Tuesday, October 31, 2006

Anecdotal Reasoning Yields Anecdotal Profits

I recently encountered a market analysis that drew a parallel between the March-May runup in stock prices and the recent August-October rise. The implication was that we were due for a significant correction, as we had beginning in May.

This kind of anecdotal reasoning which, in essence, says, "This market looks like this, so it should do that" has little place in the playbooks of serious traders. Because a chart or oscillator has a particular shape doesn't necessarily reflect supply and demand in the marketplace. This is why such superficial measures of the market fare so poorly when they are put to the test. Indeed, of over 6000 common patterns in the S&P 500 market analyzed by David Aronson in his excellent text "Evidence-Based Technical Analysis", none led to statistically significant profits.

One reason that the August-October runup is different from its predecessor is that far more stocks have participated throughout the move. Above we see a chart of the number of stocks making new 65-day highs minus new 65-day lows (blue line) vs. the S&P 500 (SPY; red line). As the market moved higher from March-May, the net number of stocks making new highs was dwindling. As the market has moved higher from August-October, we've seen expanding net new highs.

Why is this important?

Since 2004, we've only had 24 occasions in which we've had 1250 or more new 65-day highs on a given day. This most recently occurred on Thursday, when we had 1450 new highs across the major stock exchanges. Twenty days after such occasions, SPY has been up by an average of 1.01% (21 up, 3 down), much stronger than the average 20-day gain of .56% (433 up, 260 down) for the entire sample.

Anecdotal reasoning may tell us that we're due for a significant correction, but my previous research, as well as these findings, just don't bear that out. How we make new highs is crucial to whether or not those new highs are sustained and extended. Another way of saying this is that the trajectory of a market move matters as much as the trend.

There are plenty of market gurus who offer trading advice based on anecdotal reasoning. Their promises and claims of success are equally anecdotal. Trading with a knowledge of market history is hardly infallible, but it beats the alternative: trading in ignorance of history.

Later this AM, we'll commence the Morning With the Doc and track the market in real time. Also, please be my guest at the free panel discussion I'll be participating in on Thursday, November 2nd at the Chicago Mercantile Exchange. It will also be Webcast; should be fun. Thanks for all your interest in the blog, as well as your insightful comments.

2 comments:

John Cook said...

Yet another piece of solid research. This is the exact reason why I sold the Mays highs and have been buying the breakouts from balance during the Aug to Oct uptrend. Keep up the great work.

Brett Steenbarger, Ph.D. said...

Thanks, John, and congratulations on your timing--

Brett