Streetwise Professor

The Treasury basis trade continues to be in the news, with one of the biggest basis traders–Citadel’s Ken Griffin–complaining that the SEC should regulate hedge fund basis trading, but should instead make sure that banks aren’t supplying too much leverage to . . . well, hedge funds mostly. This seems like a raising rival’s costs gambit. No doubt restrictions on leverage would hit Griffin’s competitors harder, whereas SEC regulation might have a more even impact.

Regardless, one question that hasn’t been asked in all the to-ing and fro-ing about Treasury basis trades is why they exist at all, let alone why they get so big. This graph (courtesy of FTAlphaville, based on CFTC data) provides a major clue:

Note the mirror image between leveraged funds (mainly hedged funds) and asset managers (ostensibly non-leveraged funds–the reason for the “ostensibly” will become clear shortly).

To the extent that hedge funds’ short positioning reflects basis trades, the graph suggests the following. Hedge funds take a leveraged market neutral position, buying bonds, funding them via repo, and selling futures. Futures are in zero net supply: the graph shows that the longs on the other side of the hedge funds’ futures short are asset managers.

Most asset managers do not, and in some cases even cannot, take leverage directly. So for example they are constrained in their ability to just buy Treasuries with borrowed money (e.g., via repo). But the basis trade allows them to lever up via futures. So in some sense, the basis trade is just an additional link in a chain of intermediation. Laundering leverage, if you will.

(A more complete picture might add swap dealers to the picture. Some managed money, such as leveraged ETFs, enter into swaps with dealer banks. The dealer banks in turn can hedge by taking offsetting futures positions.)

The hedge funds expect to earn a small margin on the trade–on average, though there is risk. The market is pretty competitive, so to a first approximation that margin (the difference between the actual futures price and the theoretical futures price derived from bond prices, bond vols and correlations, and repo rates) equals hedge funds’ marginal cost of supplying this intermediation. The asset managers on the long side of the futures trade are willing to pay “too high” a futures price (relative to bond prices) because this is a cheaper way of achieving a leverage target than via the available alternatives.

The March 2020 experience shows that the basis trade can be a fragile one that creates some systemic risk: this is why regulators are concerned about basis trades now, to Ken Griffin’s chagrin. Providing this leverage intermediation/laundering creates tail risks for the hedge funds that do so. This raises the question of whether there are regulatory constraints that inefficiently constrain the ability of asset managers to take leverage more directly, rather than via a longer dealer (or money market) to hedge fund to asset manager chain. If so, such constraints could give rise to unnecessary (systemic) risks.

If regulators are concerned about the systemic risks in basis trades, they should take a systemic approach–and understand more fully why basis trades exist in the first place, and why they have periodically become so large. Looking at individual links in the chain (hedge funds, or by Griffin’s lights, banks) can be misleading because it begs the question of why the chain exists in the first place. The link that is driving the process is likely the one that has escaped discussion so far–the asset managers at the end of the chain. Why do they want leverage and why is the basis trade the most cost effective way of supplying a lot of it? Could it be the most cost effective because other, more directly intermediated sources of leverage are unduly expensive because of regulatory or institutional constraints? Definitely worth regulators’ attention.