In a recent article for Bloomberg, columnist Nir Kaissar discusses the ongoing debate regarding the challenge of stock picking.

He cites a study conducted by University of Arizona professor Hendrick Bessembinder that found, for the period between 1926 and 2016, “only 4 percent of stocks accounted for all the wealth created,” and “50 of those stocks accounted for 40 percent of that wealth.”

That basket of 50, writes Kaissar, resembled a growth stock portfolio (and included four of the five FANG stocks), and while 7 of those 50 no longer trade, the other 43 “fetch prices befitting the most sought-after companies.” He points out that, given the historically weak performance of growth stocks, it is unusual that so much of the wealth creation was attributed to the 50.

While Bessembinder’s study doesn’t distinguish between growth and value stocks, Kaissar suggests that perhaps, “among growth stocks, the losers more than compensate for the wealth created by the winners.” He shares data provided by value investors Wes Gray and Jack Vogel—who ran simulations using randomly selected growth and value stock portfolios and gauged performance for the period between 1963 and 2013. Vogel and Gray found, Kaissar writes, that the worst-performing value portfolio beat the best-performing growth portfolio, and the distribution of outcomes among growth portfolios was wider than that of value portfolios.

The findings, says Kaissar, suggest that picking growth stocks could be even tougher than picking value. “I’m not suggesting that picking the winners is easy,” Kaissar concludes, “But maybe some winners are easier to spot than others.”