On the topic of using moving averages to avoid risk, one question I often get is, What actually would have happened over the past couple of decades? Everyone has an interest in avoiding a 2000 or 2008, but everyone also knows that market timing doesn’t work.
I agree that market timing does not work. And, in all cases, moving averages won’t get you out at the top or in at the bottom. You can, however, use them to get a warning of potential trouble. There is a trade-off between the amount of loss you suffer (or gain you miss) and the number of trades you make. Looking at the actual numbers can help to characterize that trade-off.
What do moving averages tell us?
For this analysis, I used monthly data for the S&P 500. I went back to the start of 1995 and looked at both 10-month (+/−200-day) and 20-month (+/−400-day) moving averages, on a monthly basis, with the following results.
Clearly, the 10-month moving average leaves a lot of money on the table; 60 percent of the time the model was out of the market, it missed out on gains. On the other hand, the four negative price changes it avoided were big losses. With the 10-month, though, you’re giving up quite a bit of upside to minimize the downside. For the 20-month, the odds are better—two out of three times the model was out of the market, there was a significant drawdown. This is why I start to get concerned when we hit the 10-month moving average—there’s a 40-percent chance of a significant drawdown—and I get very concerned at the 20-month, as the odds of a severe downturn have now risen to 67 percent.
Two points to keep in mind
Note that neither of these indicators times the market. Both have losses of around 10 percent, on average, from the peak before they trigger, which means you will get a trigger after the peak. The same applies for a buy trigger; you will not buy in at the bottom.
The valuation level of the market also makes a significant difference. When the market is expensive, drawdowns tend to be more severe, and tactical strategies such as moving averages tend to outperform. When the market is cheap, buy and hold is optimal. Right now, when the market is expensive, the use of a moving average might well add value.
What to do when a moving average breaks
This isn’t to say you should sell everything when the market drops a bit. Think of using moving averages like a hurricane warning if you live in Florida. When you get a warning, you don’t sell your house and move to Minnesota. Instead, you buy food and water, stock up on flashlight batteries, and maybe board up your windows. Similarly, when a moving average breaks, perhaps you hold off on putting additional money into stocks, or reallocate some stocks into lower-volatility funds, or even just take a deep breath, sit tight, and remind yourself you’re in it for the long haul. Hurricane predictions aren’t perfect, but they are useful, and the same goes for moving averages and other technical indicators. Right now, with the market expensive, the storm warnings aren’t flying, but it makes sense to keep an eye out for them.
| 10-Month MA | 20-Month MA | |
| Number of Transactions | 10 buys, 9 sells | 4 buys, 3 sells |
| Price Change While In | −6.2% to 108.4%, average 23.9% | 2.9% to 179.5%, average 68.1% |
| Price Change While Out | −37.8% to 14.8%, average −2.9% | −24.7% to 10.6%, average −11.5% |
| Price Changes When In | 8 positive, 2 negative | 4 positive |
| Price Changes When Out | 6 positive, 4 negative | 2 negative, 1 positive |
| Loss from Peak | −5.3% to −19.6%, average −9.1% | −7.2% to −14.1%, average −11.3% |