March 20th, 2020 | by Ashvin Viswanathan, CFA
Does Market Timing Work?
The recent surge in market volatility and associated plunge in stock prices have been quite remarkable. In less than four weeks, the S&P 500 Index plummeted 27% from a record high on February 19 to its close on March 16, signaling the arrival of a bear market and a close to a remarkably long, 11-year bull run for US equities. Some investors are wondering if it’s too late to sell their stocks; others are asking if now is an appropriate time to start picking up shares at current marked-down prices. In this context, we thought now is an opportune time to bring some research discipline to bear on the rather controversial (and emotional) topic of stock-market timing.
The financial news media are filled with self-described market experts who counsel investors to buy or sell based on opinion based on different arguments. Some of these pundits may use trend-following charts, such as 200-day moving averages, to come to a market-timing conclusion. Other analysts may use valuation yardsticks and, for example, advise investors to sell when stocks are in the top decile of valuation by history, or buy when they’re priced in the lowest-valuation decile.
Strategic Allocation vs. Market Timing
We looked back over more than nine decades of market data, and analyzed just how difficult it is to time markets successfully. Here’s how we conducted our study: We compared the performance of a 60/40 balanced portfolio (60% S&P 500 Index/40% 5-year US Treasury Notes) with that of a market timer. For the 60/40 portfolio, we rebalanced equities back to the 60% neutral allocation when market movements caused equities to move more than 10% relative to their target allocation (i.e., up to a 66% portfolio weighting or down to 54%). The market timer also starts with a 60/40 portfolio, but dumps equities and holds 100% bonds when he expects a market correction, and moves back to 60% equities when he feels we’re past the market bottom.
Beginning in 1926, we ran about 3,250 scenarios in which we shifted equities to bonds from between zero and 18 months prior to the market peak, and reallocated to equities from zero to 18 months after the actual market trough. We then compared the annualized return to the next market peak of this market-timing approach to that of the buy-and-hold, but actively rebalanced, 60/40 portfolio.
Figure 1 illustrates some of the results in a bar chart. For example, if an investor has perfect timing (how many of those have you ever met?) and sells stocks precisely at the peak and buys them back at the very bottom of the market, his portfolio will outperform a systematically managed balanced strategy by 2.2 percentage points annualized. But if the investor’s timing is off by more than six months on both the timing of when to sell out of the market and when to buy back into equities, he would have been better off doing nothing more than sticking to his (systematically rebalanced) 60/40 strategic allocation.
In our study, overall the 60/40 strategic allocation method held the upper hand in 57% of scenarios across 18 months prior to market peaks and following market troughs, and in virtually all scenarios where the investors misjudged the end and beginning of a market downturn by six months or greater. In reality, the discrepancy is even larger since many market-timers are off by more than 18 months—we all know investors who bailed out of stocks in 2008 and never reinvested, or declared 2013 the top of the market and sold.
We can’t predict how far the market will fall. Not every bear market presages a recession, and not every recession is accompanied by a 50% collapse in stock prices, as we saw in the so-called Great Recession of 2008-2009. We do know that, as demonstrated above, attempts at market timing can be harmful to one’s financial health. For this reason, we advocate the type of strategic rebalancing discussed earlier in this article.
Bear markets are an inevitable part of investing in stocks—and one reason long-term investors earn the equity risk premium. Stock prices have always (eventually) returned to new highs, and often much faster than market pundits have predicted. Past is not prologue, but we measured the five most recent market corrections (Figure 2) prior to the current slump, and looked at how many days stocks took to rebound. Of course, there are also outliers, such as the market crash that helped to trigger the Great Recession—in that case, the market cratered 57% from October 2007 to March 2009 and took 1,046 days to recover. Let’s hope we don’t repeat that history.
Some market observers act as if they have the ability to foretell market peaks and troughs. In fact, those with such abilities are few and far between. As our research shows, most investors may be better off adhering to a disciplined approach such as holding a periodically rebalanced, strategically allocated portfolio.