Is it possible to figure out when a bubble is going to end? An opinion piece by John Authers in Bloomberg says even though we can’t know for sure, it doesn’t stop anyone from trying, and points to the last few doses of speculation to give us some clues.

During the dot-com boom, the market narrowed sharply in its final stage. The S&P 500—excluding technology—fell more than 10% while tech rose more than 50% in the last 5 months of that bubble. That differed from earlier in the boom, when tech stocks rallied without getting in the way of the rest of the market.

Retail involvement can also be an issue. Typically, individual investors flood in at the end of a bubble and then bear the brunt when it bursts. But some analysts predict that there’s still a final paroxysm of buying to come. And in regards to valuation, the idea that an expensive market can always become more expensive is a problem that can still rear its ugly head. A final splurge of speculation could come with more liquidity, as history suggests that a bubble burst comes when cheap money is finally removed.

Trying to time the top of the bubble seems to be reserved for those with money they can afford to lose, and only those with strong constitutions will look to short low-quality small-caps. The majority of investors with a longer time horizon can look through this phase completely by putting very little money in FANG-type stocks. But money managers who are judged by their performance in the short-term will likely move the market toward a collection of assets that will do well in an atmosphere of rising bond yields.