A new strategy for trading market reversals

My recent article and its follow-up, which documented the loss of trending in the S&P 500 Index, also found evidence of reversals: Short-term moves in the S&P tended to reverse direction rather than continue. That has led me to examine reversal patterns in the S&P on my research blog, culminating in evidence I've found for countertrend equivalence. Countertrend equivalence is simply a 50 cent term that means the following:

If the market establishes a trend over X periods, the next X periods will tend to move against this trend.

Using a proprietary trending indicator, I found countertrend equivalence for both five-day periods and five-hour periods. When we had consistent uptrending in (SPY | Quote | Chart | News | PowerRating) over a five-day period, the next five days yielded subnormal returns. When we had consistent downtrending in SPY over a five-day period, the next five days yielded well above average returns. The identical pattern showed up when I looked at the next five hours as a function of the trending behavior of the previous five hours.

This is significant, because it suggests that we might be able to profit from market reversal patterns by looking at trends at multiple time frames. Selling the S&P when we are registering uptrends over the last five hours and the last five days (and vice versa) would potentially blend two sources of edge, signaling trades that might be worth scaling into over time.

I decided to try to replicate the findings by simply using five-day highs and lows in place of the trend measure. This analysis also replicates work done by Larry Connors and Conor Sen in their book "How Markets Really Work". In that book, Larry and Conor found (among other things) evidence of countertrend behavior after five-day highs and five-day lows. Going back to March, 2003 (N = 767), I found 285 occasions of five-day highs in SPY. Over the next five days, the market averaged a gain of .16% (161 up, 124 down). This is considerably weaker than the average five-day gain in SPY when we don't have a five-day high (.39%; 287 up, 195 down).

During that same period, I found 181 occasions in which we had five-day lows. Five days later in SPY, we averaged a gain of .66% (113 up, 68 down). This is much stronger than the average five-day gain of .20% (235 up, 251 down) when we have not made a five-day low. Fading five-day highs and lows, as noted by Connors and Sen, has indeed been a profitable strategy.

I then decided to look at the market intraday, by focusing on five-hour highs and lows. When the market has made a five-hour high (N = 164), the next five hours in SPY average a gain of .018% (88 up, 76 down). This is modestly weaker than occasions in which the market has not made a five-hour high. In those circumstances, the next five days in SPY average a gain of .027% (204 up, 187 down).

When we've made a five-hour low in SPY (N = 147), the next five hours in SPY average a gain of .045% (80 up, 67 down). This is stronger than the average five-hour gain in SPY when we have not made a five-hour low (.017%; 212 up, 196 down).

Again we see evidence of countertrend equivalence: strong markets are displaying weak returns over multiple time frames and weak markets are displaying strong returns. While the countertrend tendencies with five-day--and especially five-hour--highs and lows are not as pronounced as they are with a dedicated trend measure, they nonetheless point the way to potential trading strategies employing multiple time frames and stocks. I made no effort to optimize results in my selection of timeframes; it may well be that alternate periods--and trading instruments--will yield superior returns, as Connors and Sen's research indeed suggests.

Brett N. Steenbarger, Ph.D. is Associate Clinical Professor of Psychiatry and Behavioral Sciences at SUNY Upstate Medical University in Syracuse, NY and author of The Psychology of Trading (Wiley, 2003). As Director of Trader Development for Kingstree Trading, LLC in Chicago, he has mentored numerous professional traders and coordinated a training program for traders. An active trader of the stock indexes, Brett utilizes statistically-based pattern recognition for intraday trading. Brett does not offer commercial services to traders, but maintains an archive of articles and a trading blog at www.brettsteenbarger.com and a blog of market analytics at www.traderfeed.blogspot.com. He is currently writing a book on the topics of trader development and the enhancement of trader performance due for publication this fall (Wiley).