like weather forecasts, stock forecasts have less probablility of being upset, the shorter the period of forecast. 1-day forecasts are best for technical trading, but for fundamentals, 1-m at least gives you some guidelines, though you must be prepared to constantly update your forecasts.

Using the U-Matrix To Identify The Most Dense Areas Of A SOM

2010-10-16T14:24:00.010+11:00

Volatility attribute window
Beta attribute window

1. U-Matrix of clusters and selected areas

2. Statistics of each cluster

3. Area of high 1-month forecast return %

Stocks picked from S2

5. Stocks picked from S3

For explanation on technology, methodology and terminology, see: http://www.technifundamentals.com/2010/07/explanation-page.html .

A more refined way to visualize clusters created by a SOM is to use the U-matrix (unified distance matrix). In image 1 above, lighter colored areas are more dense than darker colored areas. Thus there are more stocks located on the lighter colored nodes. In this post, we use ValuEngine to screen for all stocks with a minimum average daily volume of 100000 shares and a minimum market cap of $100 million. There were 2867 stocks in this category, and they are plotted on the SOM. Image 2 shows that clusters S2 and S5 while both having low P/E and valuation, have different characteristics with regard to the other model variables. While S2 consists of smaller cap stocks, S5 stocks are big cap. S2 are lower volume stocks while S5 are highly liquid stocks. S2 stocks have high forecast 1-month forecast return [see area of high forecast return in S2] while S5 stocks have low 1-month forecast return. Also [see top images] S2 stocks have higher Beta and volatility than S5 stocks. So its a choice between big cap slower movers with lower risk or smaller caps with higher risks but greater potential returns. The lower two images are the stocks picked by selecting the densest areas of S2 and S5 according to the U-Matrix SOM. It is not surprising that S5 selected stocks include stocks like Kellogs, Campbell Soup and Kimberley-Clarke while selected S2 stocks are relative unknowns.

If you like QE2 [quantitative easing] go for S2. If you think that QE2 is going to cause inflation, continued weakness of the US$, and eventually, higher borrowing costs [yields] for Treasuries; and retaliation by other countries whose currencies are getting stronger, go for S5. these are the companies with operations in other countries, and whose earnings when translated back to US$ will grow even more. Or they are recession proof public Utility services and garbage collectors. Personally, I prefer S5 stocks. I don't think that QE2 can hold up the market for long.

SP500,DJIA,NASDAQ: What's the difference anyway?

2010-10-10T19:56:00.008+11:00

0: High P/E

1. Stocks in cluster S3

2.The three clusters: S1=Blue, S2=Red, S3=Yellow3. Cluster statistics of model variables: S3 is really different

4. Relatively high Beta of S3

5.However S3 has low M/B Ratio

6.Moderate valuation of S3

For explanation on technology, methodology and terminology, see: http://www.technifundamentals.com/2010/07/explanation-page.html .

With the DJIA, S&P500 and Nasdaq seemingly diverging, I wanted to know if they formed distinct clusters on the SOM. This would be an indication of their divergence. However, as shown in image 2, they do not form distinct clusters. The 'S's (S&P500 components), the 'D's (DJIA) and the 'N's are all mixed up on the SOM. How then, should you approach the current situation on the market if you are the fundamentals-based, medium term, big cap type of investor? I would suggest you look for stocks that stand out from the rest. On our SOM, S3 (Yellow) is the smallest cluster; but stands out from S2 and S1. This is shown in image 3 where the bars of each model variable are expressed in deviations from the Mean of the entire data set and therefore the longer each bar, whether positively or negatively, the greater the deviation from the norm. Thus, S1 which contains most of the S&P500 stocks is the most average cluster, while S2 is different but not the big difference that you see in S3. Image 1 shows you the stocks that make up S3. Again, it's a mixed bag. But I do discern some patterns here: You can see a many regional banks: Huntington, Regions, Suntrust, Zion etc. I also spot the beaten down Pulte and Lennar, as well as other Finance sector stocks like Legg Mason and ETrade. The general characteristics of the stocks in S3 are shown in the various Attribute Windows: images o (top) , 4, 5 and 6. Using the scale below each attribute window as a guide, we find that these stocks in S3:

1.Have high P/E, but low M/B (good sign?) as well as are undervalued by ValuEngine's models.

2. They are relatively high Beta stocks, so how well they perform depends very much on the performance of the market.

Combining Technical Analysis With Fundamental Analysis

2010-10-02T13:05:00.005+10:00

1. The ValuEngine Growth Model Stocks in Cluster S3
2. Clusters S1,S2 and S3

3. Forecast 1-Month Return % Map

4. MDC Holdings

$. Provident Energy PVX

5. Electronic Arts ERTS

6. Regions Financial

For explanation on technology, methodology and terminology, see: http://www.technifundamentals.com/2010/07/explanation-page.html .

Fundamental analysis tells you what to buy while technical analyis tells you when to buy. With that in mind, we see if we can combine the two. Specifically, we want to examine the technicals of ValuEngine's fundamentals-based models. For this, we generate a SOM of the S&P500 and place on it the stocks picked by ValuEngine's Growth Model [they call it the Forecast Model]. I use the Growth Model because unlike the Valuation Model, it takes into account the short term 1-month forecast return %. In Image 2, you see the 22 stocks of the Growth Model on the SOM. I narrow down to only those stocks located in cluster 3, because their 1-month forecast return % are the highest [as shown in image 3]. And because their cluster statistics in terms of deviation from the entire data base are also the highest.

Now, a little explanation on the home-made technical indicator we are going to use to examine some of the stocks in cluster S3: My home-made indicator calls a Buy when the Red histogram is above zero AND Tushar Chande's Q_Stick [top window Blue line] is also above zero. Preferably, both indicators should be rising. If there are divergences between the two, it is a cautionary signal. The length of the Red histograms tells you about the length of each Up and Down cycle and whether they are regular. As you can see, for all the selected stocks, the cycles are irregular and unpredictable. But we can see that ValuEngine's Growth Model stocks do have some 'early' technical signals for a potential price increase. With that in mind, take a look and analyze for yourself, the potential of Electronic Arts, Provident Energy, MDC Holdings and Regions Financial.

US,HK,SG Stock Picks With Prototype TA Indicator

2010-09-27T19:08:00.007+10:00

Hong Kong: Phoenix TV

USA: Las Vegas Sands
USA:EBay

Hong Kong: China Blue Chem

Singapore: Capital Commerical REIT

Hong Kong 361 Degrees

SEE PREVIOUS POST FOR HOW TO USE THE INDICATOR

Designing A Technical Analysis Indicator

2010-09-27T00:40:00.004+10:00

Prototype technical analysis Indicator on Genting Singapore stock

Technical Analysis is just visual analytics. If you approach it from this perspective, forget about prediction and cut the mumbo-jumbo about Heads and Shoulders, Gann, Elliot Waves etc, it helps make the mind clearer. Although the past cannot predict the future, I think is fairly safe to say that the greater the frequency or regularity of a pattern occuring, the greater the probability that it will occur again. It is my belief that the proliferation of algorithmic high speed trading programs makes technical analysis more relevant. Prices are out of equilibrium with fundamentals for a longer time, and inevitably, computer generated trades leave more distinct footprints that can be detected by sophisticated programs. Long term is for dinosaurs and the markets are really casinos. The equities market is beginning to resemble the Forex and commodities market in characteristics. The individual investor has been marginalized.
So, in that sense I am a born-again technical analyst. But I will do technical analysis with technologies that break away from the linearity and Gaussian distribution assumptions of traditional technical analysis.
Technical analysis 'wrings out' information from Open, High,Low, Close and Volume data. It does this by de-noising, smoothing, compressing, and normalizing the countless possible combinations of O,H,L,C,V. All these pre-processing techniques are well-known to practitioners of data mining.
But there is a limit to what anyone can do with just OHLCV. If used as input to a neural network, any group of technical analysis indicators will suffer greatly from collinearity. That is, much of their information content overlap; and there is no point feeding so many indicators into the network. This week I try my hand at constructing a technical analysis indicator. Tentatively, it will be known as the Singapore Flyer and is to be used only for the Singapore market.
It is a trend-folowing indicator. I don't believe in oversold/overbought indicators. It is important to catch a reversal. At least with the trend, we admit we don't know when it will end, but ride it for part of the way. The Singapore Flyer Indicator is to be used in conjunction with Tushar Chandes' Q-Stick Indicator. Go Long when SF crosses zero upwards AND Q-Stick is >zero. Go Short when SF crosses zero downwards, AND Q-stick is <>Technical analysis indicators if they are to be any good , will have to be customized for each market. Markets like Singapore are more affected by other financial markets from the DJIA, to the Hang Seng, to US$, Commodities and Bonds. Whereas for NYSE/NASD, domestic developments are more important. Hong Kong market is highly correlated with events in China, and Shanghai Exchange while not totally insulated from the the rest of the world, has a life of its own. To my many friends and ex-colleagues in the United States, if they are individual investors. I urge them to trade the Singapore and Hong Kong stock markets. There are better opportunities here for the individual investor who trades short term.

Technical Analysis As Nothing More Than Visual Analytics: Examples from Singapore

2010-09-26T01:22:00.005+10:00

Straits Times Index: Rahul-Mohindar Oscillator

Cosco Corp

Swiber

Yangzikiang

China ADRs:Tracking The NASDAQ_SGX China ADRs

2010-09-18T13:07:00.007+10:00

1. List of 19 SGX_NASDAQ China ADRs
2. How they cluster on SOM of S&P500

3. Statistics of each cluster

4. High PE of Cluster S3: CEA and ZNH

5. 12-m return % map

6. High Price/Sales ratio of Tech ADRs

6. Beta map

7. Forecast 1-m return %

For explanation on technology, methodology and terminology, see: http://www.technifundamentals.com/2010/07/explanation-page.html .

The Singapore Exchange [SGX] has announced that in an Agreement with NASDAQ, from October 22, nineteen China ADRs can also be traded on SGX. The list of these ADRs is in image 1. *All the fundamental data on the 19 ADRs are extracted from ValuEngine, and if you subscribe to their service, you will be able to track the ADRs in greater detail. The key question is whether, since these ADRs are also listed on the American market, will their performance be heavily influenced by performance in that market? As we have seen in a previous post, ADRs of emerging markets, as they grow bigger, will mature and behave just like any other global Blue Chip. So if you are looking for outperformance, it is best to look for those ADRs which not only have room to grow, but are not yet behaving like the big caps, maybe because of the different type of investors they attract.

1. Image 2: Contrary to what one may think, the 19 China ADRs have a wide range of characteristics in their fundamentals and technicals. Plotted against the backdrop of the S&P500, they are present in each of the three clusters. Stocks in S1 and S2 are more typical of the S&P500 component stocks, while S3 stocks are more different. While the average Beta of the ADRs is low [below 1.0], ADRs like China Aluminum, Yangzhou Coal Mining and Suntech Power have Betas >2.80.

2. There is only one stock on S2 and that is STP Suntech Power the solar energy Company. The other solar energy Company TSL Trina Solar, although about the same size in market cap as STP, is in S1 and has more similarities with YZC Yanzhou Coal Mining.

3. The most differentiated cluster is S3, and it contains the two Chinese airlines CEA China Eastern and ZNH China Southern. There are some indications that these two stocks will have higher performance in the short term. Their P/Es are very high [see image 3 the P/E bar], but their valuation % is low. Their 12-month return % has not been impressive but their 1-month forecast return % is highest among the 19 ADRs. Finally, and unusually, their Beta is also very low: 0.51 and 0.62 respectively
4. As is to be expected, the tech Companies like Netease, Shanda and Mindray medical have high Price/Sales ratio. In terms of stability, these are also the stocks whch have high 5-year Returns: Neteast, Shanda, Mindray, Ctrip, Home Inns. Changyou and Baidu. Unfortunately we do not have market/book data on these ADRs.

5. In summary, if you have intention to invest in these ADRs, it is better to focus on those that have not performed as well, are less similar to the S&P mainstream stocks, and are still Undervalued. You can see from the difference in length of bars of model variables in image 3, that only ZNH and CEA are significantly different in characteristics from the S&P 500 components. S2 ADRs are not so different. [length of bars represent standard deviation from the Mean of the whole data set. S1 approximates the S&P500. This is verified by the fact that its Beta is 1.011 where S&P500=1.0]. Note# with greater liquidity and participation when they are tradeable in Singapore and available to more Asian investors, the characteristics may change, and then these 19 ADRs will be less attached to the American market.

Note 2# The SOM above was constructed with long term fundmental and technical variables. For those who would like short term technical analysis of each of these ADRs can send me an email.

The Difference Between Stock Selection For Short Term And Long Term

2010-09-11T16:37:00.008+10:00

1. Rahul-Mohindar Trade Model for the DJIA
2. Stocks selected by ValuEngine Valuation Model

3. Location of Undervalued Stocks on the SOM

4. Momentum of Undervalued Stocks

5. Undervalued Stocks Cluster In S3

6. Alpha/Beta of Singapore Stock: Straits Asia Resource

7. Alpha/Beta of Solarfun Power

8. AlphaBeta of KKR Financial Holdings

9. Alpha/Beta of Sanmina-Sci

For explanation on technology, methodology and terminology, see: http://www.technifundamentals.com/2010/07/explanation-page.html .

Have you ever wondered why stocks which are undervalued can test your patience with regard to their performance? I mean when you read recommendations by broking houses and analysts....

This is because stocks can be out of equilibrium with their fundamentals for a very long time. This is due in part to the daily market noise caused by short term traders who play by momentum and liquidity, and by the over/under reaction caused by the availability of instantaneous information. It is now even worse when high-frequency algorithmic trading accounts for 30-60% of trade volume, depending on the Exchange. We take a look at stocks selected by ValuEngine's Valuation model (the nodes of the SOM that are labelled 'vl') and examine their longer term fundamental characteristics as well as their short term technical characteristics.

Image 1: A technical outlook for the DJIA after yesterday's Close using the Rahul Mohindar technical analysis model shows that the DJIA is still in the Bullish zone, and the upward move is still intact as indicated by the Blue-colored price bars as well as the Blue arrow. Third window from top also shows the most recent bar is still above the Red take-profit line. So we can go ahead with stock selection for Long.

Image 2 are the stocks selected by ValuEngine valuation model and the Pink highlighted stocks are shortlisted for our technical analysis.

Image 3 shows the location of the stocks on the SOM, [South-East corner and labelled 'vl']and the scale at the bottom of the Valuation map shows Valuation +/- % from Blue to Red. So the undervalued stocks are dark Blue.

Image 4 is a map of the Momentum attribute and the more Red the higher the momentum. We can see that our Undervalued stocks have a high Momentum. But this momentum is a long term momentum of 12-months rate of change.

Image 5 shows that L stocks of the ValuEngine Valuation, Growth and Quality models are neatly positioned in cluster S3 thus confirming that the cluster is homogenous.

Images 6-9 plot the 21-day Alpha and 21-day Beta of a stock. *Refresher: Alpha is the intercept of the linear regression line of a stock with its market Index and Beta is the slope of that line. Roughly speaking, Alpha level is an indicator of a stock's 'resilience' versus the market. And Beta is an indicator of a stock's sensitivity to the market's movements. The choice of period for Alpha/Beta determines the values you get and the volatility of Alpha/Beta. There are many dis-advantages for using such a crude measure as Alpha/Beta, among the most important of which is that it assumes linearity in the markets. But Alpha/Beta is the most direct way of measuring changes, and gets right to the root of the matter unlike traditional technical analysis indicators. For this article U.S stocks are measured against the DJIA and the Singapore stock was measured against the Straits Times Index (STI). In the charts, the 'price chart' of the Index against which the stock is measured is superimposed on the stock's rpice chart.

Image 6 shows the Alpha/Beta of a stock on the Singapore Stock Exchange: Straits Asia Resources. This stock was selected based on a technical analysis screen-nothing to do with fundamentals. Note the high violatility of its Alpha/Beta lines and how as at yesterday Beta is going down while Alpha is going up- a strong sign of resilience to market movements.

Image 7,8,9 shows stocks selected by fundamentals criteria of Valuation: Note that their Alpha/Beta lines are much more tame compared with the Singapore stock. This is despite the fact that these stocks not only are undervalued, but have high momentum and Beta [see Beta column in image 2] too. It's just that the time period selected for calculating momentum and Beta is a long time viz one year.

Now we have to wonder if investing strictly by fundamentals is a good idea in the light of core changes happening in the investment environment. With derivatives,algorithmic trading, 1000 trades a second, and the inter-connectedness of all financial markets, your stocks never have a chance to get near the fundamental valuation they have. And you don't have the time to wait and wait. The analysts can write all they want about P/E ratios, M/B Ratios, Cash Flow and so on but in the short term they all don't matter. And as for the long term, as Keynes said, "in the long run we are all dead". Is Warren Buffet style of investing for dinosaurs? Why not forecast the direction of movement just for tomorrow, and if no earthquake, political scandal or North Korea cause an external shock to the system, then tomorrow should be fine. Stay tuned to this Blog.....

S&P500 vs DJIA vs ADRs

2010-09-04T13:24:00.004+10:00

1. Where are the ADRs on our SOM?
2. Turbo-charged ADRs in Cluster S4

3. ADRs that behave like DJIA Blue-Chips

4. Turbocharged S4 ADRs have high Beta

5. But Turbo-charged S4 ADRs have high 12-month Return % too

For explanation on technology, methodology and terminology, see: http://www.technifundamentals.com/2010/07/explanation-page.html .

Let's leave the weekly market outlook aside for the moment, because the forecast horizon is as cloudy as ever. But continuing from last week's theme that DJIA stocks are widening their gap with S&P500 and slowly forming a cluster of their own with its unique characteristics: This week, we look at ADRs. Now, there are ADRs and there are ADRs. With globalization and the strong growth of the emerging economies, their bigger ADR's are becoming a part of every big fund manager's portfolio which by necessity must be a global portfolio. The big market cap and high liquidity ADRS are starting to take on the characteristics of the DJIA. [In image 1, you can see how closely positioned they are to the DJIA stocks ['d's']. So this type of ADR have become mature and do not have the volatility they had when they were 'young'. Among these ADRs are Vale, Banco Santander, Amer-Mobil and TEVA. [see image 3]. Then there are smaller cap, more volatile ADRs with higher growth potential, but also higher Beta and volatility. For these, see image 4.

For this week I constructed a SOM with the S&P500 as a backdrop. And on this backdrop, I placed components of the DJIA with the label 'd'. Then, about 500 ADRs were filtered for Average Daily Volume of last 3 months greater than 100,000. This reduced the number of ADRs to 285, and the SOM positioned them according to their combination of fundamental characteristics [model variables]. The ADRs retained their ticker symbol. In addition, ValuEngine's Long and Short screens for their Valuation, Growth and Quality models were also included for the SOM's calculation. Labeled with 'l' for Long and 's' for Short, they automatically formed Long and Short clusters on the SOM, and by looking at the ADRs location in a certain cluster, we can make inferences on it's characteristics. To summarize:

1. Image 1 shows the clusters of the SOM and the ADRs spread over the four clusters. S1 is where most of the S&P500 stocks are. There is a large group of ADRs there, but S1 is an ambivalent cluster with 'l' and 's' equally represented, and sparsely positioned. S2 is also generally a Short cluster with the additional comment that there are a lot of 'd' DJIA components there. S3 is a Short cluster. S4 is a strong Long cluster (see the 'l's'), and has a group of ADRs in it.

2. Image 4 shows that these smaller ADRs have high Beta [see scale at bottom of map] Contrast with the DJIA-like ADRs in the North-East corner, with their lower Beta. [These ADRs have darkened nodes]

3. Image 5 shows that even though they are more volatile, these smaller ADRs have a very high 12-month return %. For 12-month return % of big and small cap ADRs, you can see the difference in the 12-month return % column of image 2 and 3.

LESSON OF THE DAY: The world is indeed becoming a smaller place, and soon it doesn't really matter where a Company is physically located, or on which stock Exchange it is listed. The real difference is between having an international business or a domestic business. Big cap ADRs are maturing and assuming the characteristics of DJIA components. They are stable and will grow, but at a slower pace. To catch the next wave, move on to the smaller ADRs. Look in our Self-Organizing Map to discover where the gold fields are.

Afterthought: Fundamentals cannot forecast the short, short term, like tomorrow or the next 2 or 3 days. And often, we are more interested in the short, short term. For that, we are working on a Proof of Concept Analysis for a new paradigm in short-term forecasting. Stay tuned to this Blog

Spread Between DJIA and S&P500 Will Increasingly Widen

2010-08-28T12:26:00.008+10:00

1. DJIA-S&P500 spread increasingly widens
2. DJIA stocks and some basic statistics

3. DJIA stocks cluster in our Self-Organizing Map

4. Long outlook for this week is not as strong as Short outlook

5. Cluster statistics:Difference between Long and Short clusters widen

For explanation of methodology and terminology, see http://www.technifundamentals.com/2010/07/explanation-page.html

For this week, let's abandon our usual format for analysis. After all, it will be more of the same in terms of market outlook- rangebound. But if you look at image 3 which shows that S2 is the Long cluster and S3 is the Short cluster, and then look at image 4 which shows that Red area denotes Long while Purple denotes Short and Green/Yellow is wishy-washy, you will find that most of S2 is Green/Yellow. This means that Long signal is not as strong as Short signal. Will there be a melt-down? As at now, [excluding chain-linked events which may yet emerge during the week] the danger is not significant, though probability of intensity of Bearishness has increased during this week. As shown by difference in length of bars of model variables of clusters S2 and S3 [see image 5].
But the topic of discussion this week is about what will happen to US stocks for the months ahead. There is one clear trend that has emerged and should not be ignored: THERE WILL BE A WIDENING OF SPREAD BETWEEN THE PERFORMANCE OF COMPANIES WITH INTERNATIONAL BUSINESS [WHETHER DIRECT EXPORTS OR WITH OPERATIONAL FACILITIES OVERSEAS] AND COMPANIES WHICH STILL DEPEND ON THE U.S. DOMESTIC MARKET. This should be the main criteria when analysts do their analyzing. It is difficult for me to get direct data of a Company's percentage of exposure to international business. But a good proxy is to look at the components of the DJIA. Everyone of them, a household name internationally. With their brand name, big market cap and established status, it is safe to assume that they have ventured abroad, and that international business comprises a significant proportion of their revenue and profit. If that is true, then the spread between the S&P500 and the DJIA should be widening? Is this true? Image 1 shows that the evidence is irrefutable. Since 2005, the big blue chips of the DJIA have increasingly moved away from the S&P500 [of course the S&P500 also contains all the 30 component stocks of the DJIA, but out of the other 470, many are still largely dependent on the domestic economy. This is a clear wake-up call for American companies to venture abroad NOW! They should get all the help they need to venture out into the world, and the U.S. government should extend all the help that these Companies need.

