Trying To Time The Value Premium

Market timing the value premium is a non-starter; slow and steady wins the race.

Whenever small-cap stocks produce a run of years when they outperform large stocks, the media is filled with warnings from Wall Street that “this is the year of large stocks.” The same is true when value stocks outperform growth stocks for a few years. The noise from the Street has been particularly loud since we have now had the second longest run of value outperforming growth since 1927—seven straight years (2000–06). Should an investor listen to such warnings? Should they alter their investment strategy? This paper addresses these questions.

The value premium is one of the most well documented facts in finance. To calculate the value premium, financial economists take the return of value stocks (defined as stocks within the top 30 percent of stocks when ranked by book-to-market [BtM] value) and subtract the return of growth stocks (defined as stocks within the bottom 30 percent when so ranked). The term HmL—the return of high (H) BtM stocks minus (m) the return of low (L) BtM stocks—is the term used for the value premium. For the eighty-year period 1927–2006, HmL was 5.0 percent on an annual average basis. The annualized (compound) HmL was 4.3 percent. Let’s look at some other important data related to HmL.

• The premium appeared with a high degree of persistence—HmL was positive in 51 of the 80 years, or 64 percent of the time. Of course, that means it was negative 36 percent of the time. That is the nature of risk. If value stocks always outperformed growth stocks, they would not be riskier.
• The standard deviation of HmL was over 2.5 times the annual premium at 12.7 percent. This high figure shows the nature of the risks of investing in value stocks. The highest HmL occurred in 2000, when it reached 37.8 percent. Ironically, the lowest occurred the prior year when HmL was a negative 26.1 percent (probably causing many investors to panic and sell). 
• There have been relatively long periods where HmL has been both positive and negative. For example, in the five-year period 1927–31, HmL was negative in four of the five years. It was also negative in four of the six years from 1934–39, three of the five years from 1949-53, three of the five years from 1956–60, four of the six years from 1966 through 1971, all three years from 1978–80, and five of the seven years from 1985–91. On the other hand, it was positive all nine years from 1940 through 1948, all five years from 1961 through 1965, all six years from 1972 through 1977, and all seven years from 2000 through 2007.

Hopefully, you can see that while HmL has been fairly persistent, there has been no predictable pattern to the premium. The only way investors could have reliably earned the value premium is if they had the discipline to maintain their exposure throughout both the good and bad times. In order to have done that, investors would have been required to ignore the clarion calls from Wall Street and the media that we often here after a few years of value outperformance. Calls like “This is the year of growth stocks,” have been heard almost continually since 2002.

If, after experiencing the huge value premiums of the first three years of the new century (2000–02) when HmL was 37.8, 14.4 and 12.2 percent, respectively, investors listened to such calls and sold their value holdings to buy growth stocks they would have missed the next four years when HmL was 3.1, 8.3, 8.2 and 5.8 percent, respectively.

It is important to note that the data on the overall stock market is fairly similar. Over the same eighty-year period, the equity risk premium (calculated as the return to stocks minus the return on one-month Treasury bills) has been 8.2 percent (annual basis), with a standard deviation of 20.1 percent. Also, stocks have outperformed Treasury bills with a high degree of persistence (the risk premium was positive 74 percent of the years). The evidence also demonstrates that there is a negative correlation between future returns and current valuations—low P/Es forecast relatively high returns and vice versa. However, the evidence shows that there has been no persistent ability to time the market—increasing equity allocations ahead of the bull and lowering them ahead of the bear emerging. Consider the following example: A study of one hundred large pension funds and their experience with market timing found that while they all had engaged in at least some market timing, not one had improved its rate of return as a result. In fact, eighty-nine of the one hundred lost as a result of their efforts, and their losses averaged an incredible 4.5 percent over the five-year period.1

There are two reasons that trying to time the market is unlikely to be successful. First, ex-ante (before-the-fact) there should always be an equity risk premium because stocks are always riskier than one-month Treasury bills. All that high valuations predict is relatively low future expected returns; but those returns should still be higher than the returns on a riskless security. Second, the equity risk premium is so high that timing efforts would have to be right almost all the time to be successful.    

Just as there is no evidence supporting the view that investors are likely to succeed in their efforts to time the market, there is no evidence that they can “time” the value premium with persistence. If there were, we would see evidence of active managers outperforming passive benchmarks with persistence greater than randomly expected. Yet, we don’t see such persistence.

There is an important fact about value stocks and their returns that investors should understand because it will help them ignore the media (and their friends). It is a fact that most investors are unaware of—most of the HmL premium comes from a small percentage of value stocks that produce very high returns. Their outperformance often leads to them “migrating” out of the value asset class. That means that the remaining group of value stocks is a different group than existed the prior year. The same thing is true of small stocks—much of their outperformance results from a small group of small stocks “migrating”—they become large stocks as a result of their outperformance. Thus, the group of small stocks is different from year to year.

Even with this knowledge, there are those that believe that you can still time the value premium based on the relative spreads between the valuations of value and growth stocks. In other words, when the spread between the book-to-market (or price-to-earnings) ratios of value and growth stocks is wider than the historical average, investors should load up on value stocks. On the other hand, when the ratio is relatively low, they should abandon value stocks and move to growth stocks. This would seem to make sense since studies have found that when the spread in book-to-market (BtM) ratios between value stocks and growth stocks is high, the subsequent value premium tends to be high. The reverse is also true. Based on that information, if next year’s value premium is expected to be high, it would seem logical to own value stocks. If it were expected to be low, then growth stocks would seem to become the logical choice. Is it really that simple to earn abnormal returns? Does a statistical relation always translate into a viable portfolio strategy? These are the questions Jim Davis asked and answered in his study “Does Predicting the Value Premium Earn Abnormal Returns?”2  The study covered the period July 1927–June 2005.

Davis found that style-timing rules did not generate high average returns despite being able to use future information about BtM spreads. In fact, he concluded that that the expected excess return of style timing is probably negative—for the same reasons that efforts to try to time the overall market are likely to fail. Just as ex-ante there should always be an equity risk premium, ex-ante there should always be a value risk premium. And as is the case with the equity risk premium, the value premium is so large that any trading strategy would have to be right almost all of the time to deliver successful results. It would be like switching from the high-speed carpool lane to the center lane on a crowded freeway. Your “freeway algorithm” might help predict when the carpool lane or center lane will move faster or slower than normal. But will it be accurate enough to justify switching into the slow lane in an effort to get there quicker? The evidence suggests that you are better off staying in the carpool lane. The lesson for investors is that just because a statistical relation exists does not necessarily imply that a profitable trading strategy based on that relationship exists, especially after taking into account trading and other costs.

The bottom line for investors is that the prudent strategy is to ignore the calls to action you hear from Wall Street and the media and adhere to your investment plan. The only actions you should be taking are to rebalance your portfolio and to harvest losses when that can be done in a tax-efficient manner.

1.      Charles Ellis, Investment Policy (Irwin Professional Pub 2nd edition 1992).

2.      James L. Davis, “Does Predicting the Value Premium Earn Abnormal Returns?” January 2007.