Ask HN: Are bubbles caused by inaccurate data?

So, yes, but in a more general way than you're describing. It's more like "Uber will displace car ownership, and the automobile industry is worth at least a $trillion, so $40B is actually undervaluing Uber", and then if it turns out that Uber doesn't actually displace car ownership the whole valuation equation falls apart. Or sometimes it's using assumptions in places where they don't hold, eg. "Google makes $X per search result shown, and Facebook has more pageviews than Google, therefore Facebook should be worth more than Google" even though people search when they're further down the purchasing funnel than looking at social media and so search clicks are often worth a lot more.

Anyway, capitalism is naturally self-correcting in this regard. People who build their businesses on invalid assumptions go out of business; what's left are the people whose assumptions turned out to be right. All bubbles eventually pop. A lot of people view bubbles as a bad thing, but they're really part of the normal operation of capitalism. If you believe that somebody else is being an idiot, take the other side of the trade.

I think the inaccuracies naturally swing in both directions (undervaluing certain aspects of a business, overvaluing others), so in a calm marketplace, this does not cause a bubble.

But when the natural level of inaccuracies is combined with human emotions like greed and fear, inaccuracies can start swinging in one direction or the other more prominently (and this trend manifests itself in the form of a bubble, or a recession).