Robinhood Was Behind Phantom Surge in Berkshire Hathaway Trade Volume, Study Finds

The study suggests that volume data for many other stocks were inflated during the pandemic. Its authors blamed a “misguided” Financial Industry Regulatory Authority rule for how brokers should report fractional trades.

“The Finra reporting rule for fractional trading has created significant distortions,” wrote the authors, who are professors at University of California, Berkeley, Columbia University and Cornell University.

Robinhood declined to comment. A Finra spokesman said the regulator was aware of the concerns about its rule. “Finra is already actively working on the issue and is engaged in ongoing discussions with firms and regulators,” he said.

Fractional trading allows investors to buy or sell slices of a share, instead of the whole thing. The feature has gained popularity in recent years as Robinhood and other brokers have rolled it out to their customers.

Although fractional trading is available in many stocks, it is especially popular for stocks with high share prices, where it can be prohibitively expensive for a small investor to buy an entire share. And no stock is pricier than class A shares of Berkshire because of Mr. Buffett’s decadeslong refusal to split the stock. Known by the ticker BRK.A, they closed Tuesday at $425,110 a share. Most trading in Berkshire takes place in cheaper class B shares.

Because the class A shares are so expensive, they have long been thinly traded. In 2019 and 2020, just 359 of them changed hands on an average day, translating to about $109 million worth of daily trading activity, according to

FactSet.

But reported volumes of the class A shares shot higher in mid-February 2021 and have stayed elevated ever since. From Feb. 18, 2021, through last week, average daily volume was nearly 1,900 shares, or more than $800 million in dollar terms, FactSet data shows.

The volume surge prompted speculation in financial media outlets and Reddit forums. Some articles suggested that a mystery buyer was amassing class A shares. At Berkshire Hathaway itself, Mr. Buffett was puzzled by the reported volume numbers and believed they were wrong, but didn’t know the reason why, a person close to the company said.

The reason, according to the new study, is that Robinhood began to comply with an obscure part of Finra’s rule on trade reporting.

When firms execute private, off-exchange stock trades, they are required to report them to huge databases called trade reporting facilities. When such trades involve fractional amounts of less than one share, Finra requires firms to report the trades as though they were for one whole share. Thus, a small investor’s purchase of just 1/100th of a share would count as a purchase of one whole share. The authors call it the “Rounding Up” rule.

Since the trade reporting facilities are a key source of publicly available data on buying and selling activity, investors end up seeing inflated numbers for trading volume, the authors write. Such distortions are bigger for stocks with a large degree of fractional trading.

The Finra spokesman noted that most systems for reporting U.S. stock trades weren’t designed to handle the entry of trades with a fractional quantity of shares. “The Finra guidance on trade reporting needs to be understood in that context,” he said.

Robinhood launched fractional trading in late 2019, but initially it didn’t report such trades to Finra’s trade reporting facility, the brokerage disclosed earlier this year. “Since then, Finra has informed us that such trades should be reported. As a result, we began reporting fractional shares in January 2021,” Robinhood said in a securities filing.

The three professors write that Robinhood began reporting Berkshire class A trades on Feb. 12, 2021, and stepped up its reporting of the shares over the following week. They pinpointed the trades reported by Robinhood by identifying its electronic “signature”—a slight delay, measured in tiny fractions of a second, in the time it takes for trades to be broadcast over public data feeds. The delay associated with Robinhood’s trades differs from that of other brokers, making it possible to distinguish which trades originated with Robinhood customers.

Using the same technique, the authors found that another brokerage started reporting trades in Berkshire class A shares last year: DriveWealth LLC. Though it isn’t a household name, DriveWealth handles stock trades for a variety of investing apps including

Block Inc.’s

SQ 4.50%

Cash App and Revolut. It was also an early player in fractional trading, saying on its website that it “helped pioneer fractional share trading in 2015.”

Starting in October, hundreds of trades in Berkshire class A shares with DriveWealth’s signature started appearing each day on public data feeds, the authors found. That fueled a further increase in reported trading volumes. Since then, some 95% of the trades reported in Berkshire class A shares have borne either Robinhood’s or DriveWealth’s signatures—evidence that the vast majority of reported volume in the shares is being caused by small investors’ fractional trades, the authors write.

DriveWealth declined to comment.

The authors of the study called for Finra’s rule to be revised. Until that happens, they warned that reported trading volumes for many stocks—not just Berkshire class A shares—may be distorted.

“Given its unusually high stock price, BRK.A is in many ways simply an exaggerated example of how Finra’s rule along with the rise in fractional trading creates phantom, non-existent trading volume across all stocks,” they write.

Write to Alexander Osipovich at alexander.osipovich@dowjones.com