As for investors take a look at image 3 where S1 is the cluster that represents the S&P500 stocks. Note that all the DJIA stocks are also in this cluster [meaning they will have the characteristics of this cluster]. But the DJIA stocks are very closely positioned to each other and merit forming a sub-cluster by themsleves. If I increase the resolution of the SOM, there will be a fourth cluster comprising the DJIA stocks. This sub-cluster will become increasingly distinct in its characteristics and behavior. [Except for Alcoa, Bank of America and Travelers, which seem to be different kind of stocks as evidenced by their locations on the SOM] So if the market does a melt-down, either buy the DJIA components or ETFs of Asian stock markets like Singapore, Hong Kong, China, Indonesia, India or Brazil. [The fundamental statistics of the DJIA components were obtained courtesy of ValuEngine, and if you subscribe to their ValuEngine Institutional package, you'll get access to all Valuation, Growth and Quality models].

Market Outlook plus Paper, Packaging & Wood-Based Industries

2010-08-21T15:54:00.010+10:00

1. S1=Approximately S&P500, S2= Long S3=Short
2. Cluster Statisics: EPS Surprise and P/E are main difference between Longs and Shorts

3. Cluster Statistics Summary

4. Darkened Selection Area= High EPS Surprise %

5. Selected Stocks From Paper & Packaging Industry And Machinery

Rahul Mohinder Oscillator of International Paper

7. RMO of Meadwestvaco

8. RMO of Office Max

Explanation Page of Methodology and Terminology is at http://www.technifundamentals.com/2010/07/explanation-page.html

Again, it's a toss-up which way the market will go next week. But I would stick my neck out and say there won't be a melt-down, despite the ominous looking Rahul Mohinder Oscillator (RMO) in the S&P500 chart. Image 2 Cluster Statistics shows that in general the difference in length of bars of S2 (L cluster) and S3 (S cluster) is not great. The difference in Beta between S2 and S3 is also minimal [S2 Beta=1.80, S3 Beta= 1.66]. So the market has less anxiety.

EPS Surprise and good P/E will again be the dominant themes for investors. I used image 4 to select the high EPS nodes on the SOM, and the list of stocks in the selected area is in image 5. One piece of information stands out- and that is, four of the stocks are from the category Forest Products Industry or Chemicals. If we read the business profiles of MVW,IP, OMX and DOW, they are all Companies involved in the paper, packaging and wood-based specialty chemicals business. Now if we take a look at the RMO charts of MVW, IP and OMX, (images 6,7,8 above) we find that all three are still in the Bearish zone with the RMO (Green color top chart) below zero, though the charts also show some strengthening in the Price, as indicated by their Candlesticks. MVW is the strongest of the lot with a lower Beta of 1.77 versus 2.31 for IP and 2.68 for OMX.

Could our SOM and ValuEngine model be a leading indicator of whats in store for these stocks? We'll keep in mind their market price as at Friday, and see how they do next week, taking into account the market's direction.

As for Deere (DE) and Cummins (CMI), perhaps investors are beginning to realize that these Companies are increasingly export-oriented and their products are in demand by developing economies building up their infrastructure and their industries.

Technicals Look Bad (contd. from previous post]

2010-08-14T16:02:00.007+10:00

S&P500 Dillards- Consumer Services

Dow Chemical-Basic Industries

4. Sandisk- Technology

5. Tenet- Health Care

In the post before this, we selected the stocks based on their fundamentals. [please read previous post]This post will look at the technicals of some of the selected stocks. The purpose is to show that there is a time to buy, and a time to stay out of the markets. With globalization and the proliferation of high frequency algorithmic trading, all financial market are highly correlated, and the average Beta of stocks has risen. It is probable that 70-80 % of a stock's price movement is dependent on the general market no matter how strong its fundamentals. We use the Rahul Mohindar Oscillator system which is available in the latest version of MetaStock, the technical analysis software to illustrate this point. Technical Analysis is not magic and does not have prediction value. But the sophistication of present day technical indicators provide very useful visual analytics. The RMO has three Rules for calling a Buy and a signal is only valid when all three rules are fulfilled.(1) The Oscillator [top Green] must be in positive territory i.e. greater than zeroAND (2) The price chart must be in Bulllish zone i.e. Blue and not Red AND (3) The most recent Buy arrow must be Blue. [For Sell signals, do the opposite].Take a look at the images above starting from top:

1. S&P500- Green oscillator is about to go below zero. Most recent arrow on price chart is Red. The beginning of the Bear zone?

2. Dillards has been red for some time although its valuation is -41 % and the EPS Surprise % is a high 107 %

3. Dow Chemical has also only recently gone into the Bearish zone.

4. Same for Sandisk the tech stock

5. Tenet is the same story, only worse.

So, whatever the fundamentals, there is a time to buy and a time to stay out of the market. The price movement of nearly all stocks on NYSE and NASD are highly correlated with the market (Beta >2) and in fact, although I don't have the space to show, the RMO of nearly all markets (Shanghai Exchange, Hong Kong, Singapore, Commodity Research Bureau Index, US Dollar Index, 10-year Treasuries, Gold) paint a similarly depressing picture. Gold and Treasuries are not yet going down, but at the same time they are not exactly moving up too. It may be a deflationary outlook, but with the quality of U.S. debt in doubt, Bond yields may yet go up. As for currencies, all the major currencies: Euro, USD and Yen have a lacklustre economy behind them and any one currency's rise is more by default and being the best of the worst by rotation than for any substantial reason. The big question is whether the Asian economies (China, India, South-East Asia) have enough dynamism to form a bottom for the rest of the world. In that regard, it may be useful to monitor Shanghai, Hong Kong and Singapore markets for signs of a bottom. * Baltic Dry Index is going up, but it is too volatile on a day to day basis for it to be a meaningful indicator. Note# In the RMO, the third window from the top shows a Blue line and a Red line. If you are Long on a stock and a bar crosses the Red line on the way down, it is a Stop Loss signal. Looking at the Stop Loss Indicator of the S&P500 and all the stocks, we should have been out of the market some time ago.

EPS Surprise % Drives The MArket

2010-08-14T12:56:00.007+10:00

1. Clusters: S1 approximates general market:S2=Long, S3=Short 2. Cluster Statistics

3. High EPS Surprise % Stocks [HEPS]

4. Darkened selected area = area of high HEPS stocks

5. HEPS stocks do not have high 1-m forecast return%

6. HEPS stocks score do not have good long term performance track record

7. HEPS stocks are smaller market caps

8. HEPS stocks are lower priced

9. HEPS stocks fundamentals show low M/B Ratio

10. Lower P/E Ratio too

11. HEPS stocks have High Beta

12. HEPS stocks have good valuations

13. High Volatility too

14. But they are not the top in terms of daily average volume

Explanation page on methodology and terminolgy is at http://www.technifundamentals.com/2010/07/explanation-page.html

Market Outlook
The S&P500 fell by 4 % during the week from 1124 to 1080. The market was indecisive, buffeted between good news and bad news, suffering the half-empty, half-full glass syndrome. We can expect more of the same this week with a slight tilt towards Bearishness. Cluster S1 which approximates the S&P500 contains 78.10 % of the SOM's stock population and has a Beta of 1.13 which verifies that it is a good proxy of the S&P500 [since the Beta of the S&P500 is 1] The nodes on the SOM occupied by S&P500 components are labeled with "I". S2 which is 12.60 % of the stock population contains most of the L stocks. S3 with 9.3 % of the population contains most of the S stocks. The clusters are small, tight and well-defined which makes prediction more valid. Image 2 cluster statistics show S3 being more differentiated from the market than S2 [model variable bars are longer] and indicates Bearishness to be more pronounced. The Beta of S2 (L cluster) is 2.64 while the Beta of S3 (S cluster) is 1.53. This indicates that L stocks are more sensitive to changes in the general market than S stocks. So it may be safer to go Short than to go Long. The main theme driving the market continues to be EPS surprise. Therefore we will look at the EPS surprise stocks as selected by the SOM, and their characteristics. * High EPS surprise means historical high EPS surprise since by definition we cannot know the high EPS surprise beforehand. In image 4 we select with the cursor, the area on the SOM with nodes of high EPS surprise- and this is the darkened area. Therefore from images 5 to 14, the individual attribute windows of our stock population are shown, and self-explanatory comments given on the darkened area of high EPS surprise stocks.The stocks selected on our SOM darkened area are shown in image image 3. For this week, we are going to do something different: Since the market is out of whack and schizophrenic, and any semblance of equilibrium based on fundamentals is lacking for the short term, we will use technical analysis to have a look at the selected stocks. I have never been a great fan of technical analysis but with the proliferation of machine-controlled high frequency trading, it is now something not to be ignored. Also, TA is now much more sophisticated. With the influx of a new generation of practitioners, we now have comprehensive decision-making systems based on IF, THEN rules instead of stand-alone indicators. One example of this is the Rahul Mohinder Oscillator system which is available with the latest MetaStock. For an introduction to the RMO, go to http://www.forestbergen.com/download/Metastock/MasteringMetaStock10_CH6-RMO.pdf . We shall use the RMO to analyze some of the selected stocks from the different sectors : Expedia (EXPE) Dow Chemical (DOW), Textron (TXT), Tenet Health Care (THC), Sandisk (SNDK), and Dillards (DDS). This will be done in the next post, (above this post) as images are getting too many for this post.

A (Long) Colorful Tour Of The Stock Market

2010-08-07T12:46:00.017+10:00

0.Stocks

1.Value Area2. Quality Area

3. Growth Area

4. Short/Long Areas (Interpolated)

5. Clusters

6. Cluster Statistics

7. Average Volume

8. Beta

9. EPS Surprise %

10. E/P Ratio

11. Last 12-Month Return %

12. Market Cap

13. M/B Ratio

14. Sharpe Ratio

15. Valuation

16. Volatility

What can you do when market direction is very uncertain? Last week the S&P500 advanced by a mere 0.5 %. It could easily have been down by 0.5 %. Our prediction was not inaccurate. Trying to predict the direction of the market is futile at the moment. So, for this week, we'll take a long colorful tour of the market via the sixteen images of our Self-Organizing Map model above and determine just what kind of stocks have a better chance when better times resume. Before I forget, the explanation page for our methodology and terminology is here:
http://www.technifundamentals.com/2010/07/explanation-page.html . The images 1 to 16 above are in no particular order of relevance to our discussion, and so I'll refer to them by their number when the discussion requires an image to be referenced.

Market Outlook
Image 5 denotes the self-organized clusters: S1 is where most of the S&P500 stocks reside (72.97 % of the stock population), S2 is the Long cluster (14.67 %) and S3 the Short cluster (12.36 %). Image 6 cluster statistics shows that the S stocks are of a greater degree of dis-similarity than the L stocks, as measured by the length of their model variable bars, which is a hint that Bearishness is more pronounced than Bullishness. However, after interpolation, in image 4 the Red (L) area is about the same size as the Blue (S) area for S1 where most of the S&P500 stocks reside, and this is an indication that market direction will be very sensitive to unforeseen events during the week. Images 1,2 and 3 show that no particular type of stock- Value, Growth or Quality dominates market action. The significant Green, Yellow, Orange areas in images 1,2,3 denote that areas of V,G, and Q overlap (especially between V and Q). This is another factor adding to the uncertainty in the market.

So what are the characteristics of the stocks in the L cluster (S2)?

a. These are stocks which have a history of big EPS Surprise % [see image 9 the stocks in the Red,Orange, Yellow nodes of the SOM]. For the rest of the analysis, fix your gaze on this area of cluster S2. The other attributes are examined in relation to this concentrated area of cluster 2 where the L stocks of the ValuEngine screens are located.

b. These are not the big Blue Chips [see image image 12 Market Cap where this area is mostly Blue].

c. More volatile: Image 16 shows these are the more volatile stocks.

d. But they have good valuations. See image 15 where undervalued stocks (negative valuation %) are in the Blue areas.

e. However their Sharpe Ratio is not high. That is going by the Returns/Standard Deviation of the last 5 years, they performed poorly. See Image 14 Sharpe Ratio, the Blue area.

f. In terms of M/B Ratio, they score highly [image 13 shows low M/B Ratio Blue areas of cluster 2.

g. In terms of E/P Ratio, (the inverse of P/E) they also score highly . See image 10 E/P Ratio where 'hot' Red, Yellow areas are in cluster 2 too.

h. In terms of last 12-month Return %, its almost half-half. Some of these stocks have had high 12-month return % and some have had low 12-month return % See image 11.

i. Mostly, these stocks have a high Beta, and are sensitive to the direction of the SP500 Index. See Beta image 8.

j. These stocks are not the high daily volume stocks too. See Image 7

SUMMARY

So the kind of stocks that will do well in the current market situation are mid-cap, lower volume, higher volatility, but with good fundamentals such as low P/E and M/B which have not had their turn in the market. No particular sector nor particular type V,G or Q dominates this list of stocks. Image 0 at the top shows this list of stocks. Where they are repeated in the list, it means they have attributes corresponding to more than one Type V,G or Q.

MOCABI Prototype 6 (Contd)

2010-07-31T16:38:00.005+10:00

1. Clusters
2. Cluster Statistics

Cluster Statistics Summary

4. Sectors and Type: Cluster 3

5. The difference between P/E Ratio and M/B Ratio

For explanation on methodology and terminology see: http://www.technifundamentals.com/2010/07/explanation-page.html

Reflections on MOCABI

Last week we predicted a positive change in the S&P500, but cautioned that it was a very weak signal. As it turned out the S&P barely budged, going from 1101 to 1104 for the week. But I am beginning to get around to the view that, for short term predictions, predicting the direction of a stock is easier than predicting the direction of a market Index. Short term predictions viz from 1 to 3 days, are more determined by technicals like Momentum, Liquidity and Volatility than fundamentals. Each stock has its 'fingerprint', its characteristics as moulded by the type of shareholders the stock has [which in turn determine their investment style], the nature of the Company's business, the sector it is in, the cyclical nature or trend of the trades on this stock etc. Some stocks are more predictable than others, and if you can identify such stocks, your short term prediction based purely on mathematical techniques, might carry through if during the fleeting time window of opportunity no big external shocks rock your prediction. For that reason, this project may in future stress on short term predictions of 'predictable' stocks. How do we identify such stocks? By using our good old friend the SOM to first measure degrees of similarity between all stocks, and clustering them. In that way, making a prediction for a homgenous group is easier. This is also why it is difficult to predict the direction of an Index-because it consists of a group of non-homogenous stocks. [Index stocks are by definition non-homogenous because they are supposed to represent the broad market].

Market Outlook

Despite what we say above, the analysis of Market Outlook using screened stocks of the ValuEngine Model screens is still pretty good. For this week:

S1 is the cluster that contains most of the S&P500 stocks and is 71.21 % of the stock population of our SOM. S2 contains most of the S stocks and is 14.77 % of the population, while S3 contains most of the L stocks and is 14.02 % of the population. The ratio of L/S is 0.62 in S1, 0.19 in S2 and 36.0 in S3 i.e. S3 contains no S stocks. Thus, the S&P500 will be on the weaker side, but stocks like those in S3 will outperform the Index. Again, the market outlook is ambivalent. Image 2 arrows show that factors like P/E Ratio and EPS Surprise will be inportant factors.

Stocks

Image 4 is revealing. This list of stocks are the screened ValuEngine stocks which appear in Cluster S3. They are all Long [as indicated by the L tag of their label] and out of the 36 stocks, 20 are Value stocks, 12 are Quality stocks and only 3 are Growth stocks. This is a great contrast to last week when most of the stocks in the L cluster were Growth stocks.

Another interesting fact can be gleaned by looking at image 5 which shows P/E and M/B windows of the SOM. The color scale at the bottom of the windows measures the P/E and M/B values. The lower the better, i.e. the more tending towards Blue the more desirable. Note that low M/B does not equate with low P/E. Some Blue areas of the M/B window are in fact areas of higher P/E as shown by their Yellow/Orange colors. Which means that more tangible assets do not necessarily translate in to higher earnings [think technology and those companies where the Company's value is its intangible assets]; or that such assets have been inaccurately valued [think banks where book value of loans are still at optimistic (or unrealistic) valuations]

Market Outlook cum AlphaBeta Indicator Prototype 5 (contd)

2010-07-24T13:44:00.005+10:00

0. Stocks

1. Clusters 2. Cluster Statistics

3. Short/Long Areas of the SOM

4. Growth Type Area

5. Quality (Risk Aversion) Type Area

6. Value Type Area

This week, we need to do some major re-thinking. The S&P500 was up 2.9 % for the week while we had predicted a Bearish market outlook. I realize that I was looking at the trees instead of the forest, though the prediction was wrong mainly because of the effect of Earnings Surprise % during the week . Since we are attempting to predict the direction of the S&P500 Index, we should be concentrating on the area(s) of the SOM occupied by the S&P500 components instead of the area (nodes) occupied by the ValuEngine screens. [BTW, in case all this sounds unintelligible, the explanation on methodology and terminology is at http://www.technifundamentals.com/2010/07/explanation-page.html ]. With that in mind, let's proceed.

Market Outlook

The first point to note is that S1 which is the cluster containing the least number of ValuEngine screened stocks [and therefore contains the most number of S&P500 omponents] has only 58.26 % of our stock population. S2 which contains most of the S stocks is 21.23 % of the stock population, and S3 which contains most of the L stocks is 20.51 %. This means that any prediction of market outlook will be less reliable. Previous week's % of stock population in areas clusters occupied by S or L stocks was under 10 %. I have added a new label 'I' to indicates nodes occupied by S&P500 components, and you can see that although most of the I's in image 1 are in S1, quite a number of 'I's are also in S2 and S3.

The second point to note is that we should not be comparing S2 and S3 length of bars with S1 length of bars when trying to predict the direction of the S&P500. While S1 is the closest approximation of the S&p500, it is NOT the S&P500. The S&P500 on the SOM by definition will have cluster statistics where all the bars of the model variables will be zero standard deviation i.e. length of all bars will be zero. But the best we can do is to use S1 as a proxy, just bearing in mind that the shorter the length of bars in S1, the more accurate the prediction. If we look at image 3 which shows Bullish areas (Red) and Bearish areas (Blue) we can see that in S1, the Bullish area is slightly bigger than the Bearish area. So the outlook for the market is Bullish for next week. But the signal is a weak one. Again, much will depend on Earnings Surprise %

Individual stocks

The L stocks from the ValuEngine screens are mostly in S3 and the S stocks in S2. However, as explained earlier, many non-screened stocks are also in S2 and S3 which makes the prediction for the screened stocks less reliable. This is seen in images 4,5 and 6 which show the areas occupied by the Type V, G and Q stocks. Red means presence of the Type, (value=1) while Blue means absence of the Type (value=0). However, a SOM is capable of interpolation between 1 and zero taking into account the values of the neighborhood of the nodes, and so we have non-Red, non-Blue areas too, e.g. Green, Yellow. The more the Green/Yellow areas, the less clearly defined by Type. Thus we see that V and Q have larger areas of Green/Yellow, while G is more clearly defined. From this, we can generalize that G stocks selected by the ValuEngine G screen will be more reliable to base decisions on. This also indicates that the dominant characteristic of the market next week will be an emphasis on Growth stocks. Which growth stocks? See image 0 (except Whirlpool)

* Note I have put aside Alpha and Beta for the moment while I am working towards a quantitative interpretation for Alpha.

Market Outlook cum AlphaBeta Indicator Prototype [4] (contd)

2010-07-17T13:14:00.007+10:00

1. Self-Organized Clusters 2. Cluster Statistics

3. Short/Long Window

4. EPS Surprise % Window

5. Beta Window

6. Last 12-Month Return % Window

7. Market Cap Window

8. Selected Stocks

The S&P was down by about 1 % last week while we had predicted a mild Bullish outlook. Most of the down action happened on Friday (-2.88 %), so our prediction was not such a disaster. This week, we have made some improvements to the model. The attribute windows are now much more defined [more colorful] as we plotted the transforms of the attributes instead of the original values. Defining nominals and using interpolation we can now plot Short/Long as well Type [V,G,Q] on the SOM where 1= the presence of a defined nominal and 0= absence. Just to remind that the explanation of methodology and terminology is at http://www.technifundamentals.com/2010/07/explanation-page.html . And also to remind that the difference between our indicators and ordinary technical analysis indicators is that our technicals are based on fundamentals viz using ValuEngine pre-screened stocks. That's the reason for the name of this Blog (Technifundamentals). Now, being based on fundamentals means that it's not unusual to see selected stocks fall, as short-term noise clouds the fundamentals, only to rise again the week after. This is the case for one of last week's stocks- Gannet GCI-which despite earnings of 61 cents per share , up from last year's 30 cents a share and beating analyst's estimate of 53 cents a share- dived 10.7 %. ValuEngine screening criteria also include a minimum market cap and minimum average daily volume, so that leaves out the smaller and more illiquid stocks and reduces volatility.
Market OutlookIn image 1, the L stocks are mostly in cluster S2 but a number of them are also in S3 which has a majority of S stocks. This 'dilutes' the strength of the signal, and we have a rather ambivalent market outlook next week. This can be seen in image 2 cluster statistics where the bars for Short/Long are negative in S1 [which contains most the S&P500 components] and S3 and only positive in S2. In image 3, Red area is Long while Blue area is short. Red area is slightly bigger than Blue area, but Blue area is more defined, with S stocks bunched in a sub-cluster i.e. nodes of S stocks are closer together. On the other hand besides the sub-cluster of L stocks, the other L stocks are spread sparsely over the Red area. On a SOM it indicates a less homogenous grouping. On balance the outlook is more Bearish than Bullish. Type-wise or Sector-wise there is no clear pattern. V, G and Q stocks as well stocks from Sectors B,C,D, E,F,H,ND,S,T,TP,or U are distributed almost randomly over the topology of the SOM. Also, comparing the length of the bars in S2(L cluster) and S3(S cluster), in general the S3 bars are longer. Since the length of bars denote the standard deviation from the Mean of the entire data set, this implies that Bearish indicators as represented in S3 have a stronger signal strength than the Bullish indicators in S2. S3 Lastly in S1 [Blue cluster] which most approximates the Index, the ratio of L to S stocks is 8/26 or a very low 0.30.

