Contrarian Investment Strategies

by David Dreman

  Not unexpectedly, Congress’s “buyer’s remorse” at the actions it authorized would rival that of any investor caught up in a very nasty market surprise. Patriotic emotions had again swept aside more reasoned considerations—yes, shades of Affect—and had momentous consequences. As history bears out, warning signs may not be seen clearly during the moment, but nasty political surprises can almost predictably lead to overreactions by officeholders and the general public alike.

  In the markets, investors also react to surprise in a fairly predictable way. In the 2000s alone, the collapse of Enron in 2000–2001, followed by that of WorldCom in 2002 and then Bernie Madoff in 2008, soured millions on stocks. The distrust has been heightened by the flash crash of May 6, 2010, in which the Dow Jones Industrials dropped more than 600 points in minutes and the intraday swing was over 1,000 points, the second largest in market history. At the time of this writing the SEC and CFTC still have not taken appropriate action to prevent another flash crash from occurring; their inaction has further disillusioned investors and resulted in a flood of cash pouring out of stocks into Treasury bonds at almost zero yields.

  That’s what surprises do on Wall Street. They continually change our outlook on owning bonds or stocks or gold. Within the stock market they do the same for companies’ earnings outlooks—and thus inevitably their stock prices. The market is always adjusting to surprises or anticipating them or discounting them: the standard fare of investment news you hear or read every day. The key players, of course, in initiating a market surprise are the analysts we met in the previous chapter, the men and women who predict what will happen barring any surprise. Ironically, here we have a self-fulfilling prophecy. It’s their pinpoint forecasts, which are rarely met, that create the majority of earnings surprises on stocks.

  After covering the many aspects and systematic patterns of surprise in this chapter’s first section, the last part of the chapter will introduce a relatively new field of science dubbed “neuroeconomics,” which breaks down some traditional boundaries. This field combines research from neuroscience, psychology, and economics to analyze how people make decisions. It catalogs the inner activity of the brain using advanced monitoring technology to spot what actually happens when we evaluate decisions, categorize risks and rewards, and interact with other people. In short, one of its aspects is an objective, biologically based extension and confirmation of important findings in cognitive psychology. The results are often surprising!

  Surprise and the Market

  Time to think twice about that stock purchase? Possibly. But far more important, maybe it’s time to reconsider the whole mechanism of surprises, not simply from an anecdotal perspective but from a solid statistical basis. Yes, negative surprises have enormous influence on us, both as citizens and as investors. But there is the other and far more upbeat side of the coin: positive surprises put spring in our step and add heft to our portfolios. In this chapter we will look at earnings surprises. Although, as we just saw, they occur frequently, they don’t necessarily have to be anxiety-producing events. Quite the opposite; if you know what you’re looking for, they can bring up some very cheery days. We’ll see that earnings surprises have a consistent and predictable effect on stock prices that you can use to your advantage.

  Specifically, they have a dramatically different impact on stocks that people like as contrasted to those people don’t like. Importantly, the new and rapidly growing field of neuroeconomics uses brain scans to study some of our major emotional reactions to surprise, providing a solid explanation of what we shall see. And understanding the nature of surprises provides a high-probability method of beating the market.

  Surprise: No Price Is Too High for Growth

  At times, no price seems too high for aggressive growth stocks or IPOs. Investors repeatedly pay through the nose and just as repeatedly get stung. Nevertheless, as we’ve seen, strong psychological forces compel people to buy sizzling issues and then prevent them from analyzing where they went wrong.

  The pattern is, not unexpectedly, repeated for larger companies. Investors believe they can forecast the prospects for both exciting and unexciting stocks well into the future. They have high hopes for “best” stocks and high confidence that their expectations will be met. Similarly, they have low expectations for stocks that appear to have lackluster or poor prospects but again high confidence that their estimates will be dead-on. The previous chapter showed just how dead-on such forecasts actually are.

