Is It a Horse?

Even though it looks like a horse, smells like a horse, and gallops like a horse, is it really a horse? In a word, though we have demonstrated the important qualities of contrarian sti-ategies, can we really prove that investor psychology is the cause of these overreactions, rather than some other undetected factors? This is an important question, because if we can prove it, we can develop even more powerful sti-ategies to take advantage of overreactions. So far as I know, this question has not been explored before.

The key is whether the fundamentals of favored stocks deteriorate, resulting in sharply lower prices, while those of out-of favor issues improve, markedly boosting their performance. If this is the case, then the price changes are merely a reflection of deteriorating fundamentals for "best" stocks, an argument which will no doubt be made by our efficient market friends. Stock prices, they will say, are doing no more than responding to new events, which is exactiy what they are expected to do in an efficient market. Sorry, guys, but no.

A study I did in collaboration with Eric Lufkin pinpoints exactiy how powerful the psychological effects of overreaction are. It shows how ove riced the "best" (and underpriced the "worst") stocks become before they correct, as well as how small the underlying changes in a stocks fundamentals actually are that cause such major price moves.

To do tills we used tiie Compustat 1500 from 1973 to 1996 and selected five of the most widely used fundamental measures to characterize how good or bad the outiook for a company was. These were retum on equity (ROE), profit margin, growth in sales, growth in cash flow, and growth in eamings. The higher each of these measures is, the more promising a company appears to investors. We averaged each of these measurements for five-year periods.* These measurements, which well call growth and profitability indicators, demonstrate how rapidly a company is growing and how profitable this growth is.

the market, while "worst" stocks move higher, for a relatively long time following a surprise.

It may fall short of a ten commandants for contrarians, but remember, all five predictions of the investor overreaction hypothesis have been confirmed to a high level of statistical probability in the studies reviewed in chapter 6.

-5 years 0 +5 years

Measiuing points of Best and Worst Groups

Year 0 -► Point at which portfolios of best and worst stocks are assembled Period A:

Years 0 to +5-► Performance from point 0 to 5 years later

Period B:

Years -5 to 0 *" Backplaying performance for the 5 years before point 0

Next we determined how much the growth and profitability indicators changed for both the lowest and highest price-to-eamings, price-to-cash flow, price-to-boolc value, and price-to-dividend quintiles. This tells us if improvement in the underlying indicators is the real cause of outperformance for the out of-favor stocks and underperformance for the favorites.

Figure 11-1 and Table 11-2 show the results. Figure 11-1 measures the retums of the lowest 20%, and the highest 20% of price-to-book value stocks for the five years after the portfolios are set up (Period A, 0 to -1-5 years). We can see at a glance that low price-to-book value stocks do significandy better than high P/BVs. For the five-year holding periods between 1973 and 1996 they retum 153%, some 64% more than the highest price-to-book value group, and 34% more than the market.°

The next question is whether the fundamentals of low price-to-book stocks have improved (and those of high price-to-book deteriorated) enough to be responsible for this impressive price reversal.

To answer this question, we take the five growth and profitability indicators and average them over the same five-year periods that we just

Figure 11-1

Back from the Future

How Stocks Perform Before and After They Move Into the "Best" and "Worst" Groups 1973-1996

Price Movement Before Price Movement After

-5 to 0 Years 0 to +5 Years

Low P/BV 56.2% 153.3%

High P/BV 363.5% 88.8%

Market 141.3% 119.1%

Cash Flow Growth

Sales Growth

Eamings Growth

Return on Equity

Profit Margin

An investor buys stocks based on growth and profitability indicators:

• 0: Measuring point of best and worst groups (point at which portfolios are

formed).

• 0 to +5 years (Period A): The period during which the portfolio is held.

• -5 to 0 Years (Period B): Investors use this period 5 years before to 0, and 0 to

+5 years, to determine pricing in Period A (0 to +5 years) in Figure 11-1.

• -5 to -10 Years: Investors use this period, as well as -5 to 0 years, to determine

pricing in Period (-5 years up to 0), in Figure 11-1.

measured performance (Period A, 0 to +5 years). Table 11-2 (column 3), gives these resuks.

We then look back at the growth and profitability indicators from the point we began our stock performance measurements (point 0). When evaluating stocks at this point, investors examine the record of each of the five important indicators for the previous five years (-5 to 0), to decide which stocks have the most favorable and least favorable prospects. This second set of growth and profitability indicators is shown in Table 11-2, column 2.

Table 11-2

Growth and Profitability Indicators 1973-1996