Alpha/Beta

EPS Surprise % will still play a big part in selection of Long stocks. Earnings Surprise % here is of course ex-poste i.e. a stock's past record in Earnings Surprise i.e. deviation from analyst estimates (Please see the EPS Surprise % bars in image 2]. And here we are assuming that a stock that has a good record of springing Earnings Surprise is more likely to do so again. Using the EPS Surprise attribute window (image 4) to manually select the nodes [darkened areas] with the highest EPS Surprise %. we find that the stocks selected have certain attributes: they have a high Beta [image 5], high 12-month Return % (Momentum) [image 6] and smaller market cap [image 7]. We can sum up by saying that it's the mid-caps with high Beta and momentum that will show the greatest Earnings Surprise %. The average Beta of S2 Bullish cluster is 2.13 while the average Beta of the S3 Bearish sector is 1.71. i.e. the risk is greater on the Bullish side. Other than Earnings Surprise % there are no other signifcant factors contributing to a stock's Alpha at the moment. ValuEngine's Senior Analyst Steve Hach showed me an article titled "Has Alpha Turned To Beta?". That about sums up the market at this point of time. If you must play the market go do something safer and less volatile-like buying a reverse ETF of the S&P500.

Market Outlook cum Alpha/Beta Indicator Prototype (3)[contd]

2010-07-10T23:39:00.004+10:00

1. Self-organized clusters formed by ValuEngine model screens
2. Beta window of the Self-Organized map

3. Statistics of the clusters S1, S2 and S3

4. Summary of cluster statistics

5. Selected stocks from the Long cluster S3 The Market Outlook cum Alpha/Beta Indicator [MOCABI]is still undergoing testing although I am slowly nailing down some of the issues. Last week, I predicted that the selling had abated and there would be a mild rebound. Well, maybe I was wrong in using the word 'mild'. The S&P500 increased by 5 % or so for the week. Let's see how we do this week. I'll try to make the analysis as quantitative as possible. The Explanation Page for MOCABI methodology is at http://www.technifundamentals.com/2010/07/explanation-page.html.

1. S1 holds the S&P500 stocks, S2 most of the Short (S) stocks and S3 most of the Long (L) stocks. This is the same order as last week. The average Beta of S2 stocks was 1.60 last week and is now 1.50. The average Beta of S3 stocks was 1.98 and is now 2.62. There is more 'sensitivity' to the Index on the upside than on the down side.

2. Last week the ratio of L stocks/ S stocks that are in the S1 cluster which holds the S&P 500 stocks was 28/21=1.33. This week it is 42/29= 1.44. This is an indication of increased Bullishness for next week.

3. S3, the cluster which has Bullish characteristics holds only 6.14 % of our stock Universe for this SOM while S2 the cluster with the Bearish characteristics holds 17.33 % . This point somewhat tempers the Bullishness of (2) above.

4. In image 3 , the length of the bars denoting our Model variables above measures the deviation of the cluster Mean from the Mean of the entire data set. Thus, the longer the bars, the more those stocks in the cluster with those bars will differ from the performance of the Index as represented by S1 bars. The longest bars in S3 with the L stocks are Beta (first Green bar) Momentum [12-m return%] fourth bar Mauve color and Volatility (Red bar). In a Bullish market, high momentum, high Beta and high volatility means a sharp run-up. Also take a look at the Purple bar next to the Red Volatility bar, which measures EPS surprise % in the stock's history . Note its length in S3 as well as S2. EPS surprise will be an important factor in a stock's movement for the coming week. Thus the sharp run-up is based on fundamentals, particularly EPS surprise.

4. On the whole, it can be summarized thus: The market outlook is decidedly more Bullish this week as compared to last week. But only a few stocks will have move significantly. The stocks will be those with good fundamentals and the run-up will be speedy. The L stocks from the ValuEngine L screens will move up more than the S stocks from the ValuEngine S screens will move down.

5. Image 5 shows the selected stocks from S3 cluster. Those that are listed more than once are the output of more than one screen. Type V, G, and Q represents Valuation, Growth and Quality. The multi-listed stocks without V, G or Q are also components of the S&P500 Index. Note that most of the screened stocks come from the Q model and not the V or G model. So although the market will run on momentum, volatility and beta, it will be the Quality stocks [i.e. those with good fundamentals and risk/reward ratio] that run. This is confirmed by the statistics bars of S3 cluster for P/E, M/B (market/book) and P/S (price/sales) which are all significantly different (in a good way) from S1 the reference cluster and S2 the S cluster i.e. P/E, M/B,P/S and Valuation are all lower in value.

Market Outlook-cum-AlphaBeta Indicator Prototype(contd)

2010-07-03T16:17:00.009+10:00

1. This week's clusters

2. Last week's clusters

3. Beta window
4. Cluster statistics

5. Statistics summary

6. Stock list

Last week, our Market Outlook-cum-AlphaBeta Indicator [MOCABI] had a pessimistic view of the market, and this proved to be true. This week, we continue to explore MOCABI]. I have put all the explanation of methodology and terminology on a stand-alone page here: http://www.technifundamentals.com/2010/07/explanation-page.html.

There have been some changes in MOCABI and it will continue to have changes as it is being developed. This week's MOCABI is not a creation of a new model using this week's data. Instead, this week's data is applied by last week's model which has been refined by Dr. Gerhard Kranner of Viscovery. So we are monitoring the results. Also, at the suggestion of Dr. Kranner, Alpha should be defined and calculated in the conventional way. My 'unconventional' definition of Alpha [see explanation] I will call by another letter of the Greek alphabet, maybe Zeta. We shall then post stock lists as selected by high Alpha and high Zeta and compare their subsequent performance. We are also debating the interpretation of the SOM output and will refine the model accordingly. In the meantime, it's on with the show:

This week's market outlook

In times of low confidence and jittery nerves in the market, the average Beta of stocks rise. Thus, no matter how good a stock picker you are, for the short term, you are at the mercy of the market, and perhaps 80 % of a stock's performance is determined by its Beta. Comparing between image 1 and image 2, shows that there is not much change in the situation as evidenced by the three clusters, and their not-much-changed sizes. Most of the Long [L] stocks are in S3 but S3 only comprises 7.52 % of our stock Universe, while S2 which contains the Short stocks [S] comprises 19.63 %. The average Beta of S3 is 1.985 which is higher than the average Beta of S2 which is 1.604 so the L stocks in S3 are more sensitive to the overall market. S1 with a average Beta of 1.189 confirms that our SOM is well constructed- S1 which comprises 72.84 % our stock Universe, is approximately the S&P500 and of course the Beta of the Index is defined as 1.0. The high Beta of selected S3 L stocks is shown by the stock list in image 6. In such a wishy-washy situation, I would base my prognosis of the market outlook for this week, by looking at the number of L and S stocks in S1. There are 28 L stocks and 21 S stocks in S1. Thus the optimistic/pessimistic ratio is positive at 28/21= 1.33. Iwould venture to say that the selling has abated and this week will see a mild rebound. Is there any point in going on to select stocks even if they are high Alpha and low Beta? I do not think so. And there's no point picking stocks to Short too, as the Risk/Reward ratio for going Short does not justify it. Let's wait till next week when we see if the average Beta level has come down.

Explanation Page

2010-07-03T16:10:00.003+10:00

Explanation of MOCAB Indicator prototype.
Introduction
Stock markets are Complex Adaptive Systems [CAS] with all the unique properties of such systems. Properties of CAS such as feedback loops, emergence, self-organization, self-similarity, co-evolution and distributed connectivity require that a meaningful analysis of CAS be done with tools that take into account such complexities. To deal with the complexities of CAS requires a paradigm shift from hard computing to soft computing, from linear parametric modelling to non-linear non-parametric modelling and from exact solution to approximate solution.While most of the world’s phenomena (natural or man-made) are CAS, it is only in the last decade or so that with the exponential leap in the power of computers and software, we are able to use tools more suited to the analysis of CAS. Such tools include neural networks, fuzzy logic, evolutionary algorithms, wavelets,swarm intelligence and, in particular, Self-Organizing Maps Click www.viscovery.net/self-organizing maps for a short summary).Neural Networks are a class of Artificial Intelligence and a SOM is a class of Neural Network that mimic the biology of the human brain. Neural Networks are capable of associative memory recall, pattern recognition, classification, forecasting, optimization and noise filtering - which are all forms of generalization. In other words, neural networks learn from specific situations and apply the learning to new, hitherto unencountered situations. Most Neural Networks are supervised networks, i.e. they are ‘taught’ the relationship between input data and a target variable.On the other hand, a SOM is one of the few classes of unsupervised Neural Networks. A SOM does its work without having to be ‘taught’. The SOM algorithm self-organizes data of similar characteristics into clusters, very much like the brain does. The most useful feature of a SOM is that it can be used for the exploration, classification, analysis and visualization of large sets of multi-dimensional data. We can illustrate this point by taking an example from the stock market data of a Company. If data is available it is possible to plot the relationship between a stock’s price and its PE ratio as a two-dimensional relationship which can easily be visualized. A three-dimensional visualization is also possible if we have a third variable e.g the stock’s Price/Book Value ratio and three axes x,y,z to construct a 3D chart. But for complex models such as the ValuEngine models where close to 30 variables are involved, it is an impossible task to graphically depict the inter-relationships between all the variables. Only a SOM can do that. A SOM represents a perceptual space where data objects have been ordered in a “landscape” with respect to their overall similarity. SOMs have a wide variety of application in various fields from predictive analytics for marketing, to classification of wines, detection of credit card fraud, optical character recognition and medical imaging.
The MOCAB Indicator.
The MOCAB Indicator that is compiled at the end of each trading week, contains information to gauge the market outlook for the coming week based on the situation as at Close on Friday. (if Friday is not a trading holiday) and to select stocks with high Alpha and the appropriate Beta. The input of the MOCAB Indicator are the time-tested stock analysis Models of ValuEngine Inc of Princeton, NJ and the output and analysis is done using the SOM-based Data Mining software of Viscovery Software GmbH in Vienna, Austria.
MOCAB Indicator Information Content
Market Mode: The screening output of the three ValuEngine models (Valuation, Growth and Quality) and how they are positioned on the SOM will be an indicator of the strength of each mode.
Long/Short: Each of the three ValuEngine screens has Long/Short versions, and how the L and S stocks are clustered and positioned on the SOM will be an indicator of the market strength.
Sector Information: Each of the screened output is also labelled to indicate the sector they belong to, and how they are clustered and positioned on the SOM may yield information on sector strength .
High Alpha stocks: Alpha is the return in excess of the return of some market Index. In other words, the non-Beta part of a stock's movement or to put it even more clearly the return on a stock if the market return were zero. In the context of this Blog and its SOM technology, Alpha is defined as the degree of dis-similarity of a stock with a market Index and is represented by the cluster that has the most difference with the cluster approximating the Index in terms of exhibiting desirable values of the ValuEngine model variables. And within this cluster, the individual stocks that most represents the properties of the cluster are the high Alpha stocks. On a SOM, overall degree of similarity/dis-similarity is measured within a cluster by the Euclidean [mathematical space] distance between nodes with the greater the distance indicating the greater the dis-similarity. Among clusters on a SOM, the degree of dis-similarity of a cluster with a market Index is measured by the deviation of the Mean of the cluster from the Mean of the entire data set or a cluster approximating the market Index. * for justification of my definition see section on heuristic approach and holistic perspective below.
Beta: Beta is a measure of a stock’s performance relative to the general market. It is calculated by doing a regression analysis of a stock and the market’s price movement over a period of time. The market as represented by an Index is designated a Beta of 1, and so a stock with a Beta of 1.5 will theoretically move 50 % more than the market when it is going up and also when it is going down. Our objective is to look for high Alpha high Beta stocks when the market is trending up and high Alpha low Beta stocks when the market is trending down. On a SOM, high Beta or low Beta stocks can be identified using the Beta attribute map which plots the model variable Beta as a ‘heat map’ with color intensity scale tending towards Red for high Beta and towards Blue for low Beta.
Heuristic approach based on domain expertise.
1. It is my belief that beyond a point, it is necessary for quantitative analysis of financial markets to incorporate heuristics based on domain expertise and experience. It is also my belief that a heuristic approach brings with it a more holistic perspective which is essential for the soft sciences like Economics and Finance that have to contend with human behaviour. Shu-Heng Chan and Paul P. Wang as editors of Computational Intelligence in Economics and Finance [Springer-Verag 2004] have also mentioned the need for a heuristic approach based on domain expertise because of the special issues that economics and finance modelling involve viz extremely noisy data, behavioural changes and non-linear relationships. To which I might add the issues of long fat-tailed probability distributions, Black Swan events, missing data, heteroskedasticity, autocorrelation and frequent regime switches. To quote from their book: “In the domain of highly complex problems, precision is neither possible nor often desirable. Heuristics or approximate algorithms become the only acceptable tools” A sentiment also shared by Prof. Lotfi Zadeh, the father of Fuzzy Logic. A heuristic approach is also an inherently more robust approach with more room for accommodating higher degrees of uncertainty a point not to be dismissed considering the characteristics of modern financial markets.
2.The market can be described by three modes which can be characterized as: Valuation, Growth and Quality. Valuation mode is characterized by investors’ emphasis on fundamentals with the accompanying technical characteristics of Oversold/Overbought and reversion to the Mean . Growth mode emphasises the future, places less weighting on present fundamentals and is accompanied by the technical characteristics of Momentum and Trend. Quality mode is concerned with volatility and stocks are selected based on their risk/reward ratio as represented by a metric such as the Sharpe Ratio. At any point in time, the market is a combination of various degrees of Valuation, Growth and Quality.
3. The definition of Alpha using SOM is a ‘purer’ and more holistic definition of Alpha. The traditional Alpha which can be depicted mathematically as the point where the Beta line intersects the Y axis is an ex-post statistic calculated from the Beta and dependent on (arbitrary) choice of time period for its calculation and has little predictive value. It is a market dynamic statistic like a technical analysis indicator. The definition of Alpha that is used here measures dis-similarity based on all the variables of the SOM model which are in turn derived from the fundamentals-based ValuEngine models and therefore have predictive value since fundamentals such as earnings, sales, book value, cash flow, market cap, earnings surprise, yield on long term treasuries, etc have been proven to have predictive value for medium and longer term investment time frame.
4. Alpha values alone is not sufficient for stock selection. Alpha and Beta must be used together. On a SOM we can use the ‘heat map’ of the model variable Beta to pinpoint the high Beta or low Beta stocks among the stocks which in the previous step had been selected for high Alpha. Our methodology selects the best of the best in the sense that the selected stocks could be from any of the three ValuEngine screens based on the three ValuEngine models Valuation, Growth and Quality. Our method also does not limit us to a fixed number of selected stocks. After the high Alpha stocks have been selected we discard those with undesirable Beta values. In a up-trending market, stocks with high Alpha and high Beta are selected to take advantage of the upward move. But high Beta is a double-edged sword, and in a down-trending market stocks with high Beta will also move down more than the market. Therefore in a down-trending market, we should select stocks with high Alpha and low Beta. In a sideways trendless market, it is safer to stick to high Alpha and low Beta stocks.
Methodology
1.A SOM of the component stocks of the S&P500 is created. This SOM represents the basic topology of the market.
2. Six sets of 20 (22 stocks for Growth model) stock portfolios based on Long and Short versions of the three ValuEngine screens [Valuation, Growth, Quality] are created. (1). Valuation Long. (2).Valuation Short (3). Growth Long (4).Growth Short (5). Quality Long (6). Quality Short. * Note ValuEngine uses a different name for their screens. For our purpose, ValuEngine Standard= Valuation; ValuEngine Forecast= Growth, and ValuEngine Star= Quality. Note #2: Growth model has 22 stocks instead of 20 because it’s set-up of 2 stocks per sector out of the 11 sectors S&P500 is fixed.
3. The screened stocks are marked such that the screen from which they originated, the sector they belong to, and whether they are Long or Short positions are all included in the selection. In addition, some of the stocks are labeled in the map. The position of a label is approximately the position of the node on the SOM that the stock occupies. The S&P500 component stocks are not labelled and the empty spaces represent the nodes on the SOM that they occupy.
4.The acronyms used in the labelling are:V=Valuation; G=Growth; Q=Quality; L=Long; S=ShortB= Basic Industries C= Capital Goods D= Consumer Durables E= EnergyF= Finance H= Healthcare ND= Consumer Non-Durables S= Consumer Services T= Technology TP= Transportation U= Public Utilities* S&P500 stocks are not labeled and thus occupy the 'empty' spaces on the SOM. Thus GLT is a growth model Long stock from the Technology sector and VSS is a valuation model Short stock from the consumer services sector.

Market Outlook-cum-AlphaBeta Indicator (Prototype)

2010-06-26T17:34:00.015+10:00

1. Beta Attribute Map
2. Self-Organized Clusters

3. Cluster Statistics
4. Cluster Statistics Summary
5. Long Stocks
6. Short Stocks

This is a lengthy post explaining my methodology for a prototype Market Outlook-cum-Alpha/Beta Indicator (MOCAB). For the benefit of those not interested in explanations I'll begin with the market analysis and stock selection for this week based on MOCAB and then follow up with the explanation on methodology and labelling. But first I need to mention that this is not some kind of technical analysis indicator based on just Price and Volume data. This is an indicator based on the clustering of fundamental data and on the screens of ValuEngine models. Without the data and software provided by the good people at ValuEngine, my work would not have been possible. As mentioned, MOCAB Indicator is currently a prototype and as the weeks go by, it will be fine-tuned.
Market outlook for coming week
This is going to be another nail-biting week, with a bias towards 'down'. Cluster S1 holds most of the S&P500 component stocks, but many of the Short stocks (with 'S' as the second letter of their label) in the ValuEngine screens are also in S1, and since S1 is approximately the Index, the implication is that the Index will be bearish. Another bearish sign: Most of the other Short stocks are in cluster S3, while the Long stocks (second letter of their label is 'L') are in S2. S3 is a more cohesive cluster (smaller area) than S2 and the stocks in it are closer together i.e. more homogenous. Which means the bearish signal (as represented by the S stocks) is stronger than the bullish signal (as represented by the L stocks). Yet another sign: Image 3 the cluster statistics shows that in general S3 bars are much longer than S2 bars. Since the length of the bars represents deviations of the cluster from the entire data set, and S1 is the closest approximation of the entire data set, the much longer bars of S3 as compared to S2 means that the bearish (Short) stocks in S3 have a greater deviation from the market than the bullish (Long) stocks in S2. So there you are, thats the general outlook for the market
High Alpha, Low Beta stocks
It has been confirmed by many studies that about 70 % of the gain (or loss) from a stock is due to the market. The tide lifts all boats up and down. With such a market outlook for the coming week, we should be looking at high Alpha AND low Beta stocks. High Alpha stocks with high Beta are not going to fare very well. With that in mind, I used the Beta attribute window (image 1) to look at those areas on the map with low Beta. (see sliding scale at bottom of map). Then in those areas I selected the high Alpha stocks (see below for how high Alpha stocks are selected).
The pickings as shown in image 5 are slim. And these stocks while having the lowest Beta among the ValuEngine screened Long stocks still have a relatively high Beta value of 1.73, 1.77 and 2.13. The small number of suitable stocks and the high Beta of these 'suitable' stocks, is an indication that theres not much scope for playing the market this coming week with even the high Alpha stocks swinging up and down with it. Also, there is not a significant degree of dis-similarity between the S1 control cluster and S2 the Long stocks cluster as seen by the difference in the length of bars in the Statistics bar graph. Which is an indication that the Alpha strength is not high and movements in the general market are more likely to have an impact even on the selected stocks. Caveat: The market outlook is based on all known information as at end of Friday close, and not taking into account any external shocks such as economic and political developments and natural disasters.
BTW, Image 6 shows the Short picks. There seems to be a strong sell sign for Cemex and Dania Holding Corp.
Now comes the explanation of what MOCAB Indicator is all about.
Introduction

Stock markets are Complex Adaptive Systems [CAS] with all the unique properties of such systems. Properties of CAS such as feedback loops, emergence, self-organization, self-similarity, co-evolution and distributed connectivity require that a meaningful analysis of CAS be done with tools that take into account such complexities. To deal with the complexities of CAS requires a paradigm shift from hard computing to soft computing, from linear parametric modelling to non-linear non-parametric modelling and from exact solution to approximate solution.While most of the world’s phenomena (natural or man-made) are CAS, it is only in the last decade or so that with the exponential leap in the power of computers and software, we are able to use tools more suited to the analysis of CAS. Such tools include neural networks, fuzzy logic, evolutionary algorithms, wavelets,swarm intelligence and, in particular, Self-Organizing Maps Click ValuEngine Inc of Princeton, NJ and the output and analysis is done using the SOM-based Data Mining software of Viscovery Software GmbH in Vienna, Austria.

MOCAB Indicator Information Content

Market Mode: The screening output of the three ValuEngine models (Valuation, Growth and Quality) and how they are positioned on the SOM will be an indicator of the strength of each mode.

Long/Short: Each of the three ValuEngine screens has Long/Short versions, and how the L and S stocks are clustered and positioned on the SOM will be an indicator of the market strength.

Sector Information: Each of the screened output is also labelled to indicate the sector they belong to, and how they are clustered and positioned on the SOM may yield information on sector strength .

