> It's complicated to discuss what "evidence" means for events in the future, because it rapidly heads into deep waters of philosophy, but like the 538 election

> It's complicated to discuss what "evidence" means for events in the future, because it rapidly heads into deep waters of philosophy, but like the 538 election prediction the best we can do about the future is have a probability distribution of outcomes. And as a result there are three big effects: one is low-probability events turning out to actually happen sometimes ("black swan"), another is events you wanted to be uncorrelated turn out to be correlated (which blew up mortgage "tranches"), and the third is that other people's action is market-driving information even if it's irrational; panic selling causes more panic selling, bank runs are real, etc. Markets make point estimates. Derivatives markets make distributional estimates. The information is all there, and if you disagree with it, you can make bets to that effect. But most people won't do that, because they don't actually believe that the market is inefficient. There's very good justification for the present stock market rally, especially if you pay attention to the market's internal structure. The pandemic is driving consolidation, and changing habits. Changes that primarily benefit high scale businesses. The disconnect between the state of the economy and the state of the S&P 500 is driven primarily by this distinction. The market is correct to price in a boom. You are also correct to fear a recession. These things are only mutually exclusive if you mistake the stock market for the broader economy. It's also true that there are major tail risks present. The market is not ignoring them. The market is saying that the expected value, inclusive of those tail risks, is positive. You can see the market pricing those tail risks in if you look at the implied volatility surface of options right now. It's there. It's really important to understand that the fact that a price turns out to be wrong does not make markets inefficient, or even incorrect. Gamblers making statistically correct bets are wrong all the time. That's just the way uncertainty works in the world. There are people arbitraging uncertainty at all points of the distribution, from the tail on up. It's true that markets are less efficient around low probability events. But it's also true that they quickly learn from their mistakes