Falkenblog had a post making this argument with more than the usual gusto. Main points:

Arithmetic vs. geometric averages: when an index goes from 100 to 200 and back again, the average annual return on an arithmetic basis is 25% (200/2 + 50/2 = 125). The geometric average, which shows 0% change, would seem to be more relevant to the long-term investor.

Survivorship bias: the U.S. had the best stock market in the 20th century –it’s not a good benchmark for what to expect going forward on average.

After-tax returns: taxes applicable to the equity premium reduce the margin (even in RRSPs since they just defer taxes).

Market timing: dollar-weighted returns, reflecting high inflows at peaks and outflows at troughs, are lower than time-weighted returns.

Transactions costs: “commissions were about 60 cents/share until the 1975 deregulation and are currently about 2 cents a share (about 0.1%) on average. Plus, mutual funds often had 8.5% fees. …. the bid-ask spread will cost you about 0.25% on average….”

Add up all these factors, and the equity premium shrinks to zero or worse for the average investor, says Falkenblog.

Food for thought, as they say. Personally, one issue I am wrestling with is: even if the equity premium does exist, how can it be expected to persist? To rephrase, how can one reconcile a positive equity premium with the efficient market theorem.

The latter says a systemic opportunity to profit doesn’t persist in the stock market because market participants capture such profit opportunities by biding stock prices up or down until the anomaly is eliminated. Yet, the equity premium says there is a systematic opportunity to profit in stocks by buying and holding over the long run.

Ten or fifteen years ago, there weren’t many books or studies alerting investors to the premium, so not many were responding to it. But now everyone knows about it, so we might expect many investors to have incorporated it into their strategies (or be in the process of doing so). Just look, for example, at how much pension funds and other institutional investors have shifted out of bonds toward equities over the past 10 to 15 years.

The end result may be that stock prices have been bid up relative to long-term fundamentals such that the equity premium will turn out to be close to zero or negative (at least more so for investors who buy during the mature  bullish phases).

Say an investor put $10,000 into the stocks of the Dow Jones Industrial Average (DJIA) during the six months to April 30, 1950. Then they rolled the proceeds successively into the same six months every year to 2008. The $10,000 would have gained $531,444. If they did the same thing from May 1 to Oct. 30, the $10,000 would have gained a mere $1,021.

Even more astounding, says the Almanac, is that one can triple the return on the best six months to $1,546,114 by tweaking entry/exits points. Instead of using the first/last day of the six-month period to buy/sell, one can use a simple Moving Average Convergence Divergence (MACD) indicator. In upward trending markets, the MACD gets the investor in earlier and longer; in downward markets, it gets them in later and out sooner.

Also, it appears the Almanac analysis did not use total-return indexes (which include dividends). If they did, the gains would be even larger! On a side note, one wonders if the annual returns would be sensitive to the starting year, and how the results would have been affected if the investor averaged in a position over several years.

I did a quick search for academic studies on this seasonal pattern but came up with little. Anybody know of any? But I do know there has been academic research into “market anomalies” in general and much of it warns investors to be wary of investing on the basis of historical patterns because of the Efficient Market Theorem (EMT).

As awareness of the repeating profit opportunity spreads, goes the EMT, it is discounted by investors seeking to capture the profits. If one is coming in at the late stages of the discounting process, they’ll end up taking a random walk. The Almanac, an annual publication, has been expounding the “best six months” strategy since the mid-1980s. More details on the lags etc. in the EMT are in this July 17, 2007 post.