Saturday, April 2, 2011

Valuing a stock

I have mentioned 'buying shares at the right price' on many occasions on this blog. Benjamin Graham which was Warren Buffett's mentor and author of the stock market classic: "The Intelligent Investor", often used parameters such as P/E and Book value to determine whether a stock was undervalued or not. Reading the 'Intelligent Investor' is a must for every value investor. Although written in the fifties, it is still available in the investment section of many Chapters stores and other bookstores.

Dreman also defines stocks based on these variables, but rather than determining the value of a particular stock he compares it to other stocks in the stock market or in the same stock market sector. His thesis is that stocks are valued efficiently in the market except those that are hot and those that are in the dog house. Hot stocks are overvalued and tend to underperform the market. Stocks in the dog house are undervalued and ignored by investors and analysts alike – nobody expects anything from those stocks but bad news. Stocks that are in-between basically perform like the overall market.

According to David Dreman's research, stocks in the dog house systematically outperform the stock market. Not every one of them, but as a group they do. Some may go completely broke, but a basket comprising several stocks or all stocks in that group will typically outperform the market.

Analysts, like many experts in their field, tend to be poor forecasters; they may forecast human behaviour, the economy or the performance of companies like stock analysts do. The more data is accessible by an expert the more confident experts tend to be and ironically, the less reliable their estimates. It is not only overconfidence to leads usually to overly optimistic forecasts; the errors are also caused by limitations of the human brain. In particular, humans are linear thinkers, we are good in extrapolating the effects of one variable; but we are very poor 'configural' thinkers, i.e. figuring out the effects of many variables combined. When predicting a company's annual profit and sales performance, analysts have to take into account macro-economic factors such as the rate of inflation, consumer confidence and micro-economic data such as demand and supply data for the multiple products a company may produce, availability of raw materials, etcetera and etcetera. The integration of all these variables requires 'configural' thinking' and us humans are not very good at that. Unfortunately, analysts are human too and thus their forecasts are often off… by a wide margin!

Dreman quotes research of earnings forecast errors by analysts between 1973 and 1996 for over 1500 companies that traded on the American stock exchanges. Remember how stock prices literally collapse when earnings are 3 or 5% below analyst consensus? You would think those guys are pretty good if such small deviations caused such large changes in stock valuations. Well think again, the research showed that the average analyst makes an error of 44% in his/her earnings forecast and the median error was… 42%. Wow!

Dreman splits the stock market into quintiles or fifths: the top quintile or 20% of companies that are most expensive, the middle 60% and the bottom quintile. He determines the quitiles for each of the four standard measures of value:
  1. the ratio of earnings per share divided by the price per share is the price earnings ratio. It was also one of the most frequently used variables by Benjamin Graham to determine how expensive a stock is.
  2. the  the share price/cash flow per share ratio for each company in his extensive database of U.S. public companies.
  3. The net asset value of a company is the third measurement to determine how expensive a company is. This is expressed as the book value (assets minus debt) per share which Dreman divides by the share's price to determine a company's bookvalue over price ratio.
  4. Finally he used the dividend yield of each stock in his database
Dreman found that the bottom quintile as determined by these ratios often outperformed the market, but the results became even more striking when Dreman took earnings surprises into account. He found that a negative earnings surprise with the lower quintile companies did do little to the stock price – these were poor stocks to start out with. What else could one expect? The middle quintiles underperformed the markets by 5% in the following year and the top quintile – the hot stocks – underperformed the average market by a whopping 8.9%.! "What would be the effect of a positive earnings surprise?" you might ask.

Glad you did. Dreman's used the Compustat database containing data from 1500 publicly traded U.S. companies from 1973 until 1996. Every year he determined the top quintile, the bottom quintile and the middle quintiles for each measurement type (P/E P/CSF/ BV/P, Dividend yield). A positive earnings surprise for the top quintile was what investors expected, so a year after the surprise their stock price was only 1.5% better than that of the overall market. The middle quintile companies with a positive earnings surprise outperformed the market by 3 percentage points and the lowest quintile outperformed the market by an amazing 8.1%! Those gains were not made right away, most gains were made in the three quarters following the quarter during which the surprise was announced. So plenty of time for investors to react!

According to David Dreman, hot stocks consistently underperform. The middle quintiles are performing as the average market does. The contrarian investor who invests in the bottom quintile companies with positive earnings surprises consistently does better than the average market… by 8.1%. Since exact fundamental analysis is so difficult to achieve, Dreman does not feel that traditional value investors fare as well as in the days of Benjamin Graham.

Warren Buffett is one of the few that seems to be able to predict future corporate performance correctly or better correctly enough to spot value investments a la Benjamin Graham (with some adjustments). But for most of us earthlings, buying bottom quintile companies with positive earnings surprises may prove a successful way of investing as well.

In our next posting we will show you another method of valuing stocks.

No comments:

Post a Comment