This past week the number of bears rose to 49.5% and the bulls dropped to 20.7% so it is now at fairly extreme levels. The last time the Bull-Bear Spread was this low was in early July as the July rally was just beginning. Prior to that is was the first week of November of ’09 just as that rally was kicking off, and the time before that was early March of ’09 just before that rally began.
I looked at the data a number of different ways this weekend. (For those interested you may get all the data using the AAII link I provided above. Just scroll down and you’ll find a link to a historical spreadsheet on the right hand side of the page.) Below is a sample if the kind of results I saw when conducting some studies.
So what we see here is that an extremely pessimistic outlook from investors has been followed by a rise in the market on pretty consistent basis. Six weeks later the market has been higher 81% of the time and the average gain was over 7%. The problem is that the failures have been very, very large. Using the same 6-week time frame the average loss was 10% and the max loss was 27%. The max loss occurred in September/October of 2008. August of 1990 also saw a sharp decline in the spread to levels similar to current levels. That was followed by further selling that maxed out around a 12% decline between then and October. Other instances that were followed by large selloffs included July of 2002 with a 12% decline and February 2009 with a 14% decline.
So it appears the AAII Investment Survey is suggesting there is a good chance of a rally emerging over the next several weeks. But if the market can’t manage to rally then the probable alternative is a substantial selloff.
4 comments:
Are there any S&P trading edges during the month of September? We're down quite a bit from the highs earlier in the year and I was wondering...
I wish your tables contained MFE and MFA statistics.
Since 1987, the average percentage of Bullish individuals is 39% with a standard deviation of 11%. For the week ending 27/08/10, the Bullish percentage stood at 21%. That is over one & a half standard deviations below the total sample mean. Indeed, a very low percentage of Bullish individuals. I also note that this low level of positive sentiment coincided with a 6 week closing low for the S&P 500. By any account I would suggest that Bullishness is very Hall & Oates (She's Gone).
With this in mind I decided to test whether a change in form was likely given an AAII Bullish reading of less than 25% in conjunction with a 6 week closing low for the S&P 500.
4 weeks post this condition the average gain for the S&P 500 was 4%, Max 23, Min -5, St. Dev 5%, Win Rate 74%, Observations 31.
The data is statistically significant at greater than the 1 in 100 level. I suspect a rally is on the cards despite September being an unusually cruel month for equities.
Post a Comment