In an article in The Wall Street Journal, Jason Zweig examines his mistaken prediction from 2010 that the stock market wouldn’t return more than 6%, and what can be learned from his error, which Zweig posits stemmed from relying too religiously on the past in estimating the future. He also fixated on Robert Shiller’s CAPE (cyclically adjusted price/earnings ratio) gauge, which put stocks at a price 20.3 times their adjusted long-term profits back at the start of 2010, and Zweig believed their returns had to be lower over the following 10 years. But instead, the price for those earnings boomed.

Currently, stocks are far above their records, at a CAPE 37.1 times adjusted earnings, with many analysts and managers cautioning that the market is in a super-bubble with the biggest drop of wealth in U.S. history on the horizon. But many of them are looking at past bubbles in comparison, such as the Dot-com bust, and “[y]ou can’t just lay the past onto the present and declare that it gives you an unambiguous basis for a drastic decision,” Zweig writes. It’s impossible to know whether the mean valuation will keep rising if corporate taxes remain low, or if interest rates stay moderate, or if new tech continues to disrupt society.

Even Shiller tells the Journal that CAPE offers “a limited number of observations,” adding that “we’re constantly moving into a new era.” Even with the CAPE so high, he doesn’t advise selling everything; perhaps trim your exposure to U.S. stocks, weight more towards international stocks, or favor less-expensive sectors like healthcare, financials, and consumer essentials.

Trimming stocks as they go up always makes sense, Zweig writes in conclusion, but getting rid of your shares because they’ve skyrocketed doesn’t because if anything is certain in this market, it’s that everything is uncertain.