Monday, November 13, 2006

Anticipating Next Day Returns in the Russell 2000

The Russell 2000 has proven to be an increasingly popular trading vehicle. Indeed, average 50-day volume in the IWM ETF has risen by a factor of over 13 since July, 2003. The greater volatility of the small caps--and their outperformance through much of the recent bull market--have made the Russell an important trading market.

Looking over the data from 2004 - present (N = 719 trading days), we see that the Russell does not tend to reverse the move of the previous day. The average daily change in IWM has been .05% (378 up, 341 down). After an up day in IWM (N = 378), the next day's average change is .056% (195 up, 183 down)--no different from average.

Momentum effects do seem to occur in the Russell. When IWM is up and the number of stocks closing above their 20-day moving averages increases (N = 260), the next day in IWM averages a gain of .10% (142 up, 118 down). When IWM is up, but the number of stocks closing above their 20-day moving averages declines (N = 118), the next day in IWM averages a loss of -.04% (53 up, 65 down). Broad participation in market rallies thus tends to carry forward into strength the next day for the Russell; weak participation does not.

The Russell does show signs of reversal effects as well. When the prior five days in the Russell have been up (N = 214), the next day in IWM has averaged a loss of -.03% (98 up, 116 down). When the previous week in the Russell has been down (N = 164), the next day in IWM has averaged a gain of .17% (97 up, 67 down). Buying after a week of weakness has paid off well in the Russell; buying after a week of strength has not.

Finally, when we've had more than 1000 issues making fresh 20-day highs (N = 192), the next day in IWM has averaged a loss of -.01% (93 up, 99 down). When we've had fewer than 1000 issues making new 20-day highs (N = 186), the next day in IWM has averaged a gain of .12% (102 up, 84 down). Once again, we see that buying multi-day strength has not worked in the Russell.

It is when we combine these factors that we can sometimes discover meaningful edges. When we see strong upside participation following multiday weakness, holding for next day strength exploits a nice edge. A rise with poor participation at a multiday high offers a cautionary signal. Knowing how your stock or index has been trading--its patterns of next-day returns--tells you a great deal about whether you should trade or fade moves to the upside.