  Companies with the best prospects, fastest growth rates, and most exciting concepts normally trade at a high price relative to earnings (P/E), high price to cash flow (P/CF), and high price to book value (P/BV), and invariably provide low or no dividend yields. Conversely, stocks with poor outlooks trade at low price to earnings, price to cash flow, or price to book value and usually have higher dividend yields.*52

  Often, the disparity between what investors will pay for a favored stock and one badly out of favor is immense, as chapters 1 and 2 showed. Investors, for example, happily shelled out 100 times more for each dollar earned by the Internet wunderkind eToys.com, shortly before it went to dot-com heaven, than for a dollar earned by boring old JPMorgan Chase. Investors pay such price differentials because of their confidence in their ability to pinpoint the future. Let’s look at what happens when—surprise!—their forecasts miss the mark.

  Surprise: What Our Studies Revealed

  To find out how stocks react when analysts err, I did a number of studies in collaboration with Drs. Eric Lufkin, Vladimira Ilieva, Nelson Woodard, Mitchell Stern, and Michael Berry over time, the latest one for the thirty-eight years ending with 2010.1 To be consistent, we worked with exactly the same analysts’ consensus forecasts that were used to calculate analysts’ errors in the preceding chapter.

  We wanted to measure a number of factors important to investors. First, what do analysts’ forecasting errors do to stock prices? Second—and as important—do earnings surprises have the same impact on favored as on unfavored stocks? Stocks trading in outer space are there because of analysts’ confidence in their future—possibly mixed with a wee dose of overoptimism. Did they react the same way to earnings surprises as to stocks that are in the investor doghouse? Third, we wished to examine just how accurate investors expect analysts’ forecasts to be. To help resolve this, we measured how even tiny surprises affect a stock’s price by considering any amount over one penny a share a surprise.

  To answer the three questions, we analyzed the stocks strictly according to how exciting or dull investors believed their prospects were, using three different value measures: the price-to-earnings, price-to-cash-flow, and price-to-book-value ratios. The higher the three ratios, the more enticing the stock is to an investor and the more he or she is willing to pay. Conversely, the lower the three ratios, the more unpopular the stock. We divided the stocks in all of the quarters in our 1973–2010 study into three groups strictly by how they ranked by each of these three value measures. The 20 percent of stocks that had the highest P/Es, for example, were placed in the top P/E group (called a quintile), the next 60 percent in the middle group, and the lowest 20 percent in the bottom quintile. We did this for all three value measures. The portfolios were reassembled on this basis for every quarter in the study. We then calculated the effect of surprises on each group of stocks beginning in the first quarter of 1973 and ending in the fourth quarter of 2010, a thirty-eight-year period in all.2

  The study used the 1,500 largest companies in the Compustat database with fiscal years ending in March, June, September, and December.3 Approximately 750 to 1,000 large companies were used in each of the 152 quarters of the study.

  Surprise: What the Historical Record Shows

  Next we set a yardstick to gauge the result of market surprises. The surprises are measured against the analyst consensus forecast, the average estimate of the group of analysts following each stock, as described in the preceding chapter. A surprise is measured against actual earnings, so it do
esn’t matter whether the earnings are up or down. If a company reports a loss of 80 cents a share, as an example, and the Street expected a loss of one dollar a share, then it would be considered a positive surprise of 20 cents divided by the 80 cents reported, or 25 percent.

  Do surprises affect favored and unfavored stocks in the same way? To find out, we’ll look at all surprises, the combined effect of positive and negative surprises on the favorite stocks—the highest quintile—the middle quintiles, and the out-of-favor stocks—the lowest quintile. The results are shown in Figures 9-1, 9-2, and 9-3. In each case, the 20 percent of stocks most out of favor by one of the three value measures—price to earnings (9-1), price to cash flow (9-2), and price to book value (9-3)—are depicted by the dark bar, the 60 percent of stocks in the middle groups by gray, and the most favored 20% by white.4 The charts calculate the return above or below the market’s for each of the 152 quarters of the study.

