The basis trade meets a haircut

The Treasury basis trade earns a very small spread on a very large position. That is not a criticism of it; it is the design. The trade captures the gap between a Treasury security and the futures contract that references it, and the gap is measured in fractions of a percent.

A spread that thin only produces a return worth having if it is multiplied, which means borrowing. The position is financed in the repo market, rolled overnight, and levered many times over. The leverage is not incidental to the strategy. It is the strategy.

Supervisors have been measuring how much of the market this now accounts for. Research published by the Federal Reserve decomposes hedge funds' Treasury exposures and identifies relative-value strategies of this kind as the dominant component.

The concern is not that the trade is wrong. In ordinary conditions it is stabilising: it links the cash and futures markets and keeps the two prices honest. The concern is what happens when the financing stops being available at the price everyone assumed.

That is the mechanism worth understanding. A position financed overnight has to be refinanced every morning. If lenders raise the haircut, meaning they demand more collateral for the same loan, every holder of the trade needs cash or a smaller position on the same day.

Selling Treasuries to raise that cash pushes prices down, which widens the loss, which raises the collateral demand again. The unwind is fast because the funding is short, and it is correlated because the participants hold versions of the same position.

The policy question follows directly from the plumbing. The Fed's own assessment of leverage in the financial sector treats non-bank leverage as a live vulnerability rather than a theoretical one.

A minimum haircut is the instrument under discussion internationally. Requiring lenders to hold back a small percentage on repo against Treasuries would cap how far the trade can be levered, and it would do so continuously rather than in a crisis.

The objection from the funds is that the trade improves market functioning and that raising its cost widens the spread between cash and futures, which someone eventually pays. That is true and it is the point: the current spread is narrow partly because the leverage behind it is cheap.

Central clearing is arriving on a separate track and addresses a different part of the problem. Clearing standardises margin and makes exposures visible to a central counterparty, which helps in an unwind without limiting how large the position can get beforehand.

Both measures point the same way. A strategy this large financed this short is a public matter whether or not it is a public institution, and the argument now is about the price of the financing rather than the legitimacy of the trade.