I think the market hit irrational exuberance 2 or 3 years ago, and it has just kept on going. Anybody could tell you that the market was way too high in 1996 or 1997, but you'd still have been better off to have kept doing what you were doing. You could possibly just maybe have made a case for switching to bonds at the very top in 1999 or 2000, but bonds are ridiculously low at the moment.
This is a fascinating post because even in trying to identify the problem of predicting market movements, it incorporates the error of backward-looking bias (i.e., incorporating what you know today into what you would have known then).
Your statement is that anybody could tell you that the market was way too high in 1996 or 1997.
You think that because it seemed like it was high-flying back then then crashed spectacularly in 2001-2002.
Do you know what the S&P 500 was in January 1996? It was 600. Seriously.
It was never that low again in history. And that's not counting dividends.
It seems like it was high flying only because you know now that a significant crash was coming. But there were still several years of build up before that crash.
This is the key point here. Right now the market is up 40% from February 2016--less than 2 years ago. I bet most people would say now that a 30-35% crash would bring stocks back to a very good price. But that just takes you back to February 2016. How many people do you think on this forum were saying in February 2016 that the market was a great deal, phenomenal prices, etc.? Zero.
We only see trends in retrospect. But we believe we're good at it, which is why there are so many threads "calling the top," etc. Someone will get it right just based on pure statistics (if there are enough bets, one is bound to be right). But that doesn't make that person good at predicting the market. With very, very few exceptions in history, it just makes them the lottery winner that happened to get the call right on that particular occasion.