> Doesn't this all give more meaning to the average than is its due? Of course
> a cyclical series will always return to its average, if it didn't, that
> wouldn't be the average. It's a statistical result that seems to be given a
> great deal of concrete importance. This isn't to say it can't be a useful
> analytical touchpoint, but oughtn't we keep its construction in perspective?
Yeah, exactly. You need a good analytical reason for treating an historical average as a real point of attraction, rather than just an average that might itself get dragged around by changes in the underlying forces at work. I've been making this argument in debates with people here who argue that Australian housing costs or house prices _must_ at some point return to their historical relationship with household incomes. I don't see any a priori reason why. The stats don't go back very far, but there's been a rising trend as far back as you look so the historical average seems a silly thing to latch on to. In places with longer runs of historical statistics we've seen secular changes in the ratio of housing costs to incomes.
Things are a little different with share prices because you do have analytical reasons to think that expected returns (including valuations of liquidity and risk) would not get out of line with those from alternative investments. But it's possible for the average return on all kinds of investments to rise and fall, secularly and cyclically. Anyway, the historical experience of what happens to asset prices when central banks attempt to put a floor under collapses doesn't go back very far.
Cheers, Mike scandalum.wordpress.com