Sometimes it is "Different This Time"

I just read Peter Bernstein's Against the Gods: The Remarkable Story of Risk. It is an interesting read though I feel he often wanted to include everything he'd researched for the book, irrespective of relevance rather than explaining in more depth the ideas he covered. For example, he never really explained why and how his discussion of game theory was really relevant. I understood the discussions of stuff I was familiar with because I was familiar with those things. I only got the most superficial notion about the other topics I was less familiar with. Anyway, one key story that Bernstein repeated more than once stayed with me. This seems to have been a formative event in his own investing career. He quotes a Gilbert Burke that before 1958:

"It has been practically an article of faith in the U.S. that good stocks must yield more income than good bonds, and that when they do not, their prices must promptly fall"

Since 1958, stocks have yielded less than bonds, while from 1870 to 1958 there were only rare occasions when they did not yield more. In 1958, Bernstein's older colleagues waited for stock prices to fall. But they never did. The bull market continued into the 1960s. Today, I often read statements like "there has been no bear market so short, so this one must last longer" (what about 1990-1?). This is just one topic people like to pronounce on. There is no theory behind this statement and only a very small sample by statistical standards to support it empirically. Maybe this time, things are different. Before 1958 people assumed that as stocks are riskier than bonds they must yield more. Since 1958, more persistent inflation and increases in stock buybacks among other things have changed the behavior of both bond and stock yields. The lessons are to look for what might have changed and not to rely on small samples to make strong predictions in the absence of theory.