High Alpha stocks: Alpha is the return in excess of the return of some market Index. In other words, the non-Beta part of a stock's movement or to put it even more clearly the return on a stock if the market return were zero. In the context of this Blog and its SOM technology, Alpha is defined as the degree of dis-similarity of a stock with a market Index and is represented by the cluster that has the most difference with the cluster approximating the Index in terms of exhibiting desirable values of the ValuEngine model variables. And within this cluster, the individual stocks that most represents the properties of the cluster are the high Alpha stocks. On a SOM, overall degree of similarity/dis-similarity is measured within a cluster by the Euclidean [mathematical space] distance between nodes with the greater the distance indicating the greater the dis-similarity. Among clusters on a SOM, the degree of dis-similarity of a cluster with a market Index is measured by the deviation of the Mean of the cluster from the Mean of the entire data set or a cluster approximating the market Index. * for justification of my definition see section on heuristic approach and holistic perspective below
Beta: Beta is a measure of a stock’s performance relative to the general market. It is calculated by doing a regression analysis of a stock and the market’s price movement over a period of time. The market as represented by an Index is designated a Beta of 1, and so a stock with a Beta of 1.5 will theoretically move 50 % more than the market when it is going up and also when it is going down. Our objective is to look for high Alpha high Beta stocks when the market is trending up and high Alpha low Beta stocks when the market is trending down. On a SOM, high Beta or low Beta stocks can be identified using the Beta attribute map which plots the model variable Beta as a ‘heat map’ with color intensity scale tending towards Red for high Beta and towards Blue for low Beta.

Heuristic approach based on domain expertise.

1. It is my belief that beyond a point, it is necessary for quantitative analysis of financial markets to incorporate heuristics based on domain expertise and experience. It is also my belief that a heuristic approach brings with it a more holistic perspective which is essential for the soft sciences like Economics and Finance that have to contend with human behaviour. Shu-Heng Chan and Paul P. Wang as editors of Computational Intelligence in Economics and Finance [Springer-Verag 2004] have also mentioned the need for a heuristic approach based on domain expertise because of the special issues that economics and finance modelling involve viz extremely noisy data, behavioural changes and non-linear relationships. To which I might add the issues of long fat-tailed probability distributions, Black Swan events, missing data, heteroskedasticity, autocorrelation and frequent regime switches. To quote from their book: “In the domain of highly complex problems, precision is neither possible nor often desirable. Heuristics or approximate algorithms become the only acceptable tools” A sentiment also shared by Prof. Lotfi Zadeh, the father of Fuzzy Logic. A heuristic approach is also an inherently more robust approach with more room for accommodating higher degrees of uncertainty a point not to be dismissed considering the characteristics of modern financial markets.

2.The market can be described by three modes which can be characterized as: Valuation, Growth and Quality. Valuation mode is characterized by investors’ emphasis on fundamentals with the accompanying technical characteristics of Oversold/Overbought and reversion to the Mean . Growth mode emphasises the future, places less weighting on present fundamentals and is accompanied by the technical characteristics of Momentum and Trend. Quality mode is concerned with volatility and stocks are selected based on their risk/reward ratio as represented by a metric such as the Sharpe Ratio. At any point in time, the market is a combination of various degrees of Valuation, Growth and Quality.

3. The definition of Alpha using SOM is a ‘purer’ and more holistic definition of Alpha. The traditional Alpha which can be depicted mathematically as the point where the Beta line intersects the Y axis is an ex-post statistic calculated from the Beta and dependent on (arbitrary) choice of time period for its calculation and has little predictive value. It is a market dynamic statistic like a technical analysis indicator. The definition of Alpha that is used here measures dis-similarity based on all the variables of the SOM model which are in turn derived from the fundamentals-based ValuEngine models and therefore have predictive value since fundamentals such as earnings, sales, book value, cash flow, market cap, earnings surprise, yield on long term treasuries, etc have been proven to have predictive value for medium and longer term investment time frame.

4. Alpha values alone is not sufficient for stock selection. Alpha and Beta must be used together. On a SOM we can use the ‘heat map’ of the model variable Beta to pinpoint the high Beta or low Beta stocks among the stocks which in the previous step had been selected for high Alpha. Our methodology selects the best of the best in the sense that the selected stocks could be from any of the three ValuEngine screens based on the three ValuEngine models Valuation, Growth and Quality. Our method also does not limit us to a fixed number of selected stocks. After the high Alpha stocks have been selected we discard those with undesirable Beta values. In a up-trending market, stocks with high Alpha and high Beta are selected to take advantage of the upward move. But high Beta is a double-edged sword, and in a down-trending market stocks with high Beta will also move down more than the market. Therefore in a down-trending market, we should select stocks with high Alpha and low Beta. In a sideways trendless market, it is safer to stick to high Alpha and low Beta stocks.
Methodology
1.A SOM of the component stocks of the S&P500 is created. This SOM represents the basic topology of the market.
2. Six sets of 20 (22 stocks for Growth model) stock portfolios based on Long and Short versions of the three ValuEngine screens [Valuation, Growth, Quality] are created. (1). Valuation Long. (2).Valuation Short (3). Growth Long (4).Growth Short (5). Quality Long (6). Quality Short. * Note ValuEngine uses a different name for their screens. For our purpose, ValuEngine Standard= Valuation; ValuEngine Forecast= Growth, and ValuEngine Star= Quality. Note #2: Growth model has 22 stocks instead of 20 because it’s set-up of 2 stocks per sector out of the 11 sectors S&P500 is fixed.
3. The screened stocks are marked such that the screen from which they originated, the sector they belong to, and whether they are Long or Short positions are all included in the selection. In addition, some of the stocks are labeled in the map. The position of a label is approximately the position of the node on the SOM that the stock occupies. The S&P500 component stocks are not labelled and the empty spaces represent the nodes on the SOM that they occupy.
4.The acronyms used in the labelling are:V=Valuation; G=Growth; Q=Quality; L=Long; S=ShortB= Basic Industries C= Capital Goods D= Consumer Durables E= EnergyF= Finance H= Healthcare ND= Consumer Non-Durables S= Consumer Services T= Technology TP= Transportation U= Public Utilities* S&P500 stocks are not labeled and thus occupy the 'empty' spaces on the SOM. Thus GLT is a growth model Long stock from the Technology sector and VSS is a valuation model Short stock from the consumer services sector.

Back To The Drawing Board

2010-06-19T16:44:00.003+10:00

1. The portfolio after two weeks
2. A Market Direction Gauge-cum- Stock Picker

3. Cluster Statistics

4. The High-Alpha Stocks

After two weeks, our monthly re-balance market-neutral portfolio is down 1.42 % while the S&P500 is up 5.18 %. That's because of our Short position via the ProShares S&P500 ETF/ (Ticker symbol SH). If not for SH, our tech-dominated portfolio would be up 2.32 %. This is still underperforming the Index but we still have two weeks to see whether the tech stocks will power on. Still, the lesson to be learned from this is that we should not ignore the basic fact that the natural tendency of stocks is to have an upward bias. Thus giving a 50% weight to the Short position is a mistake. An algorithm must be devised for the weighting of the Short component of a MNS strategy.

This week we attempt to (1)create a market direction gauge with SOM and (2) refine our technique for using SOM to pick high Alpha stocks.

Approach is based on my following beliefs that:
1.The market can be completely described by three modes: Valuation, Growth and Quality. Valuation mode is characterized by investors’ emphasis on fundamentals with the accompanying technical characteristics of Oversold/Overbought and reversion to the Mean . Growth mode emphasises the future, places less weighting on present fundamentals and is accompanied by the technical characteristics of Momentum and Trend. Quality mode is concerned with volatility and stocks are selected based on their risk/reward ratio as represented by a metric such as the Sharpe Ratio. At any point in time, the market is a combination of various degrees of Valuation, Growth and Quality.
2.Alpha which is the risk-free return in excess of some market Index is defined as the degree of dis-similarity of a stock with the market index. In other words, the non-Beta part of a stock's movement. The higher the degree of dis-similarity, the higher the Alpha. In the context of this Blog and its SOM technology, Alpha is represented by the cluster that has the most difference with the cluster approximating the Index in terms of exhibiting desirable values of the ValuEngine model variables. And within this cluster, the individual stocks that most represents the properties of the cluster are the high Alpha stocks. On a SOM, overall degree of similarity/dis-similarity is measured in the cluster by the Euclidean [mathematical space] distance between nodes, and between clusters by the deviation of the cluster Mean from the Mean of the entire data set. Picking high Alpha stocks means picking stocks which outperform the market index when the market is rising and under-perform the Index when the market is falling . . i.e. falling at a slower rate than the Index.

3. Our analysis is based on the assumption that the output of each ValuEngine screen accurately reflect their respective model selection criteria of valuation,growth and quality.

4. This approach is also based on my philosphy of incorporating heuristics based on experience into quantitative analysis..
Methodology
1.A SOM of the component stocks of the S&P500 is constructed. This SOM represents the basic topology of the market.
2. Six sets of 20-stock portfolios based on Long and Short versions of the three ValuEngine screens [Valuation, Growth, Quality] are created. (1). Valuation Long. (2).Valuation Short (3). Growth Long (4).Growth Short (5). Quality Long (6). Quality Short. * Note ValuEngine uses a different name for their screens. For our purpose, ValuEngine Standard= Valuation; ValuEngine Forecast= Growth, and ValuEngine Star= Quality.
3.Prior to overlaying on the SOM, each of the screened stocks are labeled such that the screen from which they originated, the sector they belong to, and whether they are Long or Short positions are all denoted by the label. The position of a label is approximately the position of the node on the SOM that the stock occupies. The S&P500 component stocks are not labeled and the empty spaces represent the nodes on the SOM that they occupy.
4.The acronyms used in the labelling are:V=Valuation; G=Growth; Q=Quality; L=Long; S=Short
B= Basic Industries C= Capital Goods D= Consumer Durables E= Energy
F= Finance H= Healthcare ND= Consumer Non-Durables S= Consumer Services T= Technology TP= Transportation U= Public Utilities
* S&P500 stocks are not labeled and thus occupy the 'empty' spaces on the SOM. Thus GLT is a growth model Long stock from the Technology sector and VSS is a valuation model Short stock from the consumer services sector.
Market Outlook

Image 2 shows that the SOM is composed of three natural clusters: S1 the biggest cluster contains 66.33 % of our stock population. With a lot of space unoccupied by the ValuEngine screened stocks, S1 is where most of the S&P500 stocks are located. S2 contains 21.90 % of the stock population and is mostly occupied by the Short (S) stocks. S2 contains 17.77 % of the stock population and is mostly occupied by the Long stocks (L). To have a sense of the market direction for next week, we first look at the integrity of S2 and S3 clusters and their degree of dis-similarity with the market as represented by the S&P500 (the data set from which our SOM is constructed). Image 3 is an indicator of the degree of dis-similarity of S2 and S3 with the market. The degree of dis-similarity is measured by the length of the bars on the bar chart which measures the deviation of the Mean of the cluster from the Mean of the entire data set (the market). S2 and S2 have relatively long bars as compared with S1 which is the nearest approximator of the market since it contains most of the S&P500 component stocks. However, S2 which contains the S stocks is not a 'tight' homogenous cluster. There are big empty spaces in between. On a SOM, this is an indication that the clustering is not 'strong' and thus S2 cannot be a good indcator of market bearishness. S3 which contains the L stocks is also not a distinctive, cohesive cluster. Other than an area at the top, S3 also has many empty spaces.

The overall outlook of the market for next week is thus ambivalent. It could be up or it could be down. But judging from the slightly better clustering characteristics of S3 as compared to S2, market outlook is slightly positive.

High Alpha stocks

To pick high Alpha stocks as defined above, we need to identify the area of S3 which is most densely populated with the ValuEngine screened stocks. This area would contain stocks which are very similar to each other in having the characteristics of S3 cluster. The darkened area is the selected area and the stocks occupying the nodes of the darkened area is in image 4. Note that some stocks e.g. EK appear in more than one ValuEngine screen. Our methodology selects the best of the best in the sense that the selected stocks could be from any of the ValuEngine screens. Our method also does not limit us to a fixed number of selected stocks. The sectors the high Alpha stocks belong to are: Finance, Consumer_Non_Durables, Transportation, Basic Industries, Technology and Consumer Services. The absent sectors are Consumer_Durables, Energy, Capital Goods and Public Utilities and Healthcare.

Quantifying The Elusive Alpha

2010-06-12T16:43:00.013+10:00

2. Mean Statistics of the clusters
3. Clusters: ValuEngine Screens Against The Backdrop Of The S&P500

4. Cluster Statistics (Standard Deviations From Mean of Data Set)

5. Performance of Last Week's Market Neutral Strategy With Short S&P500 ETF [SH]

6. Performance of Last Week's Market Neutral Strategy With UltraShort S&P500 ETF [SDS]

First, to get it out of the way. The performance of last week's Market Neutral Strategy is shown in images 5 and 6 above. The strong performance of Technology sector stocks helped lift the portfolios up. If not for the Short on the S&P500, we would be in positive territory. But never, never regret it. The Short is a hedge and it should stay. Remember, this is a portfolio for Joe Smith and he cannot withstand stomach-churning volatility and steep drawdowns on his investment capital. Anyway, it's early days yet, as this is a monthly re-balance type of portfolio.

Now, for the serious business. Fund managers are always seeking Alpha or professing to have Alpha, but has anybody been able to measure Alpha? Alpha can be defined as the return in excess of some market index. Mathematically, if Beta measures the volatility of a stock against the market, and is measured by the slope of a line in a chart showing returns of a stock verus the market's return, then Alpha is where this line intercepts the Y axis of the chart. But although Beta can be meaningfully expressed in a quantitative way (e.g a stock with Beta of 1.5 will go up [or down] 1.5 times as much as the Index), Alpha cannot be similarly expressed. And thus it remains a nebulous concept for fund managers to foist on you.

In this post, I use a Self-Organizing Map to define Alpha as the ' degree of dis-similarity of a stock with the Index'. The implication of this is that the stock will outperform the Index when the Index is rising. Note that a stock can be dis-similar to the Index in a bad way too- which means that it will 'outperform' the Index negatively when the market is falling. However, Alpha is only concerned with 'good' dis-similarity.

This is a valid and practical definition and a SOM is a good way to measure dis-similarity in a holistic way. SOMs measure dis-similarity taking into account all the attributes [variables] of the stock against all the attributes of all other stocks at a certain point of time t. Degree of dis-similarity/similarity is demarcated by the boundaries of each of the self-organized clusters and a metric like Euclidean distance measures degree of similarity between nodes in the SOM's topology, thus making our Alpha measurable.

So much for the theory. Now for the real-life example. As in last week's post, ValuEngine's three screens: 20 stocks Standard (Valuation), 22 stocks Forecast (Growth), 20 stocks Star (Quality) are plotted against the backdrop of the S&P500 component stocks. The terminology: V=Valuation, G=Growth, Q=Quality. Sectors: B=Basic Industries, C=Capital Goods, D=Consumer Durables, ND=ConsumerNonDurables, S=Consumer Services, E=Energy, F=Finance, H=HealthCare, T=Technology, TP=Transportation, U=Public Utilities. Thus VU is a Valuation stock in Public Utilities sector. GE is a Growth stock in Energy etc.

Our objective is to meet two conditions (1) first to pinpoint the cluster(s) which is most dis-similar to the market. (2) and is the most densely populated. (3) Finally, we pick the stocks from the three ValuEngine screens which inhabit the densely populated zone of the cluster which is the most dis-similar to the Index. In image 3, the space on the SOM that is not labeled represents the SP500 stocks which are not output by the ValuEngine screens. This includes most of the area of cluster S1, S2 and parts of S3. In image 4, the length of the bars represents the degree of difference [standard deviation] from the Mean of the entire data set. Thus S5 cluster is greatly different from S1 and S2 clusters and in turn, S1 and S2 by their short bars are not so different from the Mean of the entire data set, which in our case essentially means the market as defined by the S&P500 . S5 is different from S1 in a good way too: The different colored arrows show that for example, S5 is more undervalued, yet has greater momentum and has a much higher 1-month forecast return %. Other traits of S5 include smaller market cap, higher volatility and higher Beta. These last three traits may or may not be good depending on the mode and mood of the market. If confidence returns [for the coming week at least], and risk appetite increases, being a smaller market cap stock with higher volatilty could be a plus in terms of outperforming the market. Neverthess, we are assured by the fact that our crop of stocks is a good mix of Valuation, Growth and Quality (see the V, G, and Q prefix in the image 1 stock list]. A good mix of sectors is also present: see the suffix of B, E, F, ND etc]. And don't forget that we can have a Short position in our portfolio via the SP500 reverse ETF SH or SDS. All in all, a good portfolio for poor Joe Smith. We won't track this portfolio as last week's portfolio one is only one week old and not due to be re-balanced for another three weeks. But at least, we have defined Alpha in a quantitative way.

Designing An Investment Strategy For Joe Smith The Babyboomer

2010-06-06T17:36:00.005+10:00

1. Clusters
2. Cluster Means

3. Cluster Statistics

4. The Market Neutral Portfolio

The financial objectives of Joe Smith Babyboomer will be quite different from a younger person. Joe is still shell-shocked from the damage to his 401K that was caused by the financial crisis. Joe doesn't trust Wall Street anymore. All he wants is a stable return that is significantly more than the pittance that banks currently pay for his deposit. But Joe will not do risky investments that may endanger whatever sum of money he still has. Still, Joe thinks it would be nice if his portfolio is able to generate a little extra pocket money for that fly-fishing trip, a trip to Shanghai with the Missus, or that Gibson archtop guitar in the shop window that's staring at him and making him drool. Let's try to design an investment strategy for Joe that is simple and doesn't involve having Joe's blood pressure rocketing as volatility causes huge drawdown on his investment capital. This will be an experiment- an ongoing exercise- with records updated every week, so that Joe can see how well his investment is doing. First, the perimeters and parameters of this strategy:

1. It will be for an investment capital of $30000 of spare cash that Joe has.

2. Since it is impossible to predict the market, and Joe can't buy-and-hold for as long as Warren Buffet, it will be a market neutral strategy [MNS], with monthly re-balancing of portfolio. Another reason why Joe has lost faith in Buy and Hold is that he knows everything in the world changes faster than it used to, and today's Blue chip could be tomorrow's has-been.

3. For ease and practicality the Short side of the portfolio will be the ProShares Short S&P500 ETF. [ticker symbol SH] or the ProShares UltraShort S&P500 ETF [ticker symbol SDS]. *SDS goes up/down twice as much for every point of change in the S&P500.

4. The MNS portfolio will have equal $ weighting. 50 % Long and 50% Short and equally spread among the Longs.

5. Transaction cost will be $15 per trade. But for ease of calculation, we'll leave it out of our figures and you can include it in your own calculations.

6. The maximum number of Long stocks will be 10, but the actual number may be less if there are no stocks which meet the selection criteria.

7. The methodology for selection is to plot all 3 ValuEngine screens [Standard (for our purpose denoted as valuation mode)], Forecast [growth mode] and Star [Quality mode] on a Self-Organisng Map [SOM] and select those stocks which have good clustering properties whether they be from the Value, Growth or Quality screens. *clustering properties and SOM technology is discussed elsewhere throughout this Blog. So here we go:

1. In image 1, the stocks from ValuEngine's Value [V], Growth [G] and Quality [Q] screens are plotted over a SOM that has been constructed from the 500 stocks of the S&P500.

2. In image 2 and 3 we select cluster 2 [Red] to be the cluster with the most attractive clustering characteristics. The V, G and Q labels denoting nodes that are occupied by the screened stocks are packed densely together unlike S 1 and S3. Cluster statistics also show Undervaluation but with Momentum [12-m return %] The current market situation is risk-averse and places emphasis on valuation and quality.

3. Image 4 shows the final list of stocks selected, the number of shares of each, and their last Closing price. Our total investment is Long 14864 and Short 14988 if SH is used and 14979 if SDS is used. We will monitor the results to see if SH or SDS is a better choice for the Short side.
Lastly, please remember that this is an experiment, the results of which will be used to tune the SOM. Ultimately it is hoped that a good newsletter can be written that meets the financial objectives we discussed.

A Market Neutral Strategy Using Self-Organizing Map

2010-05-22T16:44:00.010+10:00

1. Self-Organizing Map of ValuEngine Engine-Ratings [Composite Ranking] of SP500 Component Stocks.

2. Average Composite Ranking Score of Clusters

3. Self-Organized Clusters of SP500 Component Stocks Labeled WIth Their ValuEngine Engine Ratings Composite Ranking Score.

4. SOM Top 10 Overall Best Stocks

5. ValuEngine Top 10 Overall Best Stocks.

6. SOM Bottom 10 Overall Worst Stocks

7. ValuEngine Bottom 10 Overall Worst Stocks

The aim of a market neutral strategy is to go Long on the best stocks, and Short on the worst stocks. It is a good strategy when markets are very unpredictable, as the gains come not from market direction, but the 'spread' between best and worst. In theory, the best stocks will outperform the market positively when the market rises, and outperform the market negatively when the market falls. But for a quantitative strategy, 'Best' and 'Worst" has to be quantitatively defined. ValuEngine's approach is to give stocks their unique Engine Ratings;1-Engine to 5-Engine with 5 being the most desirable. "Engine Ratings: ValuEngine's proprietary Engine Rating system provides an overall assessment of a stock's attractiveness. It combines valuation, risk-return tradeoff, momentum, market capitalization and forecasted future return data into an easy to understand and actionable format." And as can be seen from the top image, Engine Ratings have proven to be quite a good indicator of stock performance. The Engine Ratings in turn are based on a Composite Ranking score system. Using ValuEngine's Engine Rating system, we could have a Long portfolio comprising, say, the x number best stocks and a Short portfolio of the y number worst stocks, where x=y.

In this post, we use a SOM to find the 10 best stocks and the 10 worst stocks, and we compare the list of stocks that the SOM identifies, with ValuEngine's list. A SOM offers a more holistic way of looking at 'best' and 'worst'. It simultaneously takes into account all the relevant model variables and their non-linear inter-relationship with each other. Using the self-clustering properties of the SOM, and Euclidean distance on the SOM to measure degrees of similarity, it was hypothetised that the SOM's list of 'best' and worst' would be different from ValuEngine's list.

Some of the assumptions we used in this exercise are:

1.Our Universe of stocks comprises only the component stocks of the SP500 Index because (a) we want stocks with good liquidity and reasonable size [market cap].(b) Having only Index component stocks will also lower the volatility of the portfolio, making our strategy more suitable for ordinary folks who don't indulge in high frequency algorithmic trading.

