Sharks or Dolphins? Part 2

The basis trade is here to stay, because the authorities created it

In the first part of this two part look at the basis trade I tried to quantify the spread of the trade to non-US markets. The problem with concentrating on the risks from hedge fund leverage in the cash/futures basis is it distracts from the fundamental reasons for the emergence of the trade. Today I attempt to look at the wider dynamics behind the trade. These originate from long-running institutional changes in the structure of capital markets by fiscal, monetary authorities and regulatory changes to ‘improve’ market functioning. The authorities created conditions for the cash/futures basis trade and the Treasury market is now unable, probably, to function without it. Whether the same can be said for other government bond markets is a question for another time.

The emergence of the basis trade is based on five related factors:

1. The Treasury market outgrew its traditional intermediation model.

2. Credit overweighting: regulatory encouragement for banks to withdraw from corporate financing encouraged asset managers to allocate more capital to credit products for higher returns and, accordingly, less to Treasuries.

3. In turn, futures offer a cost efficient, liquid means of acquiring and adjusting duration as the liquidity of the cash market stagnated.

4. A flat/inverted yield curve offers as much or more return from overnight funding as from duration investment. This encouraged money-market fund inflows who have made increased use of repo a large portion of which is indirectly used as finance for hedge funds trades, including basis trade.

5. The growth of very large leveraged funds has increased the efficiency of netting and so increased potential leverage in the trade.

With the exception of the first, all of these factors are the direct or indirect consequence of reforms made to ‘improve the stability’ of the financial markets. In combination they created a long-running ‘liquidity premium’ in the futures markets. Ironically, this is the antithesis of the ‘illiquidity premium’ claimed by private equity and private credit which the same asset managers have been keen to pursue. Hedge funds availed themselves of this liquidity premium, aided by the expansion of money-market funds who turned increasingly to sponsored repo. Without this intermediation, it is likely the well-intentioned reforms would have ensured massive increase in issuance by governments was much more disruptive to global bond markets.

dolphins and sharks

Accordingly, when considering the total risks embedded in the basis trade, regulators should add in the risks created by these factors. This has rarely been done.

Treasuries have Outgrown Dealers

The IMF notes in the April 2025 Global Financial Stability Report: “the Treasury market had outgrown dealers: its size is now five times dealers’ balance sheets, a significant increase from just one-and-a-half times around 20 years ago”. That seems to be overstating it slightly, as it seems the IMF is using total Federal debt, not Federal debt available to the market. Nevertheless, their main point stands; the Treasury market is massively larger compared to dealer balance sheets.

The biggest driver for this change was the massive increase in fiscal funding, encouraged by low interest rates, search for safe-haven assets and expanding government spending.

Not only has the Treasury market increased, but dealer capacity has been restricted. Regulatory changes post-GFC, including Dodd-Frank and Basel III, increased bank and dealer capital charges which placed a constraint on dealer balance sheet allocation to market making in Treasuries. The Supplementary Leverage Ratio (“SLR”) and the “enhanced SLR” (“eSLR”) created disincentives for large bank dealers to hold low-returning assets such as U.S. Treasuries by assigning them an equal risk-weighting with far riskier and higher-returning assets. While there is some argument as to exactly how binding this constraint can be, there is no argument that it did increase disincentives to act as intermediaries.

Proposed adjustments to the SLR may help, but are very unlikely fully to offset restrictions. Treasury debt has grown from around $8.3 trillion in 2012 to nearly $30 trillion in March 2025. Over that period, dealer holdings of Treasuries have remained almost unchanged $760 billion in early 2013 to $770 billion in March 2025. Similarly, financing of Treasury holdings by dealers has grown at a slower pace than overall stock of Federal debt. For the Treasury market to continue to function efficiently either Asset Managers needed to become much more involved in primary and secondary markets, or an intermediary was required to provide intermediation previously provided by dealers. Asset Managers generally chose to expand fixed income allocations towards higher return credit products rather than involvement in the relatively lower return Treasury market. That role of intermediation was taken by hedge funds.

Risks are created by crowding when exits from investments are sought simulteaously, as the market no longer has guaranteed market making factility in line with the overall size of outstanding stock. This may have been one of the explanations for the sharp rise in yields following ‘Liberation Day’.

Asset manager Credit overweight and duration management

The effect of issuance and QT on cash prices was simultaneously met with an increased demand for futures to manage duration and liquidity by Asset Managers. In this article I use the term ‘Asset Manager’ in similar terms to that used by the CFTC to cover ‘pension funds, endowments, insurance companies, mutual funds, and other portfolio/investment managers who primarily serve institutional clients.’

There are a number of factors that propelled Asset Managers towards an increased use of futures.

1. Asset Managers increased their allocations to credit products (including CLOs and private credit). Credit products are inherently less liquid than Treasury bond futures yet require active management of interest rate risk and duration. Futures allow managers to hedge and adjust duration smoothly, without the need to trade large amounts of cash bonds. The rise in interest rates increased the sensitivity to duration. Futures offer the ability to finesse duration risk quickly and cheaply. For at least a decade, the liquidity of the bond futures market has matched and usually exceeded the liquidity of the cash market. This is especially important during periods of stress, when futures markets show consistently better liquidity than cash markets.

2. The margin requirements of futures permit investment allocations to increase in higher return products.

3. Often repo financing is included as an interest expense while futures are not. As a consequence, some large funds report they have reduced repo and switched to Treasury futures instead.

