The phrase reversion to the mean refers to a statistical concept that high and low prices are temporary and a price will tend to go back to its average over time. To trade the concept of mean reversion means that you follow this simple process:

1. Find an average price over some past period.

2. Figure out the high-low range.

3. Buy when the price has deviated to the low side of the range and sell when it gets to the high side.

Does this sound too good to be true? Well, it is. Mean-reversion trading ideas have the appearance of applying basic statistical concepts to securities prices to derive trading rules, but mean-reversion trading faces severe obstacles:

• Determine the ideal lookback period to determine the average. Say, for example, that Blue Widget stock over the past two years averaged \$20 — but that \$20 average incorporates a few abnormal prices like \$1 and \$40. An average can disguise multiple deviations that have already occurred.

• Securities prices are not actually normally distributed — they just look that way sometimes. In technical analysis, your primary goal is to determine whether your security exhibits a price trend. You also want to know how strong the trend is and whether it might be ending soon. To accept the assumption that the distribution of prices will be normal is the same thing as saying that you know in advance where the price trend will end — at or near the price represented by the average, plus one standard deviation. If the price goes higher than the price that one standard deviation dictates, the trading rule embedded in the mean-reversion trading technique would have you sell.

You consider the security “overpriced” on a statistical basis. And yet you can’t be sure that the other traders in the market performed exactly the same analysis that you did. Even if they’re using the mean-reversion concept, maybe they used a different lookback period to calculate the average. Because the other traders in this security don’t see the security as overpriced, they may keep buying, and buying, and buying — pushing the price to the equivalent of the guy in the room standing 7 feet 10 inches. The opposite is true, too. The mean-reversion process would not identify the situation in which the price keeps going to zero.