2. An arbitrary 10 stocks per portfolio is used, because a 20-stock portfolio is manageable and affordable for the average investor with an investment capital of $20000 and yet offers suffiicent diversification.

3. The portfolios are equal weighted in terms of $, as I feel that trying to give weightings for each stock is a futile exercise in optimization and is a case of false precision. Optimization can be done, but I prefer the use of Genetic Algorithms rather than raditional statistical methods for optimization.

A SOM was constructed and the results are summarized below:

1. Image 1 shows the ValuEngine composite ranking scores of SP500 stocks as a heat map where areas of higher scores tend towards the color Red. [see scale below map].

2. Image 2 shows the SOM has divided the stocks into seven distinct clusters based on their degree of similarity in overall characteristics, with algotithmic thresholds to determine cluster boundaries.

3. Image 3 shows that cluster S4 [Green] has the highest average composite score of 713 and cluster S6 [Pink] has the lowest average composite score of 525. [Cluster S7 actually has the highest score of 884 but is a one-stock cluster [Eastman Kodak] and has very volatile characteristics. Image 4 also shows that the composite scores in each cluster do not run'consecutively' i.e. you may find a score of 656 in one cluster, while the next score of 657 may be in another cluster. This is the main difference between ValuEngine and the SOM.

Image 4 and 5 compare the SOMs 10 best stocks with ValuEngine's 10 best stocks. Note the differences and how the SOM puts more emphasis on the big cap technology sector e.g. Intel, Apple, Cisco and on raw materials e.g. Noble, Corning, Freeport-Mcmoran.

Image 6 and 7 compare the SOM's 10 worst stocks with ValuEngines 10 worst stocks. Goodyear, Office Depot, SprintNextel and Tesoro appear in both list but other stocks are not in both list. The 'worst' list seems to be more stock-specific and less sector-specific.

Last thoughts: Each week, or even each day, a stock's Engine rating may change. Those near the boundary between discrete Engine ratings may bounce back and forth between two ratings. The SOM has tools for displaying the finer details of the properties of a cluster and the stocks within it e.g it's centroid, its neighborhood, slope of the cluster boundaries, contour, and U-matrix. using these tools will further enhance knowledge of a cluster's characteristics and enable us to better select the stocks for the Long/Short portfolios. But for that, I will need to consult with Dr. Kranner, the inventor of Viscovery and my mentor.

This MNS strategy will not yield as good a return as ValuEngine's current market neutral strategy using the forecast model [as in their MNS newsletter by Steve Hach]. But it may be a more suitable strategy for conservative investors with steady longer term returns in mind, and less frequency in portfolio re-balance.

Mid Caps with Good Valuations and Momentum

2010-05-15T17:07:00.007+10:00

1. Mid-Caps With Good Valuations and Momentum

2. The Nine Natural Clusters Of The S&P500

3 Characteristics of Cluster 6

At the beginning of a market run, it's the Blue Chips that make the first move as mutual fund managers from all over the world load up their portfolios. When the Bull is looking tired after a long run, investors turn to the second tier stocks. While chasing third tier stocks may be hazardous, there are mid-caps that offer better returns than big caps without being especially risky. This is especially so when the market is stuck in a tight range, and highly liquid big caps and Index components make no headway. Algorithmic trading that target highly liquid big caps are content with shaving off decimals of a percent for their profit,and the constant buying and selling in the war of the machines does much to reduce volatility [except when exceptional shocks trigger even machines to follow the herd instinct to cause a free fall [or rise].

Image 2 shows a SOM of the SP500 stocks, labeled with the sectors to which they belong. [See previous posts for key to sector acronyms]. The SOM has divided the SP500 stocks into 9 natural clusters based on their degree of similarity. #Note that two stocks from the same sector do not necessarily belong to the same cluster. Based on the statistical properties of each cluster [image 3], and taking into account the current market mode and mood, we look for a cluster with slight undervaluation as well as moderate longer term momentum, as represented by 12-month return. Cluster 6 has such characteristics together with a healthy dose of volatility [necessary in times of flat markets]. If you look at cluster statistics in image 3, the clusters C1 to C9 from left to right are arranged in order of deviations from the Mean of the entire data set, getting more and more deviant with cluster 9 being the most extreme. We will not go for the 'dead' clusters like 1,2 and 3, nor will we take the too lively clusters 7,8, 9. So cluster 6 is our choice for some safe investing suited to the current market mode [and mood]. The top image 1 shows that cluster 6 is an assortment of stocks with no one sector predominating though we do see a fair number of Finance[F] and Technology stocks [T] in the list. The market cap ranges from 0.3 $Bil to 9.3 $Bil. . Average Daily Volume is healthy and 12-month return is positive. 5-year return % is dismal, but thats to be expected, and is in fact a positive indication that these stocks do still have some ground to catch up. Most importantly, the stocks have good valuations [- %] and low P/Es.

CitiGroup: Example of Algorithmic High Frequency Trading

2010-05-09T00:56:00.006+10:00

1. CitiGroup: A one-stock cluster
2. Statistics of the Seven clusters in the SOM above

3. Citigroup compared with the rest

The past few days has demonstrated what high frequency algorithmic trading can do to the market. Individual investors who have never stepped into a current day trading room find it hard to fully realize the impact that machine-controlled has on them. See my previous post at http://www.technifundamentals.com/2009/08/high-frequency-algorithmic-trading-and.html

In today's post, I plotted the ValuEngine model variables of the SP500 component stocks on a Self-Organizing Map [SOM]. The SOM divided the stocks into seven distinct clusters [see image 1 above]. But cluster 7 right at the North-West corner, is a one-stock cluster i.e. it is a very distinct cluster, but the cluster contains only one stock -CitiGroup. Now, when we go to image 2 to have a look at each cluster's statistics [plotted as deviations from the Mean of the entire data set], we understand why C stands alone by itself. C is startlingly different from all the other stocks in the SP500. C is a prime example of a stock that is the object of algorithmic high frequency trading:

It's average daily volume is an unheard of 639,321,283. Other high liquidity big caps like WMT or PG average 15 to 20 million in Volume. What other characteristics does C have to make it machine-bait? Image 2 shows that C is very undervalued [its has a valuation of -70 % according to ValuEngine]. It also has a very high 1-month forecast return % and a very high 1-yrar forecast return %. And the expected EPS growth is 400%!.

Therefore it is very suitable for machine-controlled trades, for those machines implementing a mean reversion algorithm as well those implementing a momentum algorithm. Individual investors should stay clear of machine dominated stocks because they will be caught in the cross-fire as the machines fight it out among themselves and try to outguess each other's move. Volume may be vey high but changes in the price is limited as the machines go for big volume and are able to profit from penny-size gains.

So will other stocks that have similar characteristics as C behave like C? Unfortunately for this week, there is no other stock in the SP500 that has even a remote similarity with C for us to analyze. But next week, we will track C and see if there are stocks in the same neighborhood as C.

So Will It Be the Technology or the Finance Sector?

2010-05-03T14:59:00.011+10:00

1. SOM cluster with DJIA components as reference point.

1. Sectors ranked by valuation %

5. Sectors ranked by Momentum [12-m return %]

The U.S. stock market has come to the stage where it is needs to justify its struggle to grind upwards. One of the ways adopted by fund managers is to look past individual stocks to look at sectors. Of late, the sectors of interest have been Finance and Technology. In this post, we analyze the pros and cons of these two sectors. Image 2 shows that in terms if valuation [as defined by ValuEngine models], Finance is 7.56 % overvalued, while Technology is 1.84 % undervalued. (However, note that in terms of a conventional measure such as P/E, Finance has a P/E of 19.62 while Technology has a P/E of 28.68). That is because, ValuEngine's valuation encompasses factors like analysts estimates, EPS surprise, and market dynamics like liquidity, momentum, and market cap. So in terms of valuation, being oversold and reversion to the Mean, Technology is a better bet than Finance.Next, image 3 shows that in terms of Momentum [defined here as the 12-m return %], Technology has a higher Momentum of 66.46 % while Finance has a Momentum of only 36.93 %. With lower valuation AND higher momentum, Technology is a better choice than Finance, despite its higher P/E. The Universe for our study are the three models of ValuEngine: Valuation [V], Growth[G] and Quality[Q], and the output of each of their Long[L]/Short[S] screens, numbering 20 VL + 20 VS + 22 VG + 22 VS + 20 QL + 20 QS= 124 stocks in total. The proportion of Finance Long stocks [FL] to Finance Short stocks[FS] is FL/FS= 14/16=0.87. The proportion of Technology Longs to Technology Shorts is TL/TS= 11/5= 2.2 a very significant difference! So this is another factor in favor of Technology sector over Finance sector.

Yet another factor which you should consider is that the longer term Beta of Finance as calculated by ValuEngine is 1.44 or while for Technology it is 1.16. Which means that if the market take a tumble, Finance stocks should lose more than Technology stocks. On the the hand, if the situation is reversed, Finance stands to make bigger gains.

Finally, what is the market outlook for the coming week sans new external shocks from Goldman, Greece, North Korea etc? Image 1 shows the three natural clusters of our stock Universe plus the DJIA components. Note that there is not a single S stock from any of the three models V, G, or Q in S2, the cluster that has most of the DJIA components. [AA nd BAC are the odd ones out, being in S1 instead of in S2 with all the other DJIA components. ] Since a SOM clusters together stocks by their degree of similarity, and since all the stocks in the cluster where the DJIA components are L stocks we can imply, that for the short term, the outlook for the DJIA remains cautiously optimistic.

Time To Consider Having An ADR Portfolio

2010-04-24T16:51:00.009+10:00

1. Stocks in S3 Cluster
2. Clusters

3. Cluster statistics

4. Sharpe Ratio map

5. Sharpe Ratio stocks

ADRs have 'grown up'. Five years ago, ADRs were considered a good but more risky investment [except for the ADRs of established European companies]. But the post- financial crisis economic map of the world is now very much different- with the U.S. economy expected to grow at a much slower pace than the economies of Asia and some Latin American countries. At the same time, the bigger companies of these fast growing countries are now catching up with Western companies in terms of management, innovation and financial clout. China's BYD the battery maker has even attracted Warren Buffet. Baidu the Chinese Search Engine, India's Tata which took over Jaguar from Ford, Brazils Rio Tinto and Australia's BHP all have shown that they are fit to be franked among the world-class multinationals.

Using the fundamentals data of ValuEngine and the visual analytics of Viscovery, this post compares the quality of ADRs with the DJIA components.
Image 2. In order to make the comparison of ADRs with the DJIA stocks more meaningful, only ADRs with market cap >US$1 Billion, and minimum average daily volume of >250000 shares were considered. There were 169 ADRs in this group [or 31 % of the 542 ADRs covered by ValuEngine]. Classified by continent, 64 were European companies [37 %], 56 were Asian [33 %], 45 were Latin American [26 %] , 4 were from Australia and 3 were African. For historical and geographical reasons the European and Latin American companies were the first to list on US Exchanges. But the Asian companies are fast catching up.

Europe ADRs are labeled E on the Self-Organizing Map, Asian=F, Latin America=L, Australia=A and Africa=K. DJIA stocks are labeled X. The SOM generated three natural clusters [S1,S2, and S3]. The first impression is that there is no natural grouping of E, L or F ADRs and they are distributed among the three clusters. However all the DJIA stocks are in S2 except for Boeing in S3 and Alcoa in S1.

Image 3: Cluster Statistics: For each vriable the statistics are expressed as deviation of the Mean of the range of the cluster from the Mean of the entire data set for that variable; so as to standardize the different unit values of each variable, and the length all the bars are inter-comparable. This table is the one that can tell us a lot about the quality of the ADRs. Let's take the variables one by one starting with the first bar from the left of each set of cluster statistics, comparing S2 which has the DJIA stocks with S3 which is all ADR.

Valuation: This is the first bar from the left [Color Burnt Sienna?]. As shown by the arrows drawn from "valuation %" S2 stocks are undervalued [negative valuation %] while S3 stocks are overvalued. [or you can view it from the perspective that S3 stocks are priced for greater potential growth]

Average Volume: [Green] S3 stocks have much lower Average Volume than S2 stocks. A slight problem of liquidity here

Last 12-m Return %: [Grey]. Here's where ADRs really outshine the stocks of the DJIA

Sharpe Ratio [Ochre Brown]: And you'll be surprised that ADRs on S3 have a much higher Sharpe Ratio. Sharpe ratio is based on 5 year return %/standard deviation %.

Forecast 1-m return %, Forecast 1 yr return % and 1-year chance of gain % [Green circle] : However, looks like the ValuEngine models think that ADR type of stocks are due for a correction. S2 stocks have higher values than S3 stocks for these variables. One important point I like to make here: VE models forecast based on the latest data including analysts' expectations. But if high growth ADRs keep on outperforming the most optimistic expected earnings, then the forecast will be wrong. For such stocks it is better to look at the track record i.e. the 12-m return % and the 5-year return %.
PE Ratio, M/B Ratio and P/S Ratio: [Purple circle] . Here, the contrast continues. The ADRs have very high values for these three variables [not good fundamentally, but can also be seen as being priced for higher future growth]

Beta: Last Purple bar: The ADRs in S3 have very high Betas and are more sensitive to general market changes.

If you want to know the comparison for the other variables, see the color-coded key squares at the bottom of the table.
Well, then we can see that ADRs are very much different from American big caps.

If you like the ADR story, but are afraid of their volatility, why not take those ADRs which have the highest Sharpe Ratio? For this, we can use Image 4, the Sharpe Ratio map which shows the areas with highest Sharpe Ratio [tending towards Red color]. We select these areas with the mouse, and the list of such stocks [whether ADR or DJIA component] is revealed in Image 5.

Right at the top in Image 1, are the list of all the stocks in S3. Boeing [BA] is up there among the ADRS.

Stock Market Visual Analytics Using Self-Organizing Maps

2010-04-17T22:10:00.008+10:00

1: Self-Organized Clusters DJIA Components and ValuEngine Models Output
April 10

2 Self-Organized Clusters: DJIA Components and ValuEngine Models Output

April 173. Statistics of the Clusters: Standard Deviation of Mean of Range from Mean of Entire Data Set for each Variable

4 Non-DJIA Stocks in DJIA Stocks Cluster

This post is a continuation of recent previous posts aimed at getting a better understanding of the use of SOM as visual analytics for equity market. For methodology, terminology, and understanding of the ValuEngine models from which this SOM is constructed, see previous weeks' posts. Let's make the analysis as clear as possible by using as few words as possible:1. Cluster S3 [Yellow] has the best overall statistics. See Image 3.It has got slight Undervaluation, low Volatility, higher Sharpe Ratio, low Beta, and high market cap. A Quality type of cluster.

2. What's in S3? Mostly the DJIA components [those ending in 'X'] except for BAC and AA which are in S1. The non-DJIA stocks are all Long stocks [ending with'L'] There are no stocks ending with 'S' [Short] in S3 cluster.

3. Therefore the DJIA components stocks are currently Bullish and since the DJIA represents the market, the market is currently Bullish.

4. How Bullish? Most of the other 'L' stocks are in cluster S2 [Red]. And most of the 'S' stocks are in cluster S1 [Blue]. How different statistically are the S2 and S1 stocks? Overall, not much. Therefore the Bullish signal is not a strong one.

5. Taking a look at the non-DJIA stocks in S3, [see Image 4, they are non-consequential [see Pink highlight] with very low Average Volume except for Assured Guaranty and Seagate Technology. Therefore the broader market does not have a strong Bullish signal.

6. Comparing last week's clusters Image 1 with this week's Image 2,last week

there were more non-DJIA components stocks in the DJIA components cluster [C1] and they all had an 'L' [Long]. An indication of broader market bullishness. This week the cluster with the DJIA components has shrunk in area.

Lessons In Self-Organizing Maps: Part 2

2010-04-10T16:44:00.007+10:00

1. Clusters" (All data as at market close 9 April 2010) 2. Cluster Statistics
3. Principal Components Analysis [PCA]

4. Valuation % Map

5. Stocks with highest probability for Short

This week, we continue our lessons in Self-Organizing Map. [for those who have just joined us, please see previous posts for methodology and terminology] We add another dimension to our analysis by including the Short screens in ValuEngine. Thus our labels on the SOM now contain (a)Sector [B=BasicIndustries, F=Finance, H=Healthcare, ND=ConsumerNonDurables etc] (b) The ValuEngine model that output the stock viz V=Valuation [Standard], G=Growth [Forecast], Q=Quality [Star]. L=Long as output by Long Screen for each ValuEngine model. S=Short as output by Short screen of each ValuEngine Model. X=Components of the DJIA. Thus CX is a CapitalGoods stock that is a component of the DJIA. TGL is a Technology stock from the Growth model Long screen. SVS is a ConsumerServices stock from the Short screen of the Valuation model. We feed the data of all the 6 screens [ Long and Short of each of the three ValuEngine models] plus the DJIA components into Viscovery and create a Self-Organizing map.

Image 1: The data is self-organized into three clusters: C1[Blue], C2[Red, C3[Yellow]. C1 contains all the DJIA stocks [see the X in their labels] and an asortment of stocks from each model as well as from various sectors [look at the V,G, and Q and look at the sector labels F,H,T,E,C etc]. There is nothing unique about this cluster. It is a big cluster and therefore its distinctiveness is 'diluted'. The only thing one can say is that most of the stocks are Long [L], and by implication, since the DJIA stocks are in the same cluster, the DJIA is a 'Long' for this week. But, like I said, the area covered by the cluster is huge, and there are only three clusters [one is a very small cluster], so the signal is not so strong.

Now, if you look at C2 [Red] cluster, it can be seen that it contains all the Short [S at the end] from various models V,G,Q as well as from various sectors. One implication is that sector play is not a huge thing at the moment. Again, the Short signal from this cluster is not strong as the cluster occupies a big area on the SOM.

Pay special attention to C3 [Yellow] cluster: It is a very small cluster, and there are very few stocks in it. Small clusters with nodes very close to each other are distinct and considered a strong signal. What are the stocks in C3? See Image 5: Although 5 nodes in C3 are occupied, there are only 2 stocks: Pinnacle Entertainment [PNK] and BRF-Brazil Foods S.A.[BRFS]. The reason why there are only 2 stocks is that the stocks are present in more than one model screen. The actual nodes on the SOM which can't be seen because the area of C3 is so small are SQS,SVS, NDGS,NDQS,NDVS. SQS and SVS belong to Pinnacle since it is ConsumerService sector and it appeared in the Quality [Q] as well as Valuation[V] screen. NDGS, NDQS and NDVS belong to Brazil Foods, since it is in ConsumerNonDurables sector and appeared in all three model screens, V,Q and G. But what is important is that ALL the 5 nodes are Short [S] nodes, indicating a strong signal for Short, which when combined with the very small cluster area makes it a very strong signal for Shorting PNK and BRFS. Lets just observe that the Closing price for PNQ is 11.10 and BRFS is 13.72 and monitor the results for the next few days.

Image 3: If you look at the PCA Table for this week, the highest value PCA on the SOM model and therefore the factors influencing the market both are 1-year Chance of Gain, Actual EPS and Forecast EPS. [highlighted] I took out those with PCA1 of less than 0.1 and you can see that Momentum [12-m return], Valuation, Beta have been unchecked. Even the fundamental ratios like P/E, M/B and P/S are not significant. It looks like for this week, the market is focused on Analyst Estimates and the difference between Actual and Forecast EPS.

Image 4: Lastly, in another indication of how strong the Short signal is for PNK and BRFS, take a look at the Valuation % map. Here you see that cluster C3 stands out as the most prominently high Valuation % area [see scale at bottom, low Valuation % is more desirable than high] . if we extend the Neighborhood of C3, two more stocks LBTYA and MDZ have the same characteristics.

# Want to know even more about SOMs? Visit the Viscovery web site at www.viscovery.net and go to Resources. Also, Under Publications/Web/Special Viscovery Applications, you can find my past Blog posts on using SOM in Finance.

ITS TIME FOR A DEEPER EXPLANATION OF MY SELF-ORGANIZING MAPS

2010-04-03T13:20:00.009+11:00

1. This Week's Picks
2. Clusters

3. Cluster Statistics

4. PCA (Principle Components Analysis)

5. Market/Book Ratio Map

6. Volatility Neighborhood Map

7. Members of the Volatility Neighborhood.

This week, I decided that instead of merely recording the performance of my stock picks, I would explain in greater detail, the rationale and advantages of using Self-Organizing Map. I would explain in as simple a way as possible so that the majority of readers will understand.

* BTW, no point to re-balance the stock pick portfolio every week. ValueEngine models are fundamentals-based and work best when re-balanced monthly. If you look at Image 1 This Week's Picks, 5 of the 12 stocks also appeared last week.

1. Why use SOM? Because I believe in a syncretic approach to portfolio construction. ValuEngine has three models which cover the whole prism of market characteristics: Value, Growth and Quality. All the models also account for the technical characteristics of the market viz: Mean Reversion and Trending (Momentum)

At any time, the market is never just Value, Growth or Quality but various combinations of Value, Growth and Quality characteristics. A SOM is able to capture via visualization, the nuances of this combination of characeristics. A SOM is also able to spot more clearly the beginnings of regime switches- those sudden structural changes in the market/ See http://www.technifundamentals.com/search/label/regime-switching%20models
2. Heuristics are an important element of investor decision-making. Above and beyond the quantitative analysis, human domain expertise is still required to make the final call. Tweaking a SOM manually on the weighting of each input, the map's tension and resolution, different kind of transfer functions etc also need domain expertise.

3. Markets are Complex Adaptive Systems. A linear approach is often necessary for practical reasons, but never sufficient. A SOM is a collection of locally linear regressions which because of the infinitesimal initial differences of each, becomes globally non-linear. With non-linearity comes emergent properties that were never identified before.

