1. How to use GuruFocus - Tutorials
  2. What Is in the GuruFocus Premium Membership?
  3. A DIY Guide on How to Invest Using Guru Strategies

GuruFocus Research: What worked in the market In the Decade of 2000-2009? Revisiting the Relationship between Stock Returns and Business Predictability

March 12, 2010

The first ten years in the 2000s is called “a decade from hell” by TIME magazine. It is called “the lost decade” by many investors. Don’t blame them, they did lose money. A $1 investment in the S&P 500 on Dec. 31, 1999, was worth roughly 90 cents at the end of 2009 — and that negative return includes dividend income.

However, in the lost decade, we have seen quite a few of our Gurus who did really well. Bruce Berkowitz’s Fairholme Fund averaged more than 13% a year, and Don Yacktman’s Yacktman Fund did almost 12% per annum. (See their common holdings) Therefore, there are ways to make money even in a lost decade. You just need to do it right.

In this article we would like revisit our studies on what worked in the stock market in the past decade. We have done extensive studies on this topic before. We have found that companies with consistent revenue and earning growth outperformed the market average by great margins. Among the predictable companies, the stocks of those appeared undervalued did even better. From the study we have developed the concept of Predictability Rank of Companies and the value screens of Undervalued Predictable Companies and Buffett-Munger Screener. You can see these previous studies in these articles:

· What worked in the market from 1998-2008? Part I: Predictability Rank

· What worked in the market from 1998-2008? Part II: Role of Valuations, Under-Valued Predictable Companies and Buffett-Munger Screener

· What worked in the market from 1998-2008? Intrinsic Value, Discounted Cash Flow and Margin of Safety

In this research we study the performances of prices of 3543 stocks that have been continuously traded from Jan. 3 of 2000 to Dec. 31 of 2009, of which we have the complete 10-year financial history in our database. We will try to find the correlation of the stocks of these companies with their business performances.

As we discussed before, this study have survival bias, as we cannot study the companies that have been de-listed during the 10 years. But we believe that the survival bias favors more toward companies with poor business performances. Also the study does not include dividends in the investment returns. We will compare the results with the price changes of the S&P500, which does not include dividends, either.

Correlation of Business Predictability with Stock Performances

In this section we like to discuss the effects of the predictability of a business on the company’s long term stock performances. We have ranked the predictabilities of companies based on the consistency of the revenue and earning growth over the past 10 years. Companies need to stay profitable through all these 10 years to qualify for Predictability Rank of 2 and above. The chart below shows the overview of the stock performances of companies with different business Predictability Ranks over the past decade.

Predictability Rank All 5-Star 4-Star 3-Star 2-Star 1-Star
Total Number of stocks 3543 74 141 192 135 3001
% of Stocks in Loss 48% 15% 10% 17% 20% 54%
% in Loss > 50% 30% 0% 4% 4% 6% 35%
% in Loss > 90% 13% 0% 0% 0% 0% 16%
Max Loss/Min Gain -100% -48% -86% -77% -90% -100%
Median gain 5% 192% 193% 129% 112% -11%
Annualized Median gain 0.5% 11.3% 11.3% 8.6% 7.8% -1.2%

Predictable Companies

Among all the 3543 companies studied, only 542 or 15% of them are ranked with the Business Predictability of 2-Star and above. 215 companies or 6% are ranked 4-Star and above. You can see the List of Predictable Companies here. The ranks are updated weekly.

Probabilities of Losing Money

If we look at all the 3543 companies, almost half of them (48%) are still in loss if investors bought them on the first trading day of 2000, and hold through the last day of 2009, a complete 10 years of holding period. 30% of them lost more than 50% of per share price, and 13% of them lost more than 90%.

However, if we look at companies that are ranked 2-Star or above, the probability of losing money is much smaller, less than 20%. This is a dramatic improvement compared with the average of 48%. The percentage of losing more than 50% is reduced to less than 4%, none of them has lost more than 90%.

For the 3001 One-Star ranked companies, 54% of them are still in loss after a holding period of 10 years. These companies have experienced at least one fiscal year of loss or had long term revenue and/or earning declines.