  The market return*53 is set at 0 in the center of the vertical axis of the chart. The surprise return must be added to the market return in each period to get the total return. If a bar shows a 1 percent positive return on the left, for example, it means that the stock did 1 percent better than the market over the average three-month return of the study. The market provided a 3.5 percent return on average, quarterly, so the total quarterly return would be 4.5 percent (the market return plus 1 percent or 4.5 percent for the average quarter throughout the study). If the stock did 1 percent better for the full year, it would return 14.8 percent, as shown in the chart, plus an additional 1 percent, or 15.8 percent for the full year on average, annually, throughout the entire study. If it was a 3 percent negative quarterly return, it did 3 percent worse than the market. This type of chart lets you easily appraise how surprise affects each group of stocks.

  The figures show the effect of an earnings surprise, measured in the quarter in which earnings were actually reported—which is always the quarter following that in which the earnings surprise took place. We will call this latter quarter the “surprise quarter.” The left-hand group of bars shows the effect of the surprise in the quarter it was announced, while the right-hand group represents the effect after one year.

  A glance at each of the charts shows that all surprises (positive and negative combined) helped unpopular stocks and hurt popular ones. Looking first at price-earnings ratios in Figure 9-1, we see that all surprises to unpopular stocks returned 1.2 percent above the market’s return in the surprise quarter over the life of the study, about a third more than the market.

  What’s more, beyond the surprise quarter, the beneficial or lethal effect of a surprise increased for the full year. All surprises for out-of-favor stocks (the low-P/E group) returned an average 3.8 percent above market each year. This is 26 percent annually above the market return over the life of the study. It is also triple their outperformance in the surprise quarter itself. By contrast, as the figure shows, favorite stocks, in this case the 20 percent of stocks with the highest P/E multiples, had a return almost 1 percent below the market in the quarter, which widened to 3.2 percent annually on average for the entire study, some 25 percent under the market return.

  Surprise, as we might expect, did not seem to have much effect on the 60 percent of stocks that make up the middle grouping. These stocks are not normally over- or undervalued much. As Figure 9-1 shows, the stocks were down by less than one-third of 1 percent in the surprise quarter. A year after the surprise there was a small negative (–1.2 percent) effect.

  However, the difference in the effects of all surprises on “best” and “worst” stocks was large and increased over time. “Worst” stocks outperformed “best” stocks by 2.1 percent in the surprise quarter, then steadily rose to 7.0 percent (or approximately 50 percent of the market return) in each year of the study.

  To summarize, Figure 9-1 reveals that earnings surprises did not affect the returns of the various P/E groups the same way. Surprise significantly benefits unfavored low-P/E stocks and works against the high-P/E group, while it has a nominal effect on stocks in the middle group. Is there any difference in how surprise affects stocks ranked by the other value measures?

  Surprise: Toute la Différence

  Figure 9-2 looks at the effects of all surprises on stocks measured by price to cash flow. The chart is nearly identical to Figure 9-1 for the surprise quarter and the full year. The lowest-price-to-cash-flow group again strongly outperformed the market in both cases. Similarly, the favorite stocks, the 20 percent of highest-price-to-cash-flow issues, significantly underperformed the average in both periods, while the middle group was almost unaffected by surprises. By this value measurement, surprises in analysts’ forecasts once again work powerfully in favor of the most unwanted group and against the most highly regarded stocks.

  Figure 9-3 demonstrates the effects of all surprises measured by price to book value. Remember, the higher the price-to-book-value ratio, the more popular a stock is, and the lower this ratio, the less popular the stock is. Again, the results are similar. Favorite stocks underperformed in the surprise quarter (–0.7 percent) and did even worse for the full year (–2.8 percent). Unpopular stocks outperformed in the quarter (+0.7 percent) and took off nicely over the full year (+2.9 percent above market). Again, the middle 60 percent of stocks were far less affected by earnings surprises.

