The Short Squeeze, 2021

A crowded short, a coordinated crowd, and the moment the plumbing became the story. The clearest live demonstration of why borrow costs and clearing margin are not footnotes.

What happened

  • 2020 — a struggling US retailer becomes one of the most heavily shorted stocks on the market. Reported short interest exceeds 100% of the free float — possible because a borrowed share can be sold and borrowed again by the next buyer.
  • January 2021 — retail investors coordinating on social media buy the stock and, more importantly, short-dated call options.
  • Late January 2021 — the price rises from single digits to a peak of several hundred dollars. Several short-selling funds report very large losses; at least one requires an emergency capital injection.
  • 28 January 2021 — multiple retail brokers restrict opening trades in the affected stocks. The price falls sharply that day. Public reaction is immediate and furious.
  • The stated reason — clearing houses raised margin requirements on the brokers, in some cases by an order of magnitude overnight. Restricting new positions reduced the deposit the broker had to post.
  • February 2021 onwards — prices fall substantially from the peak, though the stock does not return to its starting level. Regulators and legislators hold hearings; no market-manipulation charges follow against the retail participants.

The mechanism

Price and borrow cost rise together, and that is the squeeze. Short sellers do not choose when to cover — margin does, and every forced buy lifts the price that triggers the next one.
Covering beginsPeakShare priceCost to borrowTime (days)Price and borrow cost
  • A short position has unlimited loss and a hard funding constraint. As the price rises, the short must post more margin. When they cannot, they buy back — pushing the price higher and forcing the next short to cover. The spiral is mechanical, not psychological.
  • The borrow cost is the second engine. Fees on the hardest-to-borrow names reached hundreds of percent annualised. As the securities lending page sets out, that fee is a real carrying cost that makes waiting unaffordable independently of the price move.
  • The gamma squeeze amplified it. Buying call options forces the market makers who sold them to hedge by buying stock. As the price rises, their delta rises, forcing more buying — the gamma mechanism running in reverse against the seller. Short-dated out-of-the-money calls give the most leverage per dollar for exactly this reason.
  • The clearing system decided the ending. Trades settle over a period during which the clearing house carries the risk, and it demands collateral scaled to volatility and concentration. A vertical price move in a concentrated name produces an enormous overnight demand on the broker. The restriction was a funding constraint at the broker, not a favour to hedge funds — and the fact that almost nobody could tell the difference in real time is itself part of the case.
  • Short interest above 100% is not fraud. It follows directly from re-lending: A lends to B who sells to C, whose broker lends the same share again. The plumbing permits it, and it makes a crowded short far more crowded than it looks.

What it teaches

  • Crowding is a risk factor. A trade's popularity determines what happens when it reverses. High short interest plus expensive borrow plus a small float is a structurally fragile configuration, whatever the fundamental view.
  • Being right about a company says nothing about surviving the position. Several of the shorts were correct about the business and still lost catastrophically. The same distinction as LTCM and Metallgesellschaft: analysis and solvency are separate tests.
  • Options are not a sidecar to the stock market. Dealer hedging flows can dominate price formation in a small float. Anyone reading the equity tape alone was watching half the market.
  • Settlement and clearing are load-bearing. The most consequential moment of the episode was a margin call inside the plumbing — infrastructure most participants had never thought about and could not observe.
  • Payment for order flow entered the mainstream argument here. The brokers' revenue model became the public explanation for the restriction, and understanding why that explanation was mostly wrong requires understanding the market's structure. The episode's lasting effect was to make retail participants ask how their orders are actually handled — which is a good outcome from a bad week.