Adverse Excursions, Part 2, 08/02/07
Another good exercise: Find the odds that an x% decline will turn into a y% decline.
This can be done by adding up all the declines of magnitude x% and their adverse excursions of magnitude y%. Divide the count of y by the count of x and you have a crude indicator of the liklihood an x% decline will turn into a y% decline. Breaking this out by era is enlightening; examples for the SP are 82-p, 82-p w/o 00-02, and 97 - p w/o 00-02.
The entire adverse excursion field is fascinating and branches nicely into, "How do my trading systems perform n days after an event of magnitude x?" and its corollaries.