4. Finally, money managers have shifted towards central clearing, encouraged by reduced counterparty credit risk, increased netting efficiency, and in some cases potential capital relief. This naturally encourages the use of futures with a well-developed central clearing mechanism in addition to capital efficiency.

The combined cost of all these developments is that Asset Managers ensure futures are reliably ‘rich’ compared to their cash Treasury counterparts.

The risks attached to these developments were reflected in the ‘LDI Crisis’ in UK Gilt market in 2022. The rise in margin requirements for LDI products forced pension funds to liquidate their Gilt holdings, forcing a vicious cycle of selling and increasing margin which ultimately led to the intervention of the Bank of England. Effectively, pension funds had become ‘leveraged’ and exposed to higher margins. An analagous development has occurred in Asset Managers usage of futures and higher allocations to credit products.

Money market funds financing HFs

Assets undermanagement in money-market funds have risen from $5 trillion in 2022 (prior to Fed rate rises) to over $7 trillion today (June 2025), as savers were increasingly attracted to higher rates offered. This increase has been accompanied by increased investment by MMFs into repo, which now stands at around $3 trillion. In the last 2 years, much of the repo used by MMFs has switched from the Fed ONRRP to repo that indirectly faces hedge funds, underpinning the finance for the basis trade.

Much of this switch to repo reflects an evolutionary change to deal with large increases in bond issuance, and reduced direct involvement of broker dealers, together with regulatory proscriptions on how money market funds could operate - another example of unforeseen regulatory reform.

Another key development was ‘sponsored repo’. As hedge funds increased positions financed by repo the new mechanism emerged which intermediated repo via dealers (mostly through FICC) with money-market funds. The BIS had a nice graphic explaining the process.

Traditional repo requires the dealer to to allocate capital against the repo exposure. In a ‘sponsored repo’ transaction, the dealer “still provides a guarantee to FICC in respect of all obligations of the sponsored counterparties, but it offsets the transactions on its balance sheet. This allows dealers to economise on capital, thus building larger leveraged exposures.”

The result is a large repo funding trade between money market funds and hedge funds, which is financing the basis trade. Another way to view this is via the growth in daily volumes of DVP repo, which follows a similar path in similar quantities. The following is an up-to-date version of the BIS chart in the left panel above.

The innovation in ‘sponsored repo’ is recent. FICC launched its Sponsored Repo service in 2017. It has become a crucial part of the financing of cash/futures basis.

Profitability of the cash/futures basis

For the basis trade to function there needs to be a profitable incentive for funds to remain active. This may come and go. An increase in Treasury issuance combined with Quantitative Tightening probably aided the reduction of relative prices of cash bonds compared to futures. The chart below shows estimates of the profitability of the cash/futures basis trade by the FRBNY & JPM. Notice the periods of highest profitability broadly coincide with balance sheet reduction (QT) by the Federal Reserve.

It remains an open question whether an end to QT will signal an end to the cash/futures basis trade. I would suggest the important role of cash/futures basis trading has assumed in intermediating Treasury issuance probably REQUIRES a permanent premium in futures markets. The trade is likely to survive an end to QT.

Innovations in position accounting

Netting and cross-margining practices reduce capital requirements. This applies mostly to the repo leg of the trade. The short futures position requires initial and variation margin posted to the clearinghouse which treats the short futures leg held by hedge funds as an outright directional position. Initial margin is counted as a capital cost in trading, whereas variation margin is usually managed separately as part of operational cash management.

Much of the literature assumes an initial margin of $2000 per contract. In practice, initial margin varies significantly across different contracts.

If the basis trade is part of a broader portfolio of negatively correlated positions, the prime broker to the hedge fund may apply minimal haircuts on repo financing. This can lead to more efficient capital allocation and favours large hedge funds with multiple overlapping positions. This is supported by Fed analysis which shows “the top 50 funds ranked by gross UST exposure – out of a universe of 2069 QHFs as of Q4 2022 – have consistently held about 80 percent of total QHFs' UST exposures since the inception of Form PF in 2012.”

The Managed Futures Association (MFA) suggests “netting and cross-product margining signifcantly contribute to the low repo haircuts hedge funds often obtain on their repo borrowing in the context of the basis trade.” The Bank of England estimated “haircuts on repo transactions backed by US Treasury securities are at or below 0.5%.”It has been suggested that some positions effectively pay zero haircut.

Based on these assumptions, the effective leverage of 40:1 can be achieved and may rise to 50:1 if repo haircuts fall below the 0.5% level or much higher if applied to the shorter dated futures with lower initial margin.

Even this simplistic calculation reveals that leverage below 40:1 is unlikely to provide sufficiently high risk/adjusted returns for an average annualised basis below 30bps. Given the size of the positions, leverage employed is probably well above 40:1, on average.

The risks from hedge fund positions in cash/futures basis are well-known. Changes to financing and margin costs, availability of repo can all impact expected returns swiftly and dramatically, leading to reductions in holdings and adding to volatility at a time of elevated stress. These risks are well flagged and monitored by authorities and may lead to the provision of further backstops, aiding the risk/return characteristics of the trade.

Conclusion

In summary, rather than seeking to limit the basis trade, regulatory authorities ultimately are likely to recognise the essential role of intermediation the trade now plays. However, it is dependent in particular on stability in repo market funding. Authorities are likely to institutionalise repo support, including offering direct or indirect backstops to funding - which already partially exist through the Standing Repurchase Facility.

Of course, this will encourage criticism from some quarters. But this is the system the regulators and the fiscal authorities themselves have constructed.