Further support of the notion that Warren Buffett hasn’t lost his touch comes from Fortune’s Stephen Grandel, who says the study critiquing Buffett’s recent performance has some major flaws.

Grandel says the study used book value to gauge Berkshire Hathaway’s performance — a metric that no longer makes sense for the firm. “When Berkshire was starting out, the company’s book value was mostly made up of Buffett’s publicly traded investments. So using book value made sense. These days, though, Berkshire owns lots of whole private companies, like Dairy Queen, Fruit of the Loom, and See’s Candies,” Grandel writes. “Those are not valued like publicly traded stocks in Berkshire’s book value. As a result, using book value to track how Buffett’s stock picks have performed really doesn’t work anymore.”

He also says that something else happened around the start of the period of underperformance that statistician Salil Mehta highlighted in questioning Buffett’s recent record. “Berkshire bought Burlington Northern, the railroad, an acquisition that added $34 billion in assets to Berkshire’s book value that don’t trade like stocks,” Grandel says. “In fact, railcars, like pretty much everything else, are guaranteed to lose value over time as they get older. That ensures that the book value of Buffett’s investment will look like a loser, even if his actual return on Burlington is great.”

Using Berkshire’s share price, Grandel says, leads to another conclusion about Buffett’s performance. “Buffett’s performance has trailed off” on that basis, he writes. “But so has the stock market’s performance, and inflation. So it’s hard to fault Buffett for that. In a lower-return world, he is still outperforming, by a lot.”