Therefore, we can clearly see that by investing in predictable companies, investors have a much better chance of avoiding permanent losses.

Investment Returns

The median gain of the stocks of the 3543 companies is about 0.5% a year. Over the same period, S&P500 lost about 2.7% of year. The difference can be caused by the different samples of companies, and the survival bias, which tends to shift the median gain to the higher side.

However, if we look at the companies with Business Predictability Rank of 2-Star or above, the picture is much brighter. As a group they have an annualized median gain of 9.4%, almost 9% better than the average.

The stocks of the higher ranked companies did even better. The stocks of 4-Star and above companies have a median gain of 11.3% per annum, more than 10% better than the average.

We can see that by investing in highly predictable companies, over long term, investors can avoid potential loss and achieve much better returns even without considering the market valuations of the stocks. If one invests in companies with consistent revenue and earning growth, the growth of the business will compensate the possible overvaluations that investors may have paid. The health of business helps investors to avoid permanent losses caused by the failures of the businesses. From this we come to GuruFocus’ first rule of investing:

Rule #1: Buy predictable companies only.

Roles of Initial Valuations of Stocks

In this section we would like to discuss the roles of initial stock valuations in the performances of stocks. The initial valuations of the stocks are obviously important.

As discussed before, we use a very simple indicator to determine if a stock is undervalued, fair-valued, or overvalued. This indicator is the ratio of the P/E ratio of the stocks on Jan. 3, 2000 over average of the EBITDA growth rate over the past 10 years. This indicator values the growth of the business, and is used just for comparison purposes.

Predictability Rank 5-Star 4-Star
Undervalued Fair-Valued Over-Valued Undervalued Fair-Valued Over-Valued
Total Number of stocks 22 21 31 45 40 56
% of Stocks in Loss 0% 10% 29% 0% 5% 21%
% in Loss > 50% 0.0% 0.0% 0.0% 0.0% 2.5% 7.1%
% in Loss > 90% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
Max Loss/Min Gain 39.0% -39.0% -48.0% 40.0% -52.0% -86.0%
Median gain 592.5% 227.0% 52.0% 402.0% 221.0% 77.0%
Annualized Median gain 21.4% 12.6% 4.3% 17.5% 12.4% 5.9%

Top Ranked Predictable Companies

The initial valuations have great impact on the performances of the stocks, as expected by us value investors. Among the 74 five-star ranked companies, 22 are undervalued. None of these 22 companies suffered loss in the “lost decade”, the smallest cumulative gain is 39%. Their median gain is an astonishing 21.4% per annum, and 592.5% cumulatively. The overvalued group did much poorer. 9 companies suffered loss, the median gain is only 4.3% per annum, and 52% cumulatively in the 10y years.

Similar results are observed for stocks of all rankings. The annualized median gains of the stocks are shown in the chart below.


The chart above clearly shows that the undervalued group always outperforms the others no matter what rank the company gets. The area that works the best is Undervalued Predictable Companies, the companies that have most consistent revenue and earning growths, and appear undervalued at the time of the purchases. From this we come to our second rule:

Rule #2: Buy predictable companies that are undervalued.

Performances of Related Model Portfolios

Since our research on the predictable companies in 2008, we have created two value screens for our users to find undervalued predictable companies. One is called Undervalued Predictable Companies: these are Predictable Companies that are undervalued as measured by discounted cash flow (DCF) model. The other is called Buffett-Munger Screener: these are the companies with sustainable profit margin and little debt, and at fair prices as defined by Warren Buffett and Charlie Munger. Two model portfolios are also created to monitor the performances of these predictable companies.