With the above thoughts in mind, let's use this week's analysis as a platform for our lessons in SOM:

a. Image 2 Clusters was constructed by aggregating ValuEngine's screens for Standard (equivalent to Value), Forecast (equivalent to Growth) and Star (equivalent to Quality) stocks. In addition, the DJIA components was added as a check and balance to see their positioning on the SOM. Each stock was then labeled by its sector and its model. For terminology, see http://www.technifundamentals.com/2010/03/stock-picking-is-back-in-fashion-part-3.html. In a SOM, the smaller the cluster and therefore the closer to each other the objects in a cluster are, the more distinct the cluster is. In Image 2, Yellow cluster (S3) is the smallest. Most of the stocks in it are from the Growth model (suffix G) screen including the two DJIA component stocks Alcoa and Bank of America. Growth stocks are the flavor of the moment. Growth stocks pay less emphasis on valuation, PE, volatility etc.

b.Image 3 shows that taking into account the fact the dominant market emphasis is on Growth, the best overall statistics belong to cluster S3 which has high values for Growth characteristics of Momentum, Liquidity, Expected EPS and Beta. Valuation, low Volatility [risk aversion], P/Es etc take a back seat.

c. Now we can use Principle Components Analysis (PCA) to analyze which of the model variables are the most influential in the positioning of stocks on the SOM. PCA is a method for representing the main structural features of a data set by sacrificing the small information. This is achieved by transforming the original non-linear coordinate system of the data set into one where the variables are linearly independent. Since PCA depends on the input, I removed those ValuEngine model variables which at the moment have very small influence [snall PCs] on the SOM model. These include the 1-month forecast, the ValuEngine ratings and the Actual EPS. The highlighted Pink variables are the ones with high Principle Component values which means they were very influential in modeling the SOM. Volatility has a PC of 0.92, Last 12-month return [equivalent to Momentum] has 0.89, Beta is 0.84 and M/B Ratio is 0.75

Now you can see that Growth stocks with high Momentum, high Beta, high Volatility [not desirable in Value model], and high M/B ratio [not desirable in Value model] are the main characteristics.

d. Next, since Volatility has the highest PC value, we can take the Volatility perspective from our SOM [image 6] and select a node of the map at the point of highest value viz tending toward color Red at the highest end of the scale. We can then use a neighborhood algorithm to construct an area around our selected node, and the Volatility neighborhood is defined: the pink dots. We can then list the stocks that are in the Volatility neighborhood. Image 7.

Summary of analysis: From the SOM , we have seen that the emphasis in the market is now on Growth with all its accompanying characteristic of disregard for valuation, volatility, high book and earnings values etc. But the SOM is only a guide for your decision making. If you are, like me, a Value and Quality guy, you may decide to take a back seat for the moment in the belief that the market rally is unsustainable. If, on the other hadd, you think that even with potential interest rate increases, gradually decreasing effects of stimulus action and high jobless situation the US economy has taken a turn for the better, then you can bet on the Growth stocks.

Last point: Distances of stocks [nodes] within a cluster is a measure of degreeof similarity, but comparison of inter-cluster distances is is not valid. "If a group of objects is close together (in data space) they have obviously much in common and can thus be assigned to a (micro)cluster. If they are spread over the whole map, they are not likely to be very similar. Data points included in a SOM cluster are always similar to each other, however the oppositie is not necessarily true: It may occur that two clusters are mapped in different regions of the SOM, although they are adjacent in the original data space. This can happen because the SOM is a perceptional space which makes a compromise in preserving the topology of the data distribution. In practice this fact doesn't play a big role because in applications one is normally interested in the clusters itself"

The Healthcare Sector on a Self-Organizing Map

2010-03-28T17:55:00.003+11:00

Healthcare stocks > market cap $1 Bil and Average Volume > 500,000
Statistics of clusters

1-month Forecast Returns %

Valuation (+/1 %)

Since the Healthcare sector is in the spotlight, and also because some readers of this Blog requested it, I did a SOM of the Healthcare sector. Here are the images, and they are self-explanatory.Those who need more explanation can write to me. Just to summarize that in terms of Valuation (the more negative the Valuation the more desirable), 1-month Forecast Return % and market cap [the bigger the better in terms of risk and liquidity], S3 is the best cluster. You can look at the cluster statistics image and see the color key for each variable. S4 is a loner and is occupied by ISIS. Of course you can also look at the Valuation and 1-month Forecast Returns % maps and pinpoint the ticker symbols that occupy the space you like.

Market Insight Using Self-Organizing Map: Mar 27 2010

2010-03-28T16:55:00.005+11:00

1. Last week's results

2. This week's clusters 3. Statistics of each cluster [standardized]

4. This week's stock picks

* Note: For reference on methodology and format please see previous post athttp://www.technifundamentals.com/2010/03/stock-picking-is-back-in-fashion-part-3.html . [It is recommended that you read and understand thoroughly before looking at the stock picks for each week]

As mentioned in previous post, last week was not a good time to use my stock picks. And it turned out right. The portfolio fell by 0.24 % while the DJIA advanced by 1 %. Not disastrous, but a mediocre situation. From the 16 to 26 March the performance of the different ValuEngine models were: Value (-3.06 %); Growth (-1.50 %) and Quality (-1.91 %). The differentiation is thus not significant and this bears some implication that the market's performance this coming week will again be lack-lustre, with fewer opportunities for good gains even with the best stock picking. Image 2 shows that the stocks gathered from the three ValuEngine models are divided into three big clusters, with the stocks [a node] spread out over each cluster. This lack of tightness with big distances between nodes indicates lack of distinctiveness which makes it difficult to outperform the market. Stocks suffixed with 'X' are the components of the DJIA and they are nearly all in S1 the Blue cluster. Image 3 cluster statistics shows S3 (Yellow) to be the best cluster and the stocks in this cluster are predominantly from the Growth model. There is no dominant sector in this cluster and stocks come from sectors as diverse as Energy and Consumer Services. Again this is an indication of a wishy-washy market.

Using Self-Organizing Maps To Identify Mispriced Stocks Within An Industry

2010-03-21T17:59:00.005+11:00

Oil Industry Stocks: Market Cap> $1 Bil and Average Volume> 500000 daily
Self-Organizing Maps [SOM] are a class of Artificial Intelligence that cluster objects according to their degree of similarity. SOM are able to simultaneously take into account many variables representing the characteristics of the objects, and place them in clusters. Within each cluster and on the map as a whole, the distance of an object [node] on the map relative to all other nodes, is a measure of the degree of similarity of the object to its 'siblings'. Taking the example of the stocks in the Oil Industry, Service Companies like BJ Services, Halliburton or Schlumberger should be in the same cluster, or at least close to each other. Or Oil majors like Total, BP, Exxon-Mobil should also be close together. There are more than 20 variables [characteristics] of each stock that determine the construction of my SOM and these include fundamentals like its P/E, M/B, analyst estimates and surprise, volatility, Sharpe Ratio, 5-year return etc.
Therefore, taking into account all these factors, stocks in the same cluster or close to each other have a greater degree of similarity. Stocks from different niches of the industry may be found close together and this is regarded as an anomaly that may be an opportunity for arbitrage. For example, if rig operator Rowan is far away from fellow operator Transocean on the map, it pays to know why and if Rowan is undervalued. Action can then be taken to go Long on Rowan and Short on Transocean.
The image above is a SOM of oil industry stocks with market cap>$1 billion, and daily average vilume > 500,000. That the oil industry is a diverse industry with many players doing quite different things can be shown by the great number of natutally formed clusters on the SOM viz 11 clusters.

NOW for a limited period only, I am offering Free to industry analysts and investors who request by ssending me an email to tiankhean@gmail.com , a SOM of their industry of interest. The industry expert will be able to identify stocks which are out of sync. In addition to the SOM [sent as a .jpg], I am able to give the figures for each of the clusters pertaining to their following characteristics:

Valuation [+/1 %], ValuEngine 1 to 5-Engine Rating, Average Volume, Last 12-Month Return %, Sharpe Ratio, 5-yr Return %, Forecast 1-m Return, Forecast 1-yr Return %, 1-year chance of gain, Volatility %, Expected EPS, EPS Surprise, Market Cap, P/E, M/B, P/S, Actual EPS, Forecast EPS, Beta.

* Please note that Industry SOM is an experimental product to gauge market demand and if successful, I will have to consult with ValuEngine Inc for future pricing and mode of distribution.

Market Insight: 20 March 2010

2010-03-20T18:46:00.011+11:00

1. Self-Organizing 'weather map' of 1-month Forecast Returns %

1.March 05 Results
2. Clusters

3. Cluster Statistics

5. Stock picks

This Blog will now be updated on a weekly basis and previous weeks' performance will be tracked. * unless there is a dramatic change in the market situation caused by unforeseen external events in which case an update will be posted earlier to seize opportunities that surface. The methodology for analysis and format will be the same. For reference on methodolgy and format please see previous post at
http://www.technifundamentals.com/2010/03/stock-picking-is-back-in-fashion-part-3.html . [It is recommended that you read and understand thoroughly before looking at the stock picks for this week]

Image 2 shows performance of the stock picks on March 05. The portfolio performance is 8.27 %, During the same period, the DJIA was up 1 %.For this week, the general analysis is that the market's performance will be mediocre or bad. Top image shows the Self-Organizing Map of the 1-month Forecast Returns % of our stock Universe comprising the portfolios generated by the ValuEngine Valuation, Growth, Quality models plus the DJIA component stocks. The intensity scale at the bottom denotes the forecast returns % for 1-month. Low forecasts tend towards color Violet and high forecasts tend towards color Red. Currently, most of the map is in the range of 1-month forecast returns -1 % to 3 %. Only a small area in the northwest corner have a higher forecast returns of 7% toi 17 %. The stocks in these high nodes are TRGT, DFT, KEY and HRBN.Image 3 shows that the stocks selected by ValuEngine's Valuation, Growth, and Quality models and the DJIA components are spread over 3 large clusters and 1 small cluster. Cluster identity of each model is fairly distinct for Growth (suffix G) and the DJIA components (suffix X). But Valuation (suffix V) and Quality (suffix Q) stocks occupy Cluster C1 [Blue] together lessening the differentiation between Valuation and Quality. Image 3 Cluster statistics indicate that although C3 has the best average statistics of the 4 clusters, the difference is not significant. Image 5 shows the stocks in the selected best cluster C3. Most of the stocks are from the Growth model (G) with high Beta and are more vulnerable to market correction.Prognosis: Unless you have a high risk profile, it is best to stay out of the market for this week.

Stock Picking Is Back In Fashion Part 3

2010-03-06T00:54:00.003+11:00

Self-Organizing Map of ValuEngine Model Portfolio and DJIA Stocks By Sector
Statistics of Clusters S1,S2 and S3 Above As Standard Deviations From Mean

List Of Stocks In Best Cluster S3

Here's a novel way to pick stocks using a Self-Organizing Map (SOM). * For introduction to SOM read my previous posts on this subject or Search Wikipedia. It's quite a long-winded process so I bullet the sequence below:

1.Using ValuEngine Institutional, generate the portfolios for the Standard(Valuation), Forecast(Growth), and Star(Quality) models. The models can be approximately equated with Valuation, Growth and Quality approach as explained in Stock Picking Is Back in Fashion Part 1 . Next, generate similar data for the components of the DJIA. So now, we have four portfolios.

2. In the Sector column of each stock of each portfolio label them using the following terminology:

Portfolios:V=Valuation portfolio stock. G=Growth portfolio stock. Q=Quality portfolio stock. X=DJIA stock

Sectors: B=BasicIndustries, C=CapitalGoods, D=ConsumerDurables, ND=ConsumerNonDurables, S=ConsumerServices, T=Technology, TP=Transportation, E=Energy, F=Finance, H=HealthCare, U=PublicUtilities.

Thus, a stock from the Growth portfolio and in the Healthcare sector would be labeled HG. A stock from Quality portfolio in the Finance sector would be FQ and a stock which is a component of the DJIA and in the Technology sector would be TX.

3. Feed all the data of these models into Viscovery, the SOM software, and generate the self-organizing clusters as shown in top image. As intended, the SOM takes into account all the 20 or so variables of the models, the degree of relationship between and among all of them [Euclidean distance in mathematical terminology], and places the stocks in their rightful position on the map. # Actually its not that simple. Values, have to be normalized, some outliers clipped, appropriate transfer functions such as Sigmoid or Logarithmic have to be applied, and a decision has to be made on how find a resolution you want the Map to be.

Now for the analysis: If you look at Cluster S2 the Red cluster, you can see that the DJIA component stocks are mostly in there (those labeled with an X at the end). And this is rightly so since DJIA component stocks do have many common characteristics, such as being big cap, having high daily volume, being less volatile and in general having lower annual returns than non-index stocks since there is always someone buying and selling them. So the SOM is correctly putting them in the same cluster despite the stocks being from different sectors.

Now that we have confirmed that the SOM is correctly configured, we take a look at which cluster has better overall statistics. (Except for Valuation and Volatility where negative values are more desirable, the other variables should be more desirable the higher in positive territory they go). * Note that these statistics have been normalized and are now represented as standard deviations from the Mean. Cluster S3 has the best overall statistics. [see middle image above]. Note that Cluster S3 contains stocks from the different model portfolios i.e. you find stocks with the extensions G,V and Q and even two DJIA stocks (FX and BX). Bottom image shows the list of stocks in S3. Those that appear more than once are stocks that have been selected by more than one model. This synthetically generated portfolio is eclectic, and boasts stocks as diverse as Tata the Indian automobile Company, to KKR the Private Equity guys to Nestle and Alcoa. One can assume that it would a be fairly robust portfolio to hold for a month or so. The next month, ValuEngine would generate a different list of stocks as the market evolves. The same methodology can then be applied to re-balance your portfolio. The total number of stocks may not be the same as the size of the cluster will be different. Actually we may also have a situation where there is not much difference in the statistics between each cluster and it would be less justifiable to use this methodology for stock picking.

Stock Picking Is Back In Fashion Part 2

2010-03-04T00:09:00.008+11:00

The final choice: Prudential, the all-round best stock

Stocks of the ValuEngine Quality Portfolio (5 Engine stocks in the 1 to 5 Engine Rating System)

Sector Valuation and Performance

In the previous post, we determined that the market was in a Quality mode, and stocks of the ValuEngine Rating (Quality) model outperformed the Valuation and Growth models. In this post we analyze the stocks in the Quality portfolio as output by ValuEngine's screening tool. Here are some observations:

Eight of the twenty stocks are foreign stocks ( 40 %) i.e. ADRs of foreign Companies. Considering that the number of ADRs in ValuEngine's stock Universe is about 600, ADRs are thus disproportionately represented in this portfolio of Quality stocks. This itself could be an indication of what we said before, that U.S. Companies with international business exposure have greater griwth prospects.

Column E denotes the % by which a stock is Under/Overvalued. Column E indicates that all the stocks in the portfolio are undervalued. So one of the components of Quality is good valuations

3/4 of the stocks are from the Finance sector. If we then take a look at the table of sector valuations and performance, we see in the ranking of undervaluation, only Healthcare (-5.81%) and Technology (-4.27 %) are more undervalued than Finance. But Technology has outrun itself with a last 12-month return of 98.91 % and a current P/E of 28.27. As for Healthcare we know that the pharmaceutical giants have many obstacles to overcome whether its the Obama administration's reform proposals or the problem of expired patents and competition from generic drugs. So the Finance sector with moderately undervalued and with moderate last 12-month return of 64% and a low P/E of 18.93 is not a bad choice.

Except for Braskem the Brazilian chemicals giant, all the stocks effectively have negative 1-month return forecast. So clearly, ValuEngine's Quality model is not about short-term forecasts.

Now we come to the next stage of picking the best stock(s) in terms of overall Quality as defined by risk/reward ratio. In ValuEngine, we can use the Composite Ranking score, and even though all the stocks in the portfolio are 5-Engine, the Composite Ranking score makes its possible to further differentiate them. But its not a straight forward method of just picking the stock with the highest Composite Ranking score. There are other factors to consider, such as the stock's liquidity and market cap. In general,for the market at this point of time, an average daily volume of > 1,000,000 shares and a market cap > $1 billion is desirable. The first ranked BNP Paribas is excluded because of its low liquidity, and British Insurer Prudential is the preferred stock pick. (See Table in top image)

*Note: Of course we know that Prudential has been in the news lately because of its $35 Billion bid for AIA, the Asian insurance arm of AIG and the necessity to issue $20 billion of toights . Its share price has suffered and as at March 3 is $52.83. But one of the advantages of a quantitative system like ValuEngine is that it helps to filter out the emotion and over-reaction to news.

Single-Stock Picking Is Back In Fashion Part 1

2010-03-03T18:16:00.009+11:00

Valuation Model Portfolio Performance

Growth Model Portfolio Performance

Quality Model Portfolio Performance

After the Black Swan events of 2007 and 2008, comes what economist Noriel Roubini calls the 'New Normal'. The age of easy credit has passed, and Americans' consumers who account for 70 % of GDP have been battered and their savings depleted. The New Normal is an age of slower growth for America. Even for world economic growth as a whole,it is not yet certain that China and the other emerging countries will be able to take up the slack. The U.S. with 4% of the world's population consumes $10 trillion. China and India, with 35 % of the world's population consume $1 trillion-much of their GDP growth is based on Exports, Investment and Government spending. Stock markets in general will dither and trade in tight bands for many years. It will be difficult for individual investors to get good returns unless they have (1) a large capital base for investing (2) or they trade more frequently and use leverage or

(3) They do very good stock picking. As the world economic scene changes, and the countries of Asia, the Middle East, Latin America and Africa account for an increasingly larger proportion of world economic growth, Companies in the developed world with international business opportunities and operations will be the only ones with high earnings growth. WalMart, Proctor & Gamble, Caterpillar, McDonalds and IBMs are examples of such multi-national Companies . In the next few posts, I will outline various strategies for single stock picking using ValuEngine's models. Part 1 here, deals with identifying the current dominant market characteristic. The market at any one time places different degrees of emphasis on valuation, growth, and quality. These three characteristics (or modes) of the market can be approximated by ValuEngine's three models: 1. Valuation Model (valuation) 2. Forecast Model(growth) 3. VE Rating Model (quality). It is only by identifying the current dominant market mode that we can effectively pick the stocks of the moment.

The tables above show that from 12 feb till today, Valuation had a return of 5.99 %, Growth: 2.78%, and Quality: 7.27 %. Overall, the SP500 for this same period had a return of 1.04 %. Therefore the current dominant characteristic of the market is an emphasis on Quality. [Please read below for what Quality means]. * More details on the models: By valuation emphasis, we mean that the market is looking for stocks which are fundamentally undervalued. ValuEngine uses a closed-loop analytical [equation-based] model with variables of the stock's fundamentals, to arrive at a conclusion on valuation. By growth emphasis we mean that the market places less emphasis on current fundamentals and more on future fundamentals (growth). In this mode, variables such as the stocks' liquidity and self-feeding momentum are important. A simulation methodology to do forecasting of 1-month returns is more appropriate than a closed-loop analytical model in this case. By quality model, we mean that the market places emphasis on the overall risk/reward ratio of a stock i.e. its Quality. A stock may have fantastic fundamentals but remain undervalued for a long time. A growth stock may have fantastic momentum, but also be very volatile. Therefore quality model takes into account the risk/reward ratio. This mode of the market is best modeled with a classification algorithm where stocks are put into various categories of quality. In ValuEngine's case they are rated from 1-Engine to 5-Engine, with 5-Engine being the most desirable.The next post will analyze the stocks which make up the Quality portfolio.

Note: As at this date, the variation in performance of the different model portfolios is significant. Quite often, there will be little difference in the performance of different model portfolios. In general we can state that when the difference in performance of different model portfolios is significant, picking of stocks with the dominant model characteristics will have greater profit potential. When there is less difference in the performance of the different model portfolios it will be a moribund market where stock picking will not be much of an advantage

New Technologies For The Modeling Of Financial Markets

2010-01-17T17:15:00.010+11:00

A new generation of tools is starting to make an impact on the modeling and analysis of financial markets. For the sake of simplification, and because of the limits in computer power, modeling of financial markets was hitherto being done with the tools of traditional statistics. But the recognition that financial markets are non-linear complex adaptive systems also brings with it the recognition of the inadequacy of traditional statistics for the modeling of financial markets. Modeling for the purpose of decision-making is comprised of the following processes:
(1) filtering of relevant information ; viz de-noising (2) classification, (3) pattern recognition (4) forecasting (5) sensitivity analysis of model variables (6) optimization (7) prioritization of tasks. Here are some examples of how new technologies can be used fro modeling financial markets, with illustrations of the applications of such technologies in everday life:
De-noising: The first step is to filter out from data, what is considered irrelevant for the task at hand. Moving Averages are an example of de-noising-they filter out information above a certain threshold. Previously, much of this de-noising was done with conventional tools from digital signal processing such as Fourier Transforms. But Wavelets (seehttp://www.amara.com/IEEEwave/IEEEwavelet.html can now do a much better job. for example, digital image compression is now much more efficient thanks to the use of wavelets. Wavelets are also very good for the synthesis of musical instrument tones. They have not been able to fully replicate the sound of a Stradivarius or a 1959 Gibson Les Paul, but produce reasonably useful tones of vintage guitar amplifiers such as the Fender Twin Reverb and the Vox AC30
Sensitivity analysis of model variables: Model variables need to be tested for sensitivity to perturbation. i.e. what happens to the other variables when one variable is subject to perturbation. Traditional statistical tools cannot do a good job when it is realized that complex adaptive systems exhibit 'emergent' properties viz 2+ 2 >4. Neural Nets with their synthetic synapses and dendrites mimicking the human brain, can do a better job. The 'weights' in all synapses change in a networked and non-linear way as some parameter is changed, and these weights are measurable, thus making for a more accurate sensitivity analysis.
Pattern Recognition: The human brain can holistically sum up all the features of a face to let you recognize someone you know. The new generation of digital cameras can recognize a face, a smile, or a blink. Biometric scanners will let you through without 100 % matching of your fingerprints with their record of it. Military sofware can differentiate between an enemy and a friendly submarine. These pattern recognition tools contain classification and fuzzy logic algorithms that allow for the identification of 'subtle signals'. In a world where Black and White are just different shades of Gray, such tools are better for detecting subtle patterns laced with noise.
Classification: A major part of modeling consists of the act of classification. Here, we have seen from previous posts in this Blog, how Self-Organizing Maps can classify objects according to their degree of similarity in a more natural way than the tools of traditional statistics; where the results depend on your choice of parameters for defining the boundaries of each cluster, whereas a SOM is self-organizing. SOMs have been used for the classification of wines, classification of credit card holders, and the classification of women's erratic shopping habits
Forecasting: Decision-making often involves an imagination of future scenarios. It is hardwired into us to try and predict, even if we realize how flimsy our basis for the prediction is. One possible way to overcome this is by a 'consensus' of models; i.e. your forecast is more likely to be accurate if it is an average, or median of the forecast of a group of experts. It is now possible to use Swarm intelligence to forecast. Swarm intelligence (see http://en.wikipedia.org/wiki/Swarm_intelligence ) is the collective intelligence of autonomous, distributed, self-organizing systems. Such as ant colonies, a flock of Birds, a school of Fish, human beings in a traffic jam. 'Bots' can be created, each seeded with the objective of maximising profit, and let loose to behave and inter-act with each other. The winning Bot, or the average result of the Bots can be used as a forecast.
Optimization: Optimization used to be about Linear Programming as in Operations Research, or Gaussian steepest descent, going down the slope of steepest descent in a multi-dimensional model landscape. But the current use of Genetic Algorithms is a quantum leap for optimization. Multi-objective optimization is now much easier to do. Multi-optimization is often seen in the optimization of transportation systems capacity such as railways and airlines where there has to be compromise between different objectives. Genetic Algorithms breaks the task down into the construction of strings of genes comprising model variables and parameters, giving them a purpose in their Life and a measure of fitness for survival. There are many algorithms for the weeding out of unfit individuals some of which sound like Hitler's doctrine on Aryan supremacy.
Evolutionary AlgorithmsAll complex adaptive systems are continuously evolving, and indeed, co-evolving. Co-evolution means taking into account the agent's effect on its environment, the response of the environment, and the further response ad infinitum of all parties. In the past, trading systems were back-tested, which was a major flaw. Then there was testing using 'hold out' samples. Now, with evolutionary algorithms, we can have walk-forward testing, where each step of the time series evolves before being tested.
Are all these tools ultimately useful for trading the markets? Yes. The shorter the time frame of investment, the more useful they are. Will analysts, and the need for a financial background be redundant? No. We still need someone with an understanding of financial markets for choice of model variables, for pre-processing of the data, and for analysis against the macro investing environment. But an analyst with knowledge of and access to these technologies will be a better analyst. He will be able to know what are the current factors driving the market, what are the factors driving a particular stock, and go on from there.

