Predicting market crashes is like predicting the end of the world. When it actually happens (in both cases), there will be some subset of people whose publishe

Predicting market crashes is like predicting the end of the world. When it actually happens (in both cases), there will be some subset of people whose published prediction is close to or even exactly matches the date when it finally did happen, and those people will gloat about it. But for every one of those people there are thousands, possibly millions or more others whose prediction dates came and went with no catastrophe occurring. Given the huge number of people who publish their predictions, then, we have to ask which of these is more likely: 1. The people who got the date right -- either exact, or within a certain small window of the actual date -- had some genuinely deep insight missed by others, which allowed them and only them to make a correct prediction, or 2. Given a large enough segment of people making predictions, some subset of them will have their predicted date match the actual date purely by chance. For example, predicting market trends five years out only gives five years in which something can happen, or twenty quarters, or sixty months, or 260 weeks, or 1,826 days to peg your predictions onto. If the number of people making predictions is large enough, all dates are likely be covered by at least one prediction.