Part II: The Markets and Players
Chapter 10
Exchanges, Brokers and the Retail Boom
Early on the morning of Thursday 28 January 2021, Robinhood's clearing house told it how much collateral it would need that day. Robinhood had become self-clearing in 2018, so it, not a third-party clearer, posted the deposits that the National Securities Clearing Corporation (NSCC) requires to cover the risk of customers' unsettled stock purchases, then settled two business days after the trade. Its customers had spent the previous days buying GameStop, AMC and a handful of other heavily shorted stocks, and short-dated calls on them, as retail traders coordinating on Reddit's WallStreetBets drove GameStop from $43 to $72 in three hours on 22 January and on towards an intraday peak of about $483 on the 28th. The NSCC's requirement for Robinhood rose to about $3.7 billion, against about $696 million on deposit: a call of roughly $3 billion (written congressional testimony, as summarised in the companion). Robinhood restricted its customers to closing positions in GameStop, AMC and about a dozen other stocks. The NSCC then waived an excess-capital premium, cutting the requirement to about $1.4 billion; Robinhood raised about $1 billion that night and about $2.4 billion more within days, about $3.4 billion in all, more than its total prior funding (reported).
On the other side of the trade was Melvin Capital, a New York long/short equity fund that Gabe Plotkin, one of Steven Cohen's best-known portfolio managers at SAC Capital, had launched in late 2014. Melvin had compounded at roughly 30% a year through 2020 and started 2021 with about $12.5 billion (reported). It had been short GameStop since 2014; the SEC's staff later measured GameStop's short interest at 122.97% of its float at the peak. On 25 January Citadel and its partners (about $2 billion) and Point72 ($750 million) put $2.75 billion into Melvin for non-controlling revenue shares; Melvin closed its GameStop short on 26 January; it lost about 53% in January (Wall Street Journal, reported) or 55% (Bloomberg, reported), about 39% for 2021, and announced on 18 May 2022 that it would return its clients' money (companion's profile of Melvin Capital).
The collision produced a theory that the evidence does not support. Because Citadel Securities bought much of Robinhood's order flow and Citadel had just invested in Melvin, many customers concluded that the restrictions protected hedge funds; Vlad Tenev of Robinhood and Ken Griffin of Citadel testified to the House Financial Services Committee on 18 February 2021 that Citadel had played no role. A second popular account held that options market makers' hedging had driven the stock, a "gamma squeeze". The SEC staff's report of 14 October 2021 said it did not find evidence of a gamma squeeze. What the documents do show is plainer: under T+2 settlement, a concentrated, volatile, one-way book of customer purchases raised a clearing charge overnight, and a broker with little market risk of its own found that its balance sheet decided what its customers could trade.
The preview ends here. The full chapter is in the book.
In this chapter
- 10.1From May Day to the free app
- 10.2Who pays when trading is free
- 10.3Underneath the apps: exchanges and the clearing house
- 10.4A worked example: why one-way flow multiplies a broker's deposit
- 10.5Gamification and the regulators
- 10.6What the outcomes evidence says
- 10.7Beyond the United States
- 10.8How an expert reads a broker or an exchange