> What you describe won't account for a tumour growing in a dictator's head that makes him decide to press the big nuke button one day.

Yes, the market consensus can only reflect the information that available to (at least some) market participants.

Of course, the market mechanisms also give people incentives to acquire that information. Because any informational advantage gets rewarded with money.

> It won't account for a volcano erupting and burying a society in the ashes as happened in Pompeii, or the global equivalent.

It can account for the best estimates we have for the probability of that happening.

Again, I'm not suggesting that market estimates are perfect, nor even necessarily any good. I'm just saying that market estimates tend to be at least as good as any other estimates we have. Ie market estimates tend to be the best ones available.