He uses data over a 100 year span, in several different markets, so this was no casual study. It's possible that the difference between what you found and his studies are related to the time span involved or the frequency of the data. I don't have the books with me, but I remember one example was based on 5 day returns on the S&P.
Mine was 50 years for the S&P and 16 for the NAZ, both based on weekly close. The variation, the envelope, and the violation of that envelope were based on the relationship between those closing prices and the trend, thus eliminating some of the crazier day-to-day variation. And, I was looking at the trend and variation in the price, not the return. Analyze different numbers, get different results!
With a LTB&H focus, my interest was in long term price trends, not in short term variation. What made the recent notable violation of 3 sigma at all interesting to me was the apparent pattern of overshoot and overcorrection, more as a way to help me feel comfortable that all would return to trend in time, as it has when this pattern has been seen in the past. |