  What is remarkable is not only that out-of-favor stocks outperformed by all three measures but how similar the performance was regardless of the value measure we chose. We thus begin to see a path to making money in the stock market. Earnings surprises, whether positive or negative, affect favored and out-of-favor stocks very differently. Surprise consistently results in above-average performance for out-of-favor stocks and below-average performance for favored stocks. Has the lightbulb gone on? We can find illumination in this Psychological Guideline.

  PSYCHOLOGICAL GUIDELINE 18: Earnings surprises help the performance of out-of-favor stocks, while affecting the returns of favorites negatively. The difference in returns is significant. To enhance portfolio performance, you should take advantage of the high rate of analysts’ forecast error by selecting out-of-favor stocks.

  Conversely, buying favorites will cost you money. How much money? A look at the magnitude of the surprise effect is sobering, as I’ll demonstrate shortly.

  The Effects of Positive Surprises

  In the preceding section we looked at all surprises, both positive and negative combined, for three of the major fundamental yardsticks, the price-to-earnings, price-to-cash-flow, and price-to-book-value ratios. In this section we will separate the surprises and look first at how positive surprises—higher-than-expected earnings—effect each of the above fundamentals.

  Examine Figure 9-4. It shows the effects of positive surprises on stocks in our high, low, and middle groupings, by P/E. As you can see, positive surprises galvanize the lowest 20 percent of stocks. In the surprise quarter, they outperform the market average by 2.6 percent, or 75 percent. For the full year, the lowest P/E quintile charges ahead of the market by a remarkable 6.7 percent annually on average through the 1973–2010 period, returning 21.5 percent. Think about that for a moment. Since the mid-1920s, stocks have returned about 9.9 percent annually.5 Owning out-of-favor stocks that have positive surprises will fetch you almost double the broad market return over time. We’ll look at the reasons for this astonishing increase in return shortly.

  Positive surprises also have a noticeable but more subdued effect on stocks in the middle quintiles. The middle group outperformed the market by 1.4 percent in the same quarter. But the above-market return stayed the same for the remaining nine months. The stocks don’t continue to appreciate steadily. The price impact due to positive surprises is moderate, probably because they are the least under- or overvalued.

  Finally, positive surprises have far less effect on favorite stocks. Stocks that experience positive surprises outperform the market by 1.1 percent in the surprise quarter. The “best” stocks don’t keep i
mproving, however, as do those in the low-P/E group. Rather, they lose about half of their small gain over the next nine months.

  Although not shown, the lowest 20 percent of stocks ranked by price to cash flow or price to book value are remarkably similar. Both sharply outperformed the market for the surprise quarter and for the full year and routed the most favored stocks for the two periods. The result for the 60 percent of stocks in the middle quintiles is close to those of the other middle groups in the previous charts.

  Why do positive surprises for “best” stocks cause only a moderate rise in the surprise quarter? Since analysts and investors alike believe that they can judge precisely which stocks will be the real winners in the years ahead, a positive surprise does little more than confirm their expectations. It’s no great shakes—the top companies are expected to have rapidly growing revenues, market share, and earnings. By the end of the year, therefore, the effect of the surprise is minuscule. As we’ll soon see, some recent neuroeconomic findings seem to explain why there are these various reactions to surprise by the favored, out-of-favor, and middle groups.

  Investors react very differently to positive surprises for out-of-favor companies, no matter which of the three value yardsticks we measure them by. Those stocks moved into the lowest category precisely because they were expected to continue to be dullards. They are the dogs of the investment world and investors believe they deserve minimal valuations. A positive earnings surprise for a stock in this group is an event. Investors sit up and take notice. Maybe, they think, these companies are not as bad as analysts and investors believed. The prices of out-of-favor stocks, therefore, do not just move up in the quarter of the surprise and then drop back again, as do those of the favorites. Instead, they continue to move steadily higher relative to the market in the year following the surprise.