Exponentially Curb Your Enthusiasm

Investors and the financial media have been excited of late about the S&P 500 crossing above its 200-day moving average:

Harris Private Bank and Morgan Asset Management say the advance may indicate the bear market in U.S. equities that began in October 2007 is over, heralding more gains after a three- month, 39 percent increase. Analysts who base forecasts on price charts consider crossing above a moving average bullish because it shows stocks are rising faster than the long-term trend.

“That’s proven to be a fantastic signal,” said Jack Ablin, chief investment officer at Chicago-based Harris, which oversees $60 billion. “A lot of investors will view it as a bullish signal and will likely use that as a rationale to add exposure to the S&P 500.” [Bloomberg]

As I’ve discussed before (”Moving Averages and Human Risk“), plain vanilla moving average signals aren’t really that special: anything that keeps us from being long the market during its most volatile periods is likely to improve risk-adjusted returns, but we also want signals that have a genuine performance-boosting edge. Neither a 200-day simple moving average nor the exponential variation I’ll address below outperform enough to warrant consideration as independent indicators. But the exponential calculation certainly seems preferable, and suggests that, for the moment, investors would be better served to remain cautious.

The three backtested equity curves below display 1) a buy-and-hold portfolio tracking the S&P 500 (blue), 2) a long-only portfolio using the 200-day simple moving average (SMA) as a timing signal (red), and 3) a long-only portfolio using the 200-day exponential moving average (EMA) as a timing signal (orange).* Transaction costs are not included, nor are returns on cash, and returns are logarithmically scaled.

The portfolio using an EMA timing signal is about 16% higher at the moment than the portfolio following the SMA. The rather intuitive advantage provided by an exponential moving average is that it is more responsive to new information. When prices begin rising in the midst of a downtrend, the EMA rises in response more quickly than its simpler cousin; conversely, a selloff amidst a bullish trend will push the average lower more quickly. In both cases, “false positives” are less likely: the emphasis applied to more recent prices forces the market to work harder to prove a trend reversal.

That’s the big picture. How does it apply to the current situation? As the chart above shows, the break above the 200-day SMA (red line) that occurred in the last hour of trading on June 1 did not close above the 200-day EMA (yellow line). The index closed a few cents above its EMA the following day in SPX and SPY, though not in the Emini S&P futures. It is now caught between the two, above the simple but below the exponential average, meaning that investors using the latter as a timing signal are not yet willing to break out the Dow 10,000 hats and champagne.

* The multiplier used in calculating the EMA for this study was 2/(period+1).