Here’s what the S&P 500’s 50-day moving average is telling us now about stocks

The 50-day moving average stopped being a reliable market-timing indicator around 30 years ago. As I will discuss in a moment, that timing is not an accident. But first, let me review the performance before and after 1993 of a strategy that switched between a broad-market index fund and U.S. Treasury bills according to whether the U.S. stock market was above or below its 50-day moving average. No transaction costs were debited.

From 1928 through the end of 1992, according to my calculations, this moving-average strategy outperformed a buy-and-hold alternative by almost 5% on an annualized basis — 14.3% to 9.7%. That’s a huge margin.

In contrast, from the beginning of 1993 through the end of this year’s first quarter, the moving-average strategy lagged a buy-and-hold one by almost 3% on an annualized basis — 7.4% versus 10.3%. (See the chart below, which illustrates the number of false signals the moving average has given over the past 18 months.)

The most likely cause of this reversal of fortune, according to several finance professors who have studied technical analysis, is the creation of the first broad-market index fund ETF — the SPDR S&P 500 ETF SPY — in January 1993. A related factor was the advent of online discount brokers, which precipitated a long decline in brokerage commissions to today’s near-zero rates. Both of these factors allowed investors to easily switch at little cost between stocks and cash with a single transaction.

These developments led to a significant increase in the number of investors switching into and out of the market according to the 50-day moving average. As is so often the case, this activity wound up killing the proverbial goose that lays the golden egg.

That “egg” had remained unbroken up until then because few investors were able to actually trade into and out of the market according to the moving average’s signals. Without ETFs (and index funds with no penalties for frequent trading), a moving-average strategy would have required executing a prohibitively large number of individual transactions. Even if you had wanted to go through all that work, it would have been close to impossible to execute all those transactions simultaneously. This was before the internet, after all.

Another hurdle would have been determining the dollar amount to invest in each individual stock, since the market benchmarks are either weighted by capitalization (as in the case of the S&P 500 index) or price (in the case of the Dow Jones Industrial Average DJIA ). Yet another factor would have been significant brokerage commissions; those commissions used to be high enough that they ate up almost every strategy’s theoretical profitability. From1928 through 1992, for example, a 1% commission on each trade (which is a conservative estimate) would have reduced the 50-day moving average’s return from 14.3% annualized to just 1.0% annualized.

You’d think that a three-decade stretch of underperformance would lead followers of the 50-day moving average to throw in the towel. But you’d be wrong. MarketWatch’s Joseph Adinolfi reported earlier this week that technical analysts were telling him that the bull market is in “jeopardy” because the S&P 500 broke its 50-day moving average. And as MarketWatch’s Christine Idzelis reported, Bespoke Investment Group has indicated that it was potentially bearish that the S&P 500 had broken its uptrend.

Are other moving averages better?

What about the market-timing ability of other moving averages, such as the 200-day moving average? The answer is that they in fact do better —on occasion. But I have found their long-term track records to be equally disappointing, regardless of whether the moving averages cover periods as short as the trailing 20 trading days to as long as 200 days.

The same skeptical conclusion goes for the various ways in which market technicians have tried to tweak moving-average strategies, such as so-called exponential averages (which weigh recent days more heavily than older ones) and moving-average crossover strategies (which are based on two moving averages of different lengths, with buy signals generated when the shorter-term average crosses above the longer-term one — and vice versa).

The investment implication is that the broad stock market is not a trend follower in any straightforward way. Many investors find this hard to believe, since when looking at a chart of the stock market’s history it seems obvious to them that there are distinct upward and downward trends — bull and bear markets. But the trends the market follows are of widely varying lengths and magnitudes, and are only obvious after the fact.

The market timing systems that I have found to have better success focus on internal market divergences. These systems are based on the premise that a healthy advance is one in which there is wide participation.

Hayes Martin, president of the advisory firm Market Extremes, is one such market timer. In an email, he said that the divergences that are most worrisome currently are in the Nasdaq COMP, where a couple of megacap stocks were able to push the stock exchange to a new high last week even as many other Nasdaq-listed stocks struggled. Martin says “pullbacks in the 10-15% range” are therefore possible for the technology and small-cap sectors. That said, he adds, “there is no evidence of a major top at this time.”

Mark Hulbert is a regular contributor to MarketWatch. His Hulbert Ratings tracks investment newsletters that pay a flat fee to be audited. He can be reached at mark@hulbertratings.com

Also read: Do cash-gushing stocks outperform the S&P 500? Here’s what history has to say.

More: Cutting interest rates is misguided — the easy money would only fuel inflation