Monday, January 28, 2008

When Bearish Sentiment Hits an Extreme


The American Association of Individual Investors (AAII) conducts a weekly poll of investor sentiment that, over the years, has been reasonably good as a contrary indicator. This past week, the percentage of bears in the survey hit 59%, a historically high number.

Going back to 1987, which is when my weekly AAII data begins, we've only seen one other period with similar bearishness: very late August and mid October, 1990. Other periods with more than 50% bears have included February, 2003; July, 2006; October, 1992; October, 2002; May, 2007; and November, 2007. Most of these were superior opportunities to be buying stocks for at least the intermediate term.

The chart above takes the percentage of bears minus bulls in each weekly survey since 2002 and plots the resulting difference against the S&P 500 Index (SPY). We can see that bearishness has been increasing of late--including the current period--which is why three of the highest bearish readings since 1987 have occurred within the last year.

Going back to 2002 (N = 307 weeks), we had 88 occasions in which bearish sentiment has been greater than bullish sentiment in the AAII poll. Five weeks later, the S&P 500 Index (SPY) has averaged a gain of .90% (60 up, 28 down)--much stronger than the average five-week gain of .39% (127 up, 92 down) for the remainder of the sample.

Indeed, when bears have outnumber bulls by 20% or more since 2002 (N = 14; as is the case at present), SPY has been higher on 13 occasions, by a whopping average of 3.63%.

When bearish sentiment hits an extreme, it generally means that the bears have largely done their selling, setting up conditions for short-covering and some value-oriented buying.

RELATED POST:

Bearish Extremes in the Equity Put-Call Ratio
.

2 comments:

Angella said...

Awesome nice post and info!
Time to get out the shopping list from xmas that was completely obliterated. :)

Hunting for high probability plays!

Brett Steenbarger, Ph.D. said...

Thanks, and best of luck with those high-probability plays!

Brett