This column was originally published in June 2014.
Things change
Robert Huebscher of Advisor Perspectives, an online publication with many fine (and free!) articles, recently posed four questions about value investing. He offered his own answers, which are sound and which I recommend that you read, but he also wondered about my thoughts. Here goes.
Is there an economic reason why small-company and value stocks have performed well historically?
(For clarification: The tactic of investing according to a single characteristic, such as the company being small or its shares trading at relatively low price multiples (that is, value stocks), has many names. The approach may be called purchasing anomalies, factors, or risk factors. Recently, it has been dubbed smart beta. As that reeks of smart marketing, Morningstar prefers instead the term strategic beta. Anomalies, factors, risk factors, smart beta, strategic beta ... different ways of saying the same thing.)
Actually, there is some question whether small-company stocks really do all that well. The "small-company effect" was the first academically documented anomaly, as Rolf Banz’s 1981 paper predated the initial academic research on value investing, and was once widely believed. It stands to reason that smaller companies are riskier businesses than are larger firms, and thus deliver higher returns. But performance over the past 35 years has been disappointing, particularly when the theoretical gains are adjusted for the hard reality of transaction costs.
To the extent that smaller companies do outperform, those gains likely owe to a liquidity premium. Smaller-company shares have lower trading volume, which increases the chance of moving the stock price by putting in a trade order, and which hampers the investor’s ability to rapidly enter or exit a position. Low liquidity is a real cost that deserves to be compensated with a real return. This is fine--but properly speaking, it’s not a small-company effect.
As for value investing, yes, I do think there is an economic reason why they have performed well historically. By many risk measures, value stocks appear to be less volatile than growth stocks, but it is also true that when stock markets suffer traumatic, once-in-a-generation losses, the drawdown on value stocks is higher than the drawdown on growth stocks. As these market meltdowns tend to occur when the economy is plummeting, value stocks struggle the most at the worst possible time. That, too, is a risk that deserves compensation.
I don’t think that risk is anywhere near the size of value investing’s historic premium, though.
Has the value premium dissipated over time?
Happily--as I did not have the data immediately at hand--Huebscher answers his own question. He cites Ken French’s research as showing that value stocks outperformed by 4.57% annually before 1992, and by 2.78% annually since.
The surprise is not that the advantage declined, but that it continued to exist at all. After all, the market risk associated with value stocks appears only rarely. In addition, that danger is moderate rather than severe because the alternative to value investing, growth stocks, also gets whacked by a market crash. Yes, growth stocks figure to lose somewhat less under such circumstances, but smacking into the pavement after a 35-foot fall, rather than one of 40 feet, offers scant consolation.
Does the persistence of the value premium rely on a large group of investors behaving “stupidly”?
It’s good to win a Nobel Prize.
Many claim that the value premium owes to behavioral reasons. People focus heavily on recent events, a bias that leads them to overestimate the prospects of currently strong growth companies and to underestimate the futures of currently downtrodden, value-priced firms. They are overly influenced by dramatic examples when calculating an event’s probability; say "growth stock," and they tend to think "Apple (AAPL)." Also, value stocks can be embarrassing. Who wants to be known for owning Eastman Kodak until the bitter end? The '70s came and went, pal.
If you and I were to flatly dispute these arguments, without addressing the details, we would be naive. If Nobel Memorial Prize-winner William Sharpe does that, he is profound. Professor Sharpe maintains, to use Huebscher’s paraphrase, that "smart-beta proponents assume that, if they are 'smart' to invest in securities such as small-cap and value stocks, then there must dumb investors who hold the market portfolio and others who are "really dumb" and underweight those securities. Therefore, according to Sharpe, smart-beta strategies rely on exploiting the stupidity of others."
You know what? That is profound. The argument for the value-stock premium assumes that people collectively are brilliant at forecasting individual-company prospects, so that few investors can successfully beat the market by selecting one company over another, without regard to investment style--yet they are bone-dead stupid in pricing growth versus value stocks, year after year, decade after decade. And they neither realize their stupidity nor address it.
That said, behavioral biases are notoriously difficult to erase. They tend to occur subconsciously and require much discipline to overcome. It is not as easy to become unstupid as Professor Sharpe's argument would suggest. After the first widespread publication of its existence, the value premium declined by 170 basis points per year--but it did not disappear. It's one thing to wound the beast, quite another to slay it.
Are there reasons to expect the premium to dissipate in the future?
Yes. That process is already under way.
News of the value-stock premium has spread so widely that I can write this column, on a free and (relatively) mass-market website, with the expectation that most readers will immediately understand the context. Dimensional Fund Advisors, dedicated to the proposition of being on the smart side of the smart-stupid bet, has ballooned to become one of 10 largest fund companies. Rob Arnott’s Research Affiliates, founded just over a decade ago, now runs $170 billion of "fundamentally" invested assets that look and act much like value stocks.
It’s possible that the current enthusiasm for value stocks has temporarily made them overappreciated rather than underappreciated, and that growth stocks will be the better bet over the next decade. Even if things have not gone that far, value stocks are scarcely the sure thing that they appear to be via the rearview mirror. Those who truly don't care about tracking error and who have a time horizon that extends for decades might wish to heavily favor value stocks. Most investors, however, should not tilt more than modestly in that direction.
Postscript: Since this article was published, growth stocks have demolished value stocks, roughly doubling their cumulative returns. That somewhat strengthens the argument for value investing, but not greatly, because growth stocks didn't benefit greatly from sentiment. Most of their gains owed to good, old-fashioned earnings growth--the reason that they were considered to be growth stocks in the first place.
John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for Morningstar.com and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.