The performances of the two model portfolios are shown in the charts below:

Undervalued Predictable Companies:


Buffett-Munger Screener:


Since incepted on Dec. 31, 2008, both of the portfolios have outperformed the S&P500. But the portfolio of Undervalued Predictable Companies did much better than the other one. In less than one-year and 3 months, the portfolio gained more than 60%. It gained more than 130% since the low of March 2009. The portfolio was rebalanced on Dec. 31, 2009. It has gained more than 4.7% as the market zigzags with a gain of around 1.5% year to date. You can see the complete portfolio here: http://www.gurufocus.com/model_portfolio.php?mp=smallcap

Comparison of This Work with Ben Graham Value Approach

The idea of buying undervalued predictable companies was inspired by Warren Buffett, who looks for companies with proven and predictable earnings in his investing. This approach focuses more on the quality of the business, and hopes to buy them at reasonable or good prices.

In the original idea of Ben Graham approach of value investing, cheapness was the key. In order to avoid unexpected losses, Ben Graham used to buy a basket of cheap companies. The list of stocks that qualifies Ben Graham’s net-net is shown in GuruFocus Ben Graham Net Current Asset Screen. GuruFocus also publishes a monthly newsletter which is dedicated to this Ben Graham net current asset bargain approach.

The stocks generated from Ben Graham net-nets screener did very well. But as market recovers, the list in GuruFocus Ben Graham Net Current Asset Screen is becoming shorter. At current market valuations, we believe that buying Undervalued Predictable Companies can be more rewarding. This is the current list of Undervalued Predictable Companies.

GuruFocus Ben Graham Net Current Asset Screen and Undervalued Predictable Companies are only some of the features GuruFocus provides for its Premium Members. If you are not a GuruFocus Premium Member, we invite you for a 7-day Free Trial.

About the author:

GuruFocus - Stock Picks and Market Insight of Gurus

Rating: 2.9/5 (49 votes)


Dr. Paul Price
Dr. Paul Price premium member - 6 years ago

One obvious flaw in the reasoning given in the article...

Since the author notes that companies needed to "stay profitable for the entire ten years" it has eliminated from the data all stocks that would have qualified initially but dropped off the list due to at least one poor year later.

Buyers of 'predictable' companies today can't know whether these same companies will qualify for the same list ten years from now. So buying the last decades 'predictable' companies doesn't mean you own those who would be counted towards the backward looking data presented.

The only true test of the theory presented would have been to show what happened to a static list of companies that qualified initially and then following how they did regardless of whether they fell off the list later.
Aviramy - 6 years ago    Report SPAM
Of course predictable profit growth companies in the last decade, were good investments - in the last decade. Their profit growth predictability IS EXACTLY THE REASON they were good investments.

The question is will they be good investments in the NEXT decade.

In other words, you should grade the companies according to their profit growth predictability in the decade PREVIOUS to the last decade (1990-1999), then test the highly graded companies' stocks performance in the last decade.
Max7777 premium member - 6 years ago
Aviramy has a very good point above: To see if your theory works, Please grade the companies according to their profit growth predictability in the decade PREVIOUS to the last decade (1990-1999), then test the highly graded companies' stocks performance vs the results you are presenting to us for 2000-2010.

If it works then you have serious data that proves your theory. If it does not work then Predictability is still valuable, but this is not the way to it..

Buffett does value predictability of earnings, but in his case since he has very low turnover it makes a lot of sense. If you are going to stay with a stock for decades then Predictability is becomes paramount. On the over hand, Paulson or Soros would not pick that as an important factor, since they try to look for stocks with the highest potential today, irrelevant of past performance (I am thinking C, which has probably terrible predictability over past decade but huge potential given its fall)
Brahmsian - 6 years ago    Report SPAM
Have you done any tests on the value of insider ownership?

"Insider ownership" in my usage means the percentage of the common stock owned by people who qualify by SEC definition as "insiders", rather than the recent purchases and sales by those people on the open market.

Please leave your comment:

GuruFocus Mobile App

User Generated Screeners

pbarker46PB & PS Ideas
dwilsherQuality d/d 10 y growth
yukondeltagBig Short
tyde_dicksonLow PE / PB
Get WordPress Plugins for easy affiliate links on Stock Tickers and Guru Names | Earn affiliate commissions by embedding GuruFocus Charts
GuruFocus Affiliate Program: Earn up to $400 per referral. ( Learn More)