Last update: July 23, 2005. Contact: Mike Klein


I am not qualified, licensed, approved or authorized to make any investment recommendations or give any such advice. All data and information on this site is based on data from sources that are not under my control. I cannot guarantee the accuracy of any of the data, in fact I know of some significant data problems such as, but not limited to, incorrect stock splits that lead to radically incorrect prices. No investing decision should be taken based on the information on this site alone, and none of the information here is a recommendation to buy or sell any security!! Do the appropriate further analysis and due diligence before investing. I make no guarantees whatsoever about future returns using the methods or securities mentioned here.


I look for temporarily undervalued yet high-quality assets that can be purchased at a substantial discount from their estimated, or intrinsic, value. This definition encompasses many possible investment ideas and I am open to any and all. In fact, most, if not all, asset classes rotate in and out of favor over time. Those that are out of favor generally provide more fertile grounds for finding bargains, but no asset class should be ignored for possible ideas as bargains may appear in any asset class at any time.

To my knowledge, value investing is the only investing strategy that has been proven through many decades of real and fully documented results to provide excellent market-beating returns at low risk. Many other strategies exist but lack the basic requirements of proof: a very long track record (50 years or more if possible), a well-articulated and consistent strategy throughout, and returns with real money, regularly documented. In value investing's case, the basic tenets were set forth by Graham and Dodd in 1934 in the classic Security Analysis, adopted by Warren Buffett and numerous other extremely successful investors started in the 1950s after attending Graham's investing class at Columbia University. The finest evidence of this strategy's success is described in detail by Buffett in a speech in 1984, "The Superinvestors of Graham and Doddsville." Very few (but not zero) mutual funds continue this strategy, and all beat the market over long time periods. If anyone knows of other proven investing strategies meeting the requirements above, I would be very interested in them as well.

I actively invest my own funds and spend the majority of my "work" time on investing -- learning and doing. I earn my income by investing. I do not nor plan to sell advice, data, stock picks, or anything like that. I share methods and results here in the hope that:

Warren Buffett identified four categories of investments that he followed in the 1950s and 1960s to extremely high success, and the two that I have become comfortable with are described below (Underpriced Stalwarts and Under the Radar). The two others, Arbitrage and Control Situations, where he amasses enough stock to take control of the company and thereby directly manages its capital allocation, are unlikely to be categories that I will find myself doing.

Underpriced Stalwarts

Strong, quality companies with excellent long-term growth prospects that have had some temporary problem causing investor overreaction and unwarranted stock price drop. Buying these at sufficient discounts and holding until the market's efficiency process corrects the price should provide average annual returns in the 20-40% range. Recent examples: Merck, Coke, Nokia. The BMW Method Screen (below) is the easiest way to find candidates for this investment type.

Under the Radar

Companies too small for any larger investor and usually not covered by analysts, often with poor recent performance but a strong upswing in progress, but sometimes just great companies that are too small for large investors. These could be turnarounds or rapidly growing companies. For safety, these should be selling well below even a conservative valuation, as these smaller ships are easily buffeted around in the winds (and sometimes storms) of the markets. These take a lot more work to identify, analyze, and track; on the other hand their businesses are usually simple and easy to understand, and there is a minimum of financial shenanigans. Sometimes these are companies that have some kind of substantial value locked within them and needs a specific corporate action to unlock it. These are usually fairly explosive stocks when their improved financial performance is proved out or their latent value is unlocked, and may triple, quadruple, or more in a short time. They are, however, difficult to find. Finding candidates for this category requires scouring the results of a number of screens and is my major focus at this time. Buffett has said that he could, today, guarantee an annual return of 50% if he were managing a small portfolio ($10 million or less) by fishing around for these types of companies. They are, of course, hard to find and are not on anyone's list of great investments. You have to find these yourself, and when you find them, bet a lot of the portfolio on them.


Estimating a company's value is the hard part and there are many possible approaches. I have spent considerable time developing automated screens, and that is the focus of this web site. I have developed a number of screens that I run regularly and post results here and to The Motley Fool discussion boards. Some of these screens were developed with substantial input from other discussion board members. The screens are explained in more detail here and regular results are posted.


Being reasonably software proficient and with previous exposure to dealing with complex systems and data overload (engineering) I feel right at home developing screens for stocks, even with known imperfect and incomplete data. My goal with screens is to whittle down the universe of 8,000 or so public US companies, ADRs, ETFs, etc., with reported financial data to a small manageable number that can be analyzed in depth. To achieve any success, the screens need to have a good "hit rate"--stocks that pass the screen must have a decent number of truly good candidates among them. A "hit rate" of 20-40% is my target for decent candidates; but finding maybe 2 or 3 "great" candidates per year is the ultimate goal where the gain/risk tradeoff is such that I am willing to bet over 10% and up to maybe 30-40% of my total portfolio on such a choice. The screens do not attempt to find all companies that might be good investments, but focus on finding high-yielding small sets of companies.

