Does the market's break late last week below its 50-day moving average mean that the intermediate trend is now down?
Many technically oriented investors think it does, which is no doubt one reason why the market plunged last Friday.
But I am not so sure that breaking the 50-day moving average has the significance that technicians are giving to it. In fact, a careful analysis of the stock market in recent decades does not find that the market performed appreciably better when it was above the 50-day moving average than when below.
Take the period since early 2000, right at the top of the Internet bubble. We've had two severe bear markets since then, of course, which means that the intervening period is precisely the kind in which a market-timing system like the 50-day moving average should be able to show its stuff.
And, yet, it hasn't added value since then.
Consider a hypothetical portfolio that switched between a total-stock-market index fund and 90-day Treasury bills whenever the S&P 500 index (.SPX) crossed above or below its 50-day moving average. Since the beginning of 2000, this hypothetical portfolio produced an annualized return that was 0.2 of a percentage point below a simple buy-and-hold strategy. Note that these returns take into account the dividends the portfolio would earn while invested in the stock market and the interest earned by the T-bills when the portfolio was out of the market.
Crucially, however, these returns do not take transaction costs into account. If I had included them in my calculations, then the moving-average portfolio would have lagged a buy-and-hold by even more. In other words, even without transaction costs, this 50-day moving average portfolio lagged the market since 2000.
To be sure, the 50-day moving average hasn't always been this poor of a market-timing indicator. But you have to go back several decades before you find a period in which it beat a buy-and-hold strategy. In the decade of the 1990s, for example, the 50-day moving average strategy lagged a buy-and-hold strategy by an even greater margin.
Blake LeBaron, a finance professor at Brandeis University who has extensively analyzed various technical-analysis strategies including moving averages, says that it appears as though, around 1990, something changed in the financial markets that largely eliminated the moving average's potential as a market-timing indicator.
What might that something be? Prof. LeBaron speculates that one culprit might be the moving averages increasing popularity, coupled with the greater ease with which investors could follow them — due to lower commissions and increasingly sophisticated trading technologies. As more and more investors begin to follow a system, of course, its potential to beat the market begins to evaporate.
In any case, Prof. LeBaron noted, moving-average systems stopped being profitable in the foreign-exchange markets at about the same time that they lost their effectiveness in timing the U.S. equity market. This increases the likelihood that whatever caused the moving average to become less profitable in the stock market was more than just a fluke.
Of course, it's always possible that the stock market is, in fact, now in the early stages of a correction, or something even worse. My point is it's premature to conclude that it is, simply because the market has now broken below its 50-day moving average.