Using Visual Analytics To Interpret Econometric Forecast of S&P500 Returns for 2010

2010-01-01T17:01:00.010+11:00

1. ValueEngine's Forecast 1-month Return % of SP500 stocks
2. ValuEngine's Forecast 1-year Return % of SP500 stocks

3. ValuEngine's Chance of Gain % in 1-year SP500 stocks

Visual Analytics is the science of interpreting data by using advanced technologies. When used in the correct way, visual analytics significantly improves insight into large size and complex databases. More importantly it allows a holistic interpretation of issues presented from different perspectives. To illustrate my point, I used the screening facilities of ValuEngine to create a database of the SP500 stocks with information on their valuations, P/E, volatility, market capitalization, Beta, Market/Book ratio, liquidity etc, including ValuEngine's [www.valuengine.com ] econometric forecast of their 1-month return %, 1-year return % and % probability of the stock making a gain in a 1-year. The resulting spreadsheet was transformed into Self-Organizing Maps with Viscovery [www.viscovery.net ].* The maps are presented above. (For the moment disregard the alphabet that pins the position of each stock's location on the map. These represent the Sector the stock belongs to e.g H=Health Care, T=Technology, F-Finance etc. We will do a sector analysis in the next blog post].

Looking at the three maps we can interpret thus:

1. The first map of 1-month forecast return % shows >95 % of the stocks are forecasted to have a 1-month return of -1 % to 1 %.[see the Color scale at the bottom of the map and note how uniformly colored the map is]. The interesting fact to note is that the highest 1-month forecast is only 1%. This indicates that the market is not expected to have significant changes in the next one month. Note#: Pre-2007, forecast 1-m returns % maps were more varied i.e. more colorful, and the high end of the scale reached as high as 10 %.

2. The 1-year forecast return % map is only slightly more colorful, and the scale at the bottom ranges from -80 to 10 %. The greatest areas of the map are in the Yellow/Orange range with a 1-year forecast return of -15 % t0 10 %. A significant area of the map in the North-West corner [Orange] has a positive 1-year % return of 1 % to 10 %. Again, in comparison with past [pre-crisis] maps that I have constructed where some areas of the map posted 30-40 % annual returns, what we have now is a picture of mild optimism, and not rip-roaring bullishness.

3. Finally to further pin down the situation, the bottom map shows the probability % of making a gain in 1-year. Put in this way, there are only two outcomes for a stocks-either it makes a gain, or it makes a loss. Now if you look at the bottom scale of this map, the chances of making a gain in 1 year ranges from 11 % to 60 %. But 90 % of the map area has only a 35 % to 45 % chance of making a gain. If you analyze by sectors you will see that Energy stocks (E on the map) and Utility stocks [U on the map] have > 50 % chance of making a gain in 1 year.

* Self-Organizing Maps are a form of Artificial Intelligence with the ability to organize data into clusters with similar characteristics. Their advantage is that they are able to take into account the non-linear relationship of all the model variables with each other. Distance of nodes from each other represents degree of similarity, the smaller the distance, the greater the degree of similarity.

Having looked at these forecasts, here are two reaons why you should be cautious about statistical/econometriv forecasts.

1. Like everything else in the world, forecasts are dynamic and constantly evolving. Thus the weather man and the analyst and economists have to constantly update their forecasts. (Unless you are Nostradamus)

2. Most econometric forecasts assume a normal [Gaussian, bell-bottom] distribution curve even now, after Nasem Taleb's book The Black Swan warned us about long fat tails and how extreme events can occur more frequently than thought. A Normal distribution curve's boundaries extend only to 3 Standard Deviation. But some of the events in the recent melt-down have been calculated to have a Standard Deviation of 20-60 Standard Deviation.

Why Warren Buffet Investment Style Is Not Suitable For Babyboomers

2009-12-26T21:51:00.007+11:00

1. Annual Returns of ValuEngine 5-Engine Stocks (monthly rebalancing) 1991-2009
2. ValuEngine 5-Engine Stocks vs SP500 1991-2009

In my humble opinion babyboomers who follow the Warren Buffet style of picking undervalued stocks for buy-and-hold are at a great disadvantage. The most important factor that babyboomers should take into account of when investing, is that they do not have time on their side. And time is exactly what they need if they want to see their investments make gains. But the nature of modern financial markets makes this style of investing difficult:

1. The business environment for Companies changes at a faster pace, and changes in the technological, economic, and political environment render obsolete whole industries and business models , not sparing even very established Companies. Dotcom 'blue-chips' like Sun Microsystems, Yahoo, Nortel are a shadow of their former selves. General Motors and General Electric-will they be able to resuscitate themselves? Did we ever imagine that mighty CitiGroup, AIG and Bank of America could be a whisker away from bankruptcy? Do we have the confidence that Coca Cola, Proctor & Gamble, Exxon-Mobil will, even if they survive, grow at the same pace?

2. Modern financial markets are increasingly interlinked. Such that what happens in the equities markets affect and are affected by what happens in currency, commodities, bonds and real estate.This makes markets more volatile and babyboomers have to consider whether they can survive the drawdown on their portfolio in particular years.

3. The presence of algorithmic trading causes markets to be out of kilter (fundamental equilibirum)for longer periods. That is, while it is fine to say that stocks are over or undervalued, and they will revert to their equilibirum based on theories of investment, the fact is that stocks can go on being over or undervalued for a very long time. Algorithmic or programmed high-frequency trading feeds on momentum and liquidity and have very little relation to fundamentals. Nobody really cares how or why they make money as long as they do make money.

4. Especially for babyboomers who invest solely in the U.S. market. They have to consider that there are tectonic forces at work such as the rise of the BRIC countries (especially China). Although the U.S. market is still by far the largest, most liquid and deep, Exchanges in Shanghai, Hong Kong, Mumbai and Sao Paulo are growing at a phenomenal rate. Naturally, more and more American funds are moving to these markets, and this will in the long run, sap the dynamism of American stocks.

Two investment styles that are more in tune with the nature of modern financial markets, and thus suited to babyboomers' shorter investment horizon are:

(1) To have an investment portfolio that is global, multi-asset class, with ability to go Long or Short. This strategy can be implemented using Exchange Traded Funds. See:http://randomthoughtsofanagingbabyboomer.blogspot.com/2009/08/high-frequency-algorithmic-trading-and.html and http://randomthoughtsofanagingbabyboomer.blogspot.com/2009/09/asian-economies-gradually-decoupling.html

(2) To have an automated trading system that is based on fundamentals, but with monthly portfolio rebalancing. This strategy can be implemented with ValuEngine's models and benchmark portfolios [ see http://www.valuengine.com/]. The automation forces you to do away with the emotional aspect of investing. The monthly re-balancing is of utmost importance. Since nobody can forecast the markets, the re-balancing [with automatic screening of stocks that match certain criteria] is a means of 'adapting' to the change in the character of the market-in this case the monthly change. Monthly re-balancing is just fine. Rebalancing too frequently means you will be swinging up and down with the market 'noise'. As you can see from the graph and table above which uses the Engine Rating model, this strategy does work. And you can see, that 2008 was an exceptional year even for ValuEngine-when it underperformed the S&P500 by 16 % [SP500:-36 %. ValuEngine: -52%.

Using ETFs To Profit From Secular Downtrend Of THE US$

2009-10-31T16:02:00.012+11:00

1. Inverse relationship of US$ with US Equities
2. Relative performance of stock markets by region and country

A feature of current financial markets inter-linkage is that when US equities go up the US dollar goes down, and vice versa. Ceteris Paribus (all other things being equal) a strong demand for the assets of a country [in this case its stocks) should result in its currency rising. However, the US$ and the US stock market's dominance and liquidity means that the US$ is still a 'safe haven' currency, and so the following unusual sequence of events occurs: US equities fall-----> world markets fall----->investors flee to safety of US$. This negative correlation can be seen in the image above, where I have drawn a Red line to represent the approximate rate of increase and decrease. The 1-year time frame of this image shows that the pattern is most clear and firm after the stock market lows of March 2009. Will this firm negative correlation between US$ and US equities remain? The gradual long-term decline of the US$ versus major world currencies is most likely, considering the trillion dollar debts that the US government has created. That the decline is so gradual is because there is still strong demand for US Treasuries. And this is so because the emerging countries that are buying Treasuries are not able to re-invest their export earnings elsewhere. Also, by buying US Treasuries, they are keeping their currencies down so as not to endanger the competitiveness of their exports. But a sudden drastic and significant drop in US equities e.g. due to disastrous corporate earnings or another round of credit crunch, could unhinge the present pattern of mild negative correlation. Then we could see a falling US stock market go together with a US$ falling at a steeper rate.

In a heads-you-win and tails-you-win situation like this, why not invest in the decline of the US$. Using the Exchange Traded Fund PowerShares DB US Dollar Bearish [ticker symbol UDN] which is the inverse of the US$ Index, we can Short the US$. Since other equity markets are still strongly correlated with the DJIA, a fall in US equities will also cause a corresponding decline in Asian and other emerging markets. But as explained in a previous post on this Blog, the positive correlation between US and other Asian markets [excluding Japan] will decrease i.e. Asia will become lesss dependent on the US. Therefore, while shorting the US$ via UDN, we can at the same time take a Long position on iShares MSCI All Asia Ex Japan [ETF ticker symbol AAXJ] during a pullback of the market. If you want a more concentrated (but more volatile) approach you can gain exposure to just China through the ETF of the Xinhua China 25 (ticker symbol FXI). Or, if you want to be a pioneer, get in early on Indonesia during a pullback via the ETF Market Vectors Indonesia Index [ticker symbol IDX]. Indonesia should rightly belong to the BRIC group of countries the term coined by Goldman Sachs to denote the economic giants of the 21st Century. i.e. BRIC should be BRIIC [Brazil, Russia, India, Indonesia, China]. With a population of 237 million half of whom are under the age of 30, a [now] stable government, and rapid urbanization leading to a culture of consumerism, Indonesia is the next Asian star. Image 2 above which shows IDX having a wider gap with AAXJ and FXI seems to reflect investors' perception of Indonesia.

Outperforming The Market With Bespoke ETFs

2009-10-04T23:41:00.011+11:00

1. YTD Performance of World Stock Markets By Region
2. YTD Performance of Various Asset Classes

3. YTD Performance of Selected Special ETFs

In the last few posts, we discovered that we could cut out the market's daily noise and ride on megatrends with ETFs: e.g. the relative decline of the US economy versus the rest of the world, the gradual erosion in dominance of the USD, and the risk of inflation because of the flood of debt issued by Central Banks to 'save' the world. In this post, we analyze the performance of the world's financial markets [stocks, bonds, commodities, currency, real estate] as at year to date (YTD). We have seen how ETFs can be created to represent and replicate any asset class. And now, we are about to learn that ETFs can be more than an general market/sector/asset classindex. There are some ETF's which are 'intelligent' ETFs, that is, their underlying assets are 'not dumb' components of an index, but pre-screened, differently weighted and created to achieve a defined objective. For example, PowerShares FTSE RAFI Asia Pacific ex-Japan Portfolio (ticker symbol:PAF) is based on the FTSE RAFI Asia Pacific ex-Japan Index . The Index is designed to track the performance of the largest equities of companies domiciled in the Asia Pacific region (excluding Japan), selected based on four fundamental measures of firm size: book value, income, sales and dividends. The 1,000 equities with the highest fundamental strength are weighted according to their fundamental scores. The fundamentally weighted portfolio is rebalanced and reconstituted annually.

Another example is DOO: WisdomTree International Dividend Top 100 Fund is a non-diversified fund. It seeks investment results that closely correspond to the price and yield performance, before fees and expenses, of the WisdomTree International Dividend Top 100 Index.

Let us use simple charts showing % change YTD, to analyze the performance of a wide range of ETFs. Image 1 shows the charts of :

SPY: SPDR ETF of the S&P500

ETF: the ETF of the MSCI EAFE Index [Europe, Australia, Far East}, oreffectively the rest of the world stock exchanges minus USA and Japan

AAXJ: ETF of stocks markets of All Asia Ex-Japan

FXI: ETF representing the Xinhua 25 China big caps.

The conclusion is that all the ETFS above exhibit a broad correlation with each other, but the degree of correlation with the US has been decreasing since March 2009. The second significant point is that All Asia Ex-Japan is outperforming China. All Asia is less volatile, and besides China, includes India, Taiwan, Korea, Singapore, Malaysia, Indonesia, and Thailand. Although a significant beneficiary of China's growth, each of these countries also have their own growth drivers, both domestically and trading amongst themselves.

Image 2 shows the performance of different asset classes:

SPY: SPDR SP500 (Stocks)

GSG: Goldman Sachs Commodities Index {Commodities}

IEF: iShares Barclays 7-10 year Treasuries (Bonds)

UUP: PowerShares DB US Dollar Bullish (Currencies)

RWX: SPDR DJ ETF of International Real Estate (Real Estate)

This time, the correlation is of lesser degree. Especially for US$(UUP) and Bonds (IEF). These two classes are also the poorest performers YTD, and their lack of volatility makes them unsuitable for shorter term investment objectives. Towards the end of Spetember, we see Commodities(GSG)making a sharper fall than Stocks (SPY), a potential sign of loss of confidence in the sustainability of the recovery, and perhaps taking into account that China's appetite for natural resources has been partially satiated. If you trace back to the lows of March 2009, you can see that the biggest gainer has been Real Estate (RWX)

Now, we come to an analysis of image 3: Selected special ETFs:

PAF: described above

DOO: described above

GLD: SPDR Gold Trust

DBO: PowerShares DB Oil Fund

JNK: SPDR Barclays, Capital High Yield (Junk Bonds.)

We are especially interested in the performance of PAF and DOO versus the rest. These pre-screened 'intelligent' ETFs represent the new breed of ETFs that attempt to go beyond dumb replication of a general market/sector or asset class index. I am esepcially impressed by the performace of PAF. The only problem is very low liquidity niche ETFs often result in very wide spreads of Buy and Sell. Still, if a small portion of your ETF portfolio is devoted to niche ETFs [say 10%] with a promising theme, it could do a lot to boost your overall bottom line.

Asian Economies Gradually Decoupling From US Economy

2009-09-12T00:17:00.007+10:00

Image 1: Chart of All Asia Ex-Japan, DJIA, and US Dollar Index ETFs
Image 2: A Real Portfolio of profits and losses from AAXJ, DOG and WIP

In the two most recent posts on this Blog, I wrote on the subject of using Exchange Traded Funds [ETF] to ride on world megatrends. The heuristics for a widening gap between All Asia Ex-Japan [as denoted by the iShares MSCI All-Asia Ex Japan ETF with Ticker Symbol AAXJ] and the US Economy as represented by an ETF of the DJIA such as DIA, or its reverse the DOG were expressed in mathematical notation for greater precision. [please scroll down to the previous post to refresh yourself on the heuristics.] Image 1 above illustrates clearly the decoupling of AAXJ from the DIA, as well as the gradual decline of the US Dollar as represented by the Powershares US Dollar Futures Index ETF with ticker symbol UUP. The chart shows percentage %, [not absolute price] and is semi-log scaled to have a more accurate visual representation

From this 1-year chart you can see that when world markets hit a low in March 09, All Asia Ex-Japan and the US were at the same level. As the months went by, through all the ups and downs, Asia pulled ahead. This is represented by the Yellow area growing in size. To put it succinctly, yes the US, the world's largest economy has bounced back from near-death. But Asia has staged an even more miraculous recovery. In the meantime the US Dollar is slowly but surely beginning a long-term decline, as the US share of world GDP falls in the years ahead, and the mighty US Dollar loses some of its prominence as a reserve currency and as the currency of world trade. This megatrend of Asia decoupling from the US Economy, and the US Dollar beginning a slow decline will continue.

In Image 2, I show a real portfolio, started a month or so ago which shows the profits as at today, from AAXJ and WIP, and a loss from betting that the DJIA would go down [via DOG the reverse of DIA]. But the losses from DOG are less than the gains from AAXJ. The other ETF in my portfolio is SPDR WIP. WIP is an interesting ETF. Its underlying assets comprise inflation protected sovereign bonds of developed countries excluding Japan and the US. Thus holdings include bonds of countries like France, Australia, U.K. Sweden, and Turkey, all yielding somewhere between 2.5 to 3.5 %. WIP would ceteris paribus rise in price when investors expect an inflationary situation. This is a likely scenario in the months ahead as Asian economies rebound strongly and the rush for raw materials resumes. And with most of the world commodities as well as shipping rates being quoted in US$, a declining US$ would add fuel to the fire as producer countries raise prices to counter the effects the declining US$.

Heuristics [Rules of Thumb] for ETF Portfolio Capital Allocation

2009-08-22T17:20:00.018+10:00

Irrelevant to this post, but a rare dish of fried vermicelli from the equally rare Putien dialect group of Chinese. Chock-full of goodies: dried seaweed, shrimp, fried peanuts, clams, yam, kale, cabbage, belly pork

1. Mathematical Expression of Heuristics for ETF Portfolio Capital Allocation

** Click on Image for Full-Size and Details

The decision-making process can never be truly quantified. Human decision-making usually involves the optimization of more than one [conflicting] variable i.e.it is multi-objective. In their book “Evolutionary Algorithms for Solving Multi-Objective Problems” [Springer, 2007] Coello Coello, Lamont and Van Veldhuizan noted that despite the wealth of techniques in evolutionary Algorithms for multi-objective optimization, from Particle Swarm and Tabu Search to Simulated Annealing and Ant Systems, ultimately the human has to make a decision based on heuristics, as to which part of the Pareto Front that he would like to be on, bearing in mind that it can be a multidimensional Pareto Front. Nevertheless, it is possible to semi-quantify the basis for our decision-making by expressing the heuristics in mathematical notation.
In the previous post of this Blog, I wrote about using ETFS to invest in world megatrends and to ignore the noise created by high speed algorithmic machine-generated trading.The allocation of capital within a portfolio of ETFs (Exchange Traded Funds) provides a good opportunity for the application of heuristics based on the investor’s view of market megatrends. An ETF consists of underlying assets of the same theme, even if they are swap-based ETFs where counter-parties guarantee the performance of a theme. Thus our heuristics begins with a statement of our ‘Belief’ on the theme, initially expressed in free-flow words, and gradually converted to more precise mathematical notation. For this post, we consider the allocation of capital in a two-ETF portfolio consisting of:
[1]ProShares Short Dow30 or DOG which is the inverse of the DJIA. The aim of DOG is to replicate as closely as possible, in an inverse direction, the changes in the DJIA. Thus DOG is the inverse of DIA the ETF that tracks the DJIA.
[2]AAJX is the iShares MSCI All Asia Excluding Japan ETF. AAXJ aims to replicate the performance of Asian markets excluding Japan. AAXJ would therefore track not only China and India but also countries such as Indonesia, Malaysia, Phillipines, Thailand, Indonesia and Singapore.
Belief Statement
While the U.S. is still the largest economy in the world, the post-Lehman Brothers world will see its share of the world economy gradually decrease while Asia ex-Japan's share of the world economy will increase at an increasing rate, due to the power of compounded growth . Stated alternatively, Asia ex-Japan will be less and less dependent on the U.S. for its economic well-being. With a total population of 2.8 billion, (China 1.3 billion, India 1.0 billion, the 10 ASEAN countries 560 million), there is much scope for growth from a low base fueled by domestic consumption. Other factors for optimism on Asia include the presence of China as the driving engine, and the young, hungry, intelligent and hard-working Asian population who have just tasted the joys [sic] of urbanization and, consumerism. Additionally, the region is rich in natural resources from oil and minerals and metals to agricultural commodities, timber and fish.
This being so, as a megatrend, ignoring short-term blips and noise,
An increase of X % in DIA will result in a greater than X % increase in AAXJ [1]
A decrease of A % in DIA will result in a less than A % decrease in AAXJ [2] Our Belief statement is expressed in mathematical notation in the image above.