All screens are based on fundamentals (even the BMW Method screen, see below). Every test in every screen is based on some fundamental aspect of running a solid business. While I do some limited back testing, and collect statistics on how the screens operate, I do not tune a screen to provide the best results on historical data, since the more finely tuned a screen is to historical data, the less likely it will find good future investments.

BMW Method

This screen was inspired by the investment strategy articulated by Motley Fool member BuildMWell (BMW). It relies on detecting buying and selling opportunities for a stock by comparing its price movements over very long time periods (20, 30 years or more) to an "Average CAGR" (Compound Annual Growth Rate) curve for that stock's price. Another way to look at the BMW Method is as an implementation of a "reversion to the mean" strategy, i.e. the price of an asset may deviate, sometimes substantially, from some "mean" (average), but over time tends to "revert" back to the mean. In the case of markets driven by emotion, the deviations away from and reversions to the mean tend to overshoot, providing excellent profit potential. See this chart of the Dow average from 1896 through 2004 for an illustration.

My BMW Method screen fits an exponential curve (constant compounded annual growth rate -- CAGR) to monthly close prices of stocks for the desired time interval. Each chart also includes some useful statistical indicators of the magnitude of the deviations from the Average CAGR curve. Back testing seems to indicate that substantial profits can be made by waiting for extremely undervalued situations, i.e. the stock's price has fallen very substantially below the Average CAGR curve. The BMW Method requires a lot of patience as reversions to the mean may not occur for years.

The BMW Method may not sound like fundamental analysis, but consider the following. When the price of a company's stock, or any other asset, is looked at over a very long time period that includes at least several significant cycles (booms and busts), the long term average growth rate of the asset's price should come reasonably close to showing its true intrinsic value. Those familiar with Graham might see this as the "weighing machine" in action. If the stock price falls substantially below the long term average growth rate and yet the company is not fundamentally and permanently a different one than before, then the stock is likely selling for less than its value. Some amount of analysis must always be done to ensure this.

Latest BMW Method screen results and a more detailed introduction to the Method.

"Hidden Gems" Type

This screen was inspired by the approach taken by Tom Gardner, cofounder of The Motley Fool, in his "Hidden Gems" investment newsletter (but has evolved significantly since then). Since a lot of detailed subjective analysis is done by Tom, there is no screen that can duplicate his work, which is why the screen is called "Hidden Gems Type." The screen identifies companies whose performance is rapidly improving (in fact, accelerating) along with having good balance sheets, asset management, and shareholder-friendly stock ownership characteristics. For this screen, company size is immaterial except for the tiniest companies.

Recently some work has been done to combine the results of the HG-Type Screen with a screen that does a rough and simplistic implementation of earnings quality checks as described in Hewitt Heiserman's book "It's Earnings That Count". Essentially these checks look for self-funding (a positive Free Cash Flow) and good return on invested capital when including a total cost of capital (equity and debt). The historical results of the combined screens have been very impressive, approaching Buffett's 50%/year.

Hidden Gems-Type screen results and an introduction to the screen.

Dividend Growth (currently inactive -- I have too much going on elsewhere)

Fundamentally, an investment is the purchase today of a stream of anticipated future cash flows. For example, bonds have a predetermined stream of future cash flows. With stocks that do not pay a dividend, the future cash flows are only promises that somehow the company will return its earnings in some form to stockholders over time, and the stock price rises with that anticipation (which can be irrational at times).

With dividend-paying stocks, a portion of earnings is regularly paid out to shareholders, and the price of the stock will depend on the amount of dividends, the confidence of their continued payment, and their growth rate. Research has shown that companies that pay regular and growing dividends are conservatively financed, and appear to make conservative yet good business decisions knowing that a great deal of their access to equity capital depends on continued growing dividend payments. In short, dividend growth companies seem to be stable and conservative investments not prone to poor capital allocation decisions (for the most part). In addition, dividends are now taxed at the same low rate as long term capital gains. Finally, dividend-paying stocks have not been popular investments for some time and they could be good contrarian investments. On the potentially negative side, increasing interest rates may make dividends less attractive and could force higher dividend payments or lower stock prices.

The Dividend Growth Screen is described in more detail here along with current results.

Back testing this screen since 2002 shows that overall returns are stable yet quite rewarding (averaging about 15-20%/year including dividends), and stock prices do not tend to be nearly as volatile as non-dividend paying companies. However, it must be kept in mind that this period was one of very low and stable interest rates. Rising rates may have a negative effect on dividend paying company stock prices just as for bonds.