I like to think of the valuations as lower-bounds. There are people out there who will pay for discounted income streams, and they put a pretty good floor on p

I like to think of the valuations as lower-bounds. There are people out there who will pay for discounted income streams, and they put a pretty good floor on prices. There are people out there who will pay for book value when prices drop too far, and they provide a (different, weaker) floor on prices. There are people who will pay for brands and memes and so on, and they provide different floors again. And then throw in different discount rates, different projections of the future, different tastes, you get the picture. None of these lower-bounds provides the price. If I have a company that just holds $1 in a bank account at 0% interest, an income-based valuation says it's worth $0, and a book-value based one says it's worth $1, so it's worth at least $1. But if I have a goose that lays $100 bills every day, it's worth a lot more as an income stream than it is in book value (as a goose.) There are some things that provide upper-bounds. When prices are too high, companies should issue stock and (incidentally, not as a goal) drive the price down -- this is "raising money cheaply". But in practice this is usually a pretty weak force. The bitcoin equivalent (mining) might be stronger because there are more market participants on the supply side. Anyway, with interest rates at zero, high savings and lots of spare time and boredom, the income- and meme-based lower-bounds are causing prices to do interesting things.