Allocation of portfolio capital
Following from our argument, almost by default the initial % capital allocation for AAXJ must => 50 %
How much more than 50 % and what % to gradually increase the weight of AAXJ to can be refined by a standard 5- class trapezoidal Fuzzy Membership Function comprising: Very Pessimistic =0-20 %, Pessimistic=20-40%, Neutral= 40-60 %, Optimistic= 60-80 %, Very Optimistic= 80-100 %. Thus if you are optimistic on Asia, AAXJ should be between 60-80 % of your portfolio.
Afterthought
As time goes by, it may also be that a decrease in AAXJ could increasingly cause a decrease in DIA. As Asian bourses grow in their market capitalization and world funds increasingly invest in Shanghai, Hong Kong, Singapore etc a fall in Asian markets would have knock-on effects on the American bourses. Therefore during a market correction of AAXJ, DOG will make gains which can be used to fund the buying of more AAXJ.

High Frequency Algorithmic Trading And How It Affects You

2009-08-01T15:17:00.012+10:00

Two years ago, I made the acquaintance of a high-frequency algorithmic trader, and was invited to view his operations: In a small, darkened room a few doors away from our office in Princeton, NJ, three traders sat hunched over large multiple-screen computers. Colorful charts writhed across the screens like snakes, and numbers flashed and scrolled, updating in milliseconds. Just recently, such secretive operations were brought into the limelight with the arrest of the Goldman Sachs engineer who stole the codes for an algorithmic strategy. The public is now aware that at least 50 % of the trades on Exchanges across the world are algorithmically driven. That is, no humans are involved as trades are executed at the speed of 500-1000 trades per second. And large blocks of shares are broken up, bought or sold in split seconds to avoid detection. The machines send out probes and gather information to enable them to profit from statistical arbitrage i.e. differences in Bid and Ask of as small as 1/4 of a cent. It has now become an arms race and a war of the machines as rival houses' software (and hardware) battle each other for supremacy in speed and sophistication of strategy. (some traders like my Princeton friend actually move their operations to be nearer to the Exchange's servers for a small but crucial advantage in speed, even though we know electronic signals travel at the speed of light at 186000 miles per second). 10 years ago, Dr. Richard Olsen of the Forex house Oanda, gave me his seminal book on mining patterns from very large sets of high frequency data, and predicted the rapid advance of such techniques. His predictions have now come to past and the small investor had better know more about what the machines can do, and take counter-measures, or else he stands not a chance of surviving in the new high tech trading environment. Here then, are what machines have done to the trading environment, and what you the individual small investor can do to mitigate their effects:

1. High speed algorithmic trading [HSAT] causes the trading environment to be much more noisy. Being based not on fundamentals, but on Momentum and Volatility, they create illogical patterns which confuse the rational small investor who is trading on fundamentals. Even the day trader with his arsenal of technical analysis indicators like the RSI, MACD, or Stochastics will be confused by decisions and execution which are a hundred, or thousand times, faster; and which are gone by the time the real-time indicators move.

2. The liquidity which is generated by such trades gives a false picture. For small investors who look at volumes on the daily Close and make their decisions based on such volume, it is not accurate. Yesterday's liquidity may shrink to nothing today depending on the initial trades and how the machines weigh the probabilities for making a particular stock the flavor of the day, or even flavor of the hour or the minute.

3. Because of their increasing dominance, HSAT makes it possible for a market to act irrationally and out of sync with its fundamentals for a longer time. Thus what you see now, in the liquidty-driven global rally, in the price of crude oil futures, and in the Forex markets are in part due to the machines.

What you can do to avoid the machines:

Give up single stock, single currency, or single commodity picking. Invest in a diversified portfolio, playing on global megatrends. The machines will not be playing in this segment. It is too slow for them. They will be playing the daily ups and downs within the megatrends. By investing in the global megatrends, you are fundmanetally biased, thus avoiding the machines which play the noise.

What are these megatrends?

You can for example bet on a few things like: (1)the eventual decline of the US$, (2)the rise of inflation as stimulus measures create more debt for governments,(3) the almost inevitable rise of Asia and the Asian economies.

How do you play these megatrends?

ETFs [Exchange Traded Funds] are the instrument of choice for gaining exposure to global megatrends. With ETFs you can not only play whole sectors, and whole themes of investment. ETFs are also global, multi-asset class, and they can be Long or Short. And the costs of investing in ETFs is much lower than if you invested in mutual funds or a hedge fund. ETFs are traded exactly like a stock. ETFs tend to be less volatile, since they consist of a portfolio of underlying assets. You can also set aside say, 10 % of your ETF investment funds to actively trade short-term on the more liquid ETFs. Thus while the core 90 % is stable and move in a megatrend way, the actively traded 10 % can help to enhance total returns.

Examples of ETFs for the current economic environment

Ticker Symbol:

UDN (NYSE) Powershares DB US Dollar Index Bearish is one way to bet on the gradual decline of the US$.

WIP (NYSE) SPDR DB International Government Inflation Protected Bonds: Inflation-protected bonds of governments excluding the USA. You also get a yield from the Bonds besides capital gains.

AAXJ (NASDAQ): i-Shares MSCI All-Asia Ex-Japan Index. Gives you exposure to the 11 developing and emerging markets of Asia excluding Japan. A good way to ride on China and India with less volatility.

Links

A good way to start is to take a look at the Fulcrum fund portfolio of Singapore financial advisers New Independent http://www.ni.com.sg/

For trading ETFs in a shorter time frame, subscribe to ValuEngine's weekly ETF Report at http://www.valuengine.com/

For a platform which allows the trading of over 1000 ETFs across 22 Exchanges, try Danish Bank http://sg.saxobank.com/

Measuring Market Bearishness/Bullishness With A Self-Organizing Map

2009-06-28T22:20:00.012+10:00

1. SP500: Market Price and Fair Value
2. Longs & Shorts and the area occupied on the SOM

3. Statistics of Clusters

On a Self-Organizing Map [SOM], the distance between nodes is a measure of the degree of overall similarity in characteristics, the further, the less similar. [For an introduction to SOMs, see http://en.wikipedia.org/wiki/Self-organizing_map ] In this post we use this property of a SOM to estimate the overall Bullishness/Bearishness of the general market. Such an estimate would be very useful for a market neutral strategy where the main issue is the proportion of capital to allocate between the Longs and the Shorts. The past 3 months have seen an unexpected rally, and many are left wondering if this is sustainable. Image 1 shows the actual market value of the S&P versus its fair value. This chart is generated by ValuEngine Institutional software based on ValuEngines models. It shows that while Fair Value is rapidly falling [Green line], the Market Value is going up. We are now at the point where market value meets fair value [Green meets Blue line]. Any higher, and the market would be Overvalued.

But markets can remain in Overvalued territory for a long time if investor optimism, ample liquidity, or distortion by speculative derivative instruments fuel the fire. One way to get a different perspective of the issue is to look at the positioning of stocks in Long and Short portfolios generated for a market neutral strategy. I used the three different models of ValuEngine [Standard, Forecast, and Rating] to generate Long and Short 20 stock portfolios of each model. Thus, 60 Long and 60 Shorts were generated. Since each of the three ValuEngine models represent a different way to interpret the market, our 120 stocks do have some degree of diversity. [*ValuEngine Models: Standard is a true analytical model, Forecast additionally uses Monte Carlo simulation, and Rating additionally uses Classification algorithm]. Next, I tagged each stock of the portfolios in this manner: SL=standardlong, SS=standardshort, FL=forecastlong, FS=forecastshort,RL=ratinglong,RS=ratingshort. In this way, we can identify their location on the SOM when labels are imported. Our objective is to see how the six different categories SL,SS,FL,FS,RL,RS position themselves on the SOM.

Image 2 shows the S&P500 represented as a SOM, and the location of the portfolio stocks on it. One can immediately see that the 'L's [Longs] are positioned further part from each other than the 'S's [Shorts]. The 'S's are mainly concentrated in Cluster 3 [C3], the Yellow Cluster. The 'L's are spread out over the Blue cluster C1. An equal proprtion of L and S are in the Red cluster [C2], and in the small Green cluster [C4] are a few extreme FL stocks with poor fundamentals but high forecast prices.

The close grouping of the S stocks i.e. the short distance between each of them indicates that they have a high degree of similarity in terms of the variables of the three ValuEngine Models. On the other hand the sparse grouping of the L stocks i.e. the bigger distances between them, indicates that they are a less cohesive group.

It is possible to conclude from these observations that Bearishness is more strongly indicated in our SOM than Bullishness. How much more? Here is my heuristics (rule of thumb) for ratio of Longs to Shorts:

L=area of Longs cluster (Blue). S=area of Shorts cluster(Yellow). n=number of nodes in each cluster. Let Area be inversely correlated with strength whether Bullishness or Bearishness. In above, n in Yellow cluster is approximately equal to n in Blue cluster. So, it cancels out. S is approximately 1/2 of L, so it has twice the strength of L. Therefore ratio of Longs to Shorts is 1:2. That is, in a market neutral strategy, for every $1 allocated for Long, allocate $2 for Short.

Energy Stocks That Behave Like Healthcare Stocks

2009-06-20T16:03:00.006+10:00

1. Cluster Indicator: Heuristic Measure of Quality and Natural-ness of Clusters.
2. SP500 Stocks by Sector In Eight Natural Clusters

3. Statistics Of The Eight Natural Clusters Stocks in Cluster 6 [C6]

Sector Abbreviations: E=Energy, C=Capital Goods, B=BasicIndustries, D=ConsumerDurables, ND=ConsumerNonDurables, S=ConsumerServices, T-Technology, TP=Transportation, H=Healthcare, F=Finance,U=PublicUtilities

The aim of ordering data records according to their overall similarity in a map by Viscovery® SOMine is to form segments from the data that can be addressed with specific measurements.
Creating a segmentation from scratch is a two-step process. First, the data is clustered automatically by Viscovery® SOMine. This is done by different statistical methods, the SOM-Ward clustering, the Ward clustering or the SOM Single Linkage clustering and is described below. A segmentation based on SOM-Ward clustering is allocated automatically when a new model is computed. A Cluster Indicator is also provided by Viscovery which is a heuristic quality measure for each cluster count. It is intended as a means to help find an initial clustering.
The indicator is displayed in a chart, where the number of cluster appears along the horizontal axis. The vertical axis shows the indicator value for each cluster count, which can be interpreted as follows: If the indicator is high for a particular cluster count, the clustering may be viewed as “natural” for the map. If the indicator is low for a cluster count, that clustering is “artificial”. Usually, the peaks of the cluster indicator graph show interesting clusterings. In this exercise, the indicator is highest for 2-3 clusters. But since just 2-3 clusters will not be sufficiently detailed for the purpose of clustering SP500 stocks by their 11 Sectors, I chose 8 clusters, which is the second peak on the cluster indicator.

Image 2 shows the eight clusters, and Image 3, the statistics of each of these clusters-expressed in terms of standard deviation for each variable of the ValuEngine Model that was used for creating this Self-Organizing Map. In Image 3, we see that Cluster 6 [circled C6] is the most attractive cluster in terms of slight undervaluation, low volatility and low Beta. Attractive, that is, at a time when the market is considered to have overshot, the recent rally not being underpinned by fundamentals. Going back to Image 2, and looking at Cluster 6 [colored Pink], an interesting observation is that it contains many Healthcare [H] and Energy[6] stocks, two sectors which would seem to have opposing characteristics, with Healthcare usually regarded as a defensive sector, while Energy rides the growth story. When we identify the stocks that comprise C6, we find that it is a list almost synonymous with the well-known big caps. Pfizer and Microsoft are in this cluster, as well as Procter & Gamble, as well Cisco, JP Morgan, Goldman Sachs and so on. The Energy stocks are highlighted yellow and are: Exxon-Mobil, Conoco-Phillips, Noble Group, XTO, Transocean, Schlumberger and Chevron. If you like the Oil/Energy theme but are afraid of the volatilty of this Sector, this Self-Organizing Map points you to the 'safer' Energy stocks.

SP500: Top Eight Sharpe Ratio Stocks

2009-06-06T23:58:00.009+10:00

The Sharpe Ratio:

R=Rate of Return, Rf=Rate of Risk Free Return

E[R-Rf] is the expected value of excess of Return over Risk Free Return.

SQRT(var[R-Rf]) is the Standard Deviation of [R-Rf]

------------------------------------------------------------------------------------------------

Apple Inc, Monsanto, Gilead Sciences, Medco Health Solutions

* click on images for full-size and details

------------------------------------------------------------------------------------------------

McDonalds, Range Resources Corp, Express Scripts, Google

The Sharpe Ratio is widely used as an indicator of a stock's quality. Invented by William Forsyth Sharpe, in its most basic form it is Rate of Return- Rate of Risk Free Return/Standard Deviation of Return. The Rate of Risk Free Return was assumed to be a constant and thus the Sharpe Ratio was simply Rate of Return/Standard Deviation of Return. In 1994, Sharpe, acknowledging that Rate of Risk Free Return was a function of the economic times, revised the formula to take this into account. The formula is in image 1 at the top. If Standard Deviation is taken as an indication of Volatility, the Sharpe Ratio measures the reward/risk ratio. Since a specified rate of return is always accompanied by an expected level of risk for that rate of return, a higher quality stock is often defined as a stock that bears a lesser expected risk for a specified level of return. Now the next thing to ask is: What does the Sharpe Ratio mean in absolute terms. For example, is a Sharpe Ratio of 0.78 good, average or bad? Because of the Sharpe Ratio's formula, any increase in Rate of Return can be offset by an increase in the Standard Deviation- and vice-versa. A corollary of this is that even if the Rate of Return remains the same, the Sharpe Ratio can change if the Standard Deviation changes. Next, we have to consider the fact that the average Sharpe Ratio of a market will be different at different times. In general a Sharpe Ratio > 1 is satisfactory. During 2005-2006, strong stocks typically had a Sharpe Ratio of >1.5. It would be nice if ValuEngine can have a historical chart of the Sharpe Ratio of the S&P 500. I searched the Internet, and the only reference to historical Sharpe Ratio of the S&P 500, was an article by the Innovest Group that gave the following figures for the Sharpe Ratio of the S&P 500:

1997: 2.05 1998:1.84 1999:1.57 2000:-1.00 2001:-0.83.

Using ValuEngine Institutional, I have ranked the stocks in the SP500 by their Sharpe Ratio and listed the top 8 stocks by Sharpe Ratio above. I believe that at this juncture in time, stock selection using Sharpe Ratio as a criterion is useful. ValuEngine's calculation of Sharpe Ratio is over a period of 5 years which means starting from January 2005. About half of this 5 year period (from 2nd half of 2007 to the present) was a period of market turbulence and downtrend. Any stock surviving this period and still having a high Sharpe Ratio is worthy of consideration. From Apple to Google and Medco to Express Scripts, the stocks above have shown resilience through these extraodinary times. According to ValuEngine's calculation the market is currently about 18 % Undervalued [i.e.-18%]. Take a look at the Valuation % of the 8 stocks above. Monsanto [-18.16 %], Gilead Sciences [-17.66 %] and Medco [-8.20 %] are attractively valued in relation to the valuation of the general market. Range Resources, an independent oil Company has a high Sharpe Ratio but is Overvalued [70.57 %]. Note that because of current market condition, all our top 8 Sharpe Ratio stocks with Sharpe Ratio ranging from 0.95 to 0.71 ( a difference of 24 basis points) are have a Sharpe Ratio Rank of 99 in ValuEngine-which which means that all these stocks occupy the top 1 % of the S&P 500 in terms of Sharpe Ratio

Postscript: There is another modification of the Sharpe Ratio (I won't call it a refinement: The Sortino Ratio. In the numerator, a required or targeted Rate of Return is substracted from the Rate of Return. And in the denominator, the Standard Deviation is substituted by a targeted downside risk whose formula is rather complicated using the semivariance. (See http://en.wikipedia.org/wiki/Sortino_ratio ). The Sortino Ratio was devised by Brian M. Nom in 1986 in his study of post-modern Portfolio Theory.

Momentum: Which Sectors, Which Stocks?

2009-05-10T16:05:00.008+10:00

1. SP500 stocks in self-organized clusters
2. SP500 stocks Momentum Rank by Sector

3. Highest Momentum Stocks, their Valuation, and ValuEngine Rating

During a stock rally, be it a real rally or a dead cat bounce, momentum and volume become more important factors relative to factors such as valuation and Sharpe Ratio. But in order for investors not to be carried away especially in a puzzling market such as now, it is best to take a multi-pronged approach to stock selection, even if the investors' investment horizon is only short-term. Image 2 shows the momentum ranking of the Sp500 stocks presented on a Self-Organizing Map [SOM]. * On a scale of 1 to 100, with 100 being the most desirable, the momentum of the stocks are ranked, relative to each other. # On the SOM, the intensity scale at the bottom reflects the Momentum, with areas tending towards Red having higher Momentum. The 500 stocks have been classified by Sector with abbreviations as follows:

B=Basic Industries, C=Capital Goods, D=Consumer Durables, ND=Consumer Non-Durables, E=Energy, F= Finance, H=Healthcare, T=Technology, TP=Transportation, S=Consumer Services, U=Public Utilities.

Image 1 shows that the current rally is broad based and not confined to a few sectors. This is indicated by the fact that except for the Purple cluster, none of the 6 clusters have stocks belonging to one dominant sector. *The clusters were created based on the SOM's algorithms on measuring the Euclidean distance and regression between nodes, taking into account all the more than 40 variables of the ValuEngine model; resulting in a useful locally linear, globally non-linear model . This SOM was constructed in a finer resolution than normal, so as to have more clusters.

But, if we look at the stocks with the highest momentum ranking, we find that they are confined to a few sectors. With the SOM, we are able to pick specific nodes with the highest momentum, and the stocks occupying those nodes are shown in Image 3:

AMGN: Amgen: Healthcare

SGP: Schering Plough Corp: Healthcare

WYE: Wyeth: Healthcare

RHI: Robert Half Intn'l: Consumer Services

LEG: Legget & Platt:Consumer Services

FDO: Family Dollar Stores:Consumer Services

RSH: Radio Shack: Consumer Services

MYL:Mylan: Healthcare

FHN: First Horizon Financial Corp: Finance

WPI: Watson Pharmaceuticals

Firstly, the sectors of these highest momentum stocks is an indicator of the market's perception of the real economy. That it is not Finance of Energy that is the focus of the market's attention, but Healthcare and Consumer Services is significant. It reflects the market's optimism for a sustained bounce-back of the U.S. economy, based on the U.S. Consumer, i.e.based on domestic demand. Thus, the beaten-down stocks in Healthcare and Consumer Services are making a come back. The interest in Healthcare stocks also stems from recent M & A acitivity in this sector, as Drug giants search for new business model to revitalize themselves amid costly R &D, expring patents, generic drugs and mounting foreign competition.

Lastly, in looking at the selection of stocks pinpointed by the SOM, we should consider factors other than momentum. For example RHI, LEG and FHN are already highly overvalued, and this increases the probability that their momentum cannot be sustained. To further narrow down the list, after striking off LEG, RHI and FHN, we can look at 12-month return %. **The more undervalued stocks with lower 12-month return % and decent liquidity are the ones more likely to still have space to run in this rally, whether real or dead cat bounce. AMG and MYL are the ones with these simulataneous characteristics. As a final confirmation, note that they also have a 5-Engine Rating.

**Incidentally, this statement can be formulated quite neatly as a Fuzzy Rule using Fuzzy Memberhip Function. This Sugeno-type singleton Fuzzy Inference System(FIS) can be created using Matlab Fuzzy Toolbox's ANFIS (Adaptive Neuro-Fuzzy Inference System). But doing experiments would involve a lot of time which I haven't got. Hence, if anyone reading this would like to try, I can provide guidance especially for data preprocessing and heuristics, as well as talk to ValuEngine's Management on an Agreement to provide the data in return for some form of ownership and use of the results.

Best & Worst Stocks: 1-month Forecast Return %

2009-04-26T23:24:00.005+10:00

1. Correlation between 1-month forecast return % and 1-yr chance of gain %

2: Distribution of 1-month Forecast Return %

3:1-month Forecast return %: Best 20 stocks

4: 1-month Forecast Return %: Worst 20 stocks

Here's one way to gauge the state of the market. I used ValuEngine Institutional to generate 1-month Forecast Return % of the SP500 stocks. Then I plotted the distribution of all the forecasts. Image 2 above shows you that the shape of the distribution is skewed towards the left. It is far from being the symetrical bell-shaped curve also known as the Normal or Gaussian distribution. That is, there are more stocks at the lower end of forecast returns % than at the higher end. And that, is an indication that the market is still in a bearish mood. Now, Image 2 is another interesting finding. For stocks with a 1-month forecast return % of between -11 and 2 %, there is a significant correlation with its 1-year chance of gain %. Especially you can see that the density of data points around the 1 % to 2 % 1-month forecast return % has a very strong correlation with 1-year chance of gain %.

As for the 20 stocks with highest 1-m forecast return %, JC Penny tops the list with an incredible forecast return % of 27.99 %. But I would take this with a pinch of salt, as I think it is due to some data error. This is likely, because the 2nd highest forecast is Ford Motor with 3.03 %. But one conclusion from the list is that Energy and Health stocks dominate the list. What a combination! In my opinion, the Healthcare stocks are there because there seems to be a sort of shakeout in the sector, with mergers and acquisitions being in the news lately. But the strong Energy sector confirms my belief that inflation, higher interest rate, high energy and commodity prices will return to haunt us. The current strength of the US$ is an anomaly. With all that printing of money that the Obama administration is going to do, and with China and other countries more cautious about hoarding Treasuries, the US$ will begin a long term downward trend. And when the US$ begins to looks unattractive, the stock market will also lose steam as foreign investors begin to calculate the gains from stock versus the losses from exchange rate.

There is nothing much to say about the worst 20 stocks in 1-month forecast return %. Finance dominates this list. Except for Liz Clairborne and Harley Davidson. Note that the negative forecast return % of the worst 20 , is much higher than the positive forecast return %. Ranging from Genworth's -6.66 % to Capital One's jaw-drawing -24.54 %!