The thing about reversion to the mean, powayseller, is that you can always look back and say “see – it reverted to the mean.”
What you can do is reliably predict how and when it will do so.
One reason is, the mean changes as today’s data is added to the stats. One could say that in the long term, the mean always reverts to the data as much as the data reverts to the mean.
Second, there is an issue of time. Do you analyze the mean for the year prior to the bubble? 10 years? 100 years? 1000 years? What is the “time constant” of this thing? We don’t really know.
We could revert to the mean by starting on a new housing bubble next year, only to result in an even greater crash 5 years later than we thought. It could also be a soft-landing for 20 years. That would be a reversion to the mean, but at a different rate than you expect, and to a different level.
Perhaps this bubble is riding on top of a bust with a 40 year cycle, and part of what we have seen is a correction of that 40 year bust.
Lastly, reversion to the mean in many cases is a myth. Some would say that when you flip a coin ten times and it comes up heads 10 times, it is now more likely to come up tails so that the statistics regress to the mean. I’d say you have a rigged coin.
There is a reason why on all investment documents, it one is legally obligated to write “the past is no indication of future performance.”
Reversion to the mean is simply not the right analysis to do. I’m done w/ this thread.
