The distribution channels of new US treasury bond issues have changed radically since the GFC, and become especially marked since the Fed began tightening in 2022. Conventional asset managers have chosen to increase massively their exposure to Treasuries via derivatives. The counterpart is an equally huge increase in short derivative holdings held by leveraged accounts, who also increased their funded cash holdings. There’s been a lot of focus by regulators on the bond/futures basis trade, which forms a part of this shift to funded Treasury holdings. But focus on the basis trade misses the wider implications of increased holdings of futures by asset managers. This is larger than the basis trade and creates a kind of ‘proxy basis’ risk. The leverage inherent in this arrangement for both longs and shorts raise questions about how future volatility may disturb both holdings and funding.
A simple balance sheet outlines the main changes in Treasury holdings of key players in the last couple of years. I’m open to criticism here and I’ll try to highlight where there is room for disagreement. But I think the broad changes are clear.
Let’s start with a simplified T account description of the various important changes in balance sheet since March 2022, when monetary tightening began.
The simple balance sheet shows the addition of Treasury liabilities (top entries) and the addition of Asset Manager futures longs and cash at the bottom. The entries in between map the main route through the investor groups, including the role of the Federal Reserve and money market funds (MMF). The bottom line is there has been a ton of issuance, but conventional asset managers have taken up only some of this issuance. The difference has led to a large long/short divergence in derivative markets based on investor types. I attempt to capture the extent this intermediation in the following charts.
To read through the individual entries, the balance sheet approach shows the Treasury has added liabilities in the form of Notes/Bonds and T. Bills (top lines) while the Fed has reduced both sides of its balance sheet (second entries). The financial market impact of increased Treasury Bill issuance has counteracted the reduction in Fed’s ONRRP held by money-market funds (third entry). This can be seen clearly by showing cumulative changes in Bills outstanding & ONRRP since March 2022. T
A good part of the increased Bill issuance has gone to MMF, partly replacing ONRRP holdings and partly absorbed by higher holdings. However, overall MMF repo investment has not declined in line with the decline in ONRRP, suggesting a switch from Fed repo (ONRRP) to market repo, possibly associated with the rise in repo offered to hedge funds.
Leveraged accounts have added huge exposure to cash Treasury bonds via the cash/futures basis trade. But perhaps the most important entry is the last one, which reflects Asset Manager holdings of futures positions, as a liquid and capital efficient alternative to buying cash bonds.
Asset Managers now have the largest-ever net exposure to Treasury futures markets. Much attention has been directed to the main counterpart to this holding; namely, hedge fund holders of cash/futures basis, a trade that has been consistently profitable since the Fed began to raise rates in March 2022.
Much less attention has been given to the Asset Managers net longs as a counterpart to Leverage Accounts short positions (basis trade plus all other net short positions). This omission may be down to the conventional concerns about ‘leverage’ in the basis trade and perhaps an assumption that Asset Managers will always have cash ready to fund any volatility in margin or funding markets. That is not a certainty.
While Asset Managers (real-money accounts) may be less sophisticated than leveraged hedge funds (who manage the basis trade for profit), the very reason the net long futures position has increased must mean the net market exposure to margin and funding requirements inherent in futures positioning has increased as well. It would be wrong to think that Asset Managers are always able to fund positions held in derivatives out of available cash reserves. As the Gilt market LDI episode showed, conservative pension funds did not have sufficient funds available when bond market disturbance led LDI derivatives to call for additional cash. The resulting liquidity vacuum led to fire sales of government bonds, and the intervention of the central bank to stabilise market prices. Even if funding is not an issue, there is an implicit funding issue for the net short counterpart in futures, held by leveraged accounts. This includes the cash/basis trade, plus other net short positions.
Simply measuring the basis trade – as the Fed has attempted to do in a March 2024 report – may not measure the true risk because the net long held by AM accounts creates a proxy basis position on the other side which may be exposed (we don’t know) to the same funding and margin pressures as actual basis positions under stressed conditions. The upper limit of futures as a potential risk, therefore, is the net short position held by leveraged accounts which is equivalent to the net long position held by Asset Managers. To focus solely on hedge fund positions in the basis trade may miss a wider risk.
Most of the basis trade (in face value terms) is concentrated in 2y, 5y and 10y – with most at the shorter end of the curve. The shorter maturities have lower duration exposure, but in terms of repo financing it is face value that counts.
In the following chart long-end futures (Ultra 30-year and Bond futures) are excluded because they account only account for 10% of face value of the net futures short held by leverage accounts so don’t make a huge impact on outstanding funding amounts. Despite a large increase in long-end issuance since March 2022, but there has not been a huge increase in Leveraged net short positions. The proxy basis risk, therefore, has not increased much in long maturity positions. The focus is on issues shorter maturities.
With that caveat, we highlight the strong linear relationship between change in Note issuance from Treasury, change in DVP repo amounts outstanding (funding for individual issues) and change in the net short position of leveraged futures positions (excluding long-end futures). This relationship suggests (doesn’t prove) that the proxy basis risk is much bigger than the actual basis trade calculated by the New York Fed.
Overall, the chart suggests that Treasury Note issuance may have been intermediated via the futures market into Asset Management accounts, with hedge funds accumulating short derivative positions to match the rise in issuance, with a corresponding rise in repo to fund cash holdings, much of which is facing MMFs. It is reasonable to assume this analysis obscures the different risks in various trading strategies. It does suggest that the big question is the risk located in futures and funding in total, not just basis. And this risk has been created by Asset Managers preference for holding Treasury exposure via derivative positions.
This change reflects dramatic adjustment in the intermediation structure of the Treasury market, coinciding with historically large issuance. In the old days dealers may have increased holdings of bonds issued in anticipation of future demand. These days dealers are relatively minor figures in the intermediation of Treasury issuance. Total holdings of Bills, coupons and TIPS by dealers’ amount to $250bln, roughly 25% of the net exposure in futures markets.
How does this all matter? It means the Treasury market is more exposed to stress in any part of the cash/futures/funding nexus – including volatility in both the bond market, the bond futures market (margin) AND the repo market (funding). Any one of these can lead to forced position adjustment in either hedge funds and possibly Asset Manager accounts and will tend to transmit stress through to the other parts of the intermediation chain.
Both bond market and repo has been increasingly volatile in recent months. The volatility in repo funding could well reflect the size of positions requiring funding, especially at pinch points such as quarter-end. Whether that is a problem is an open question. It is certainly worth considering the overall implications of derivative positioning and how it relates to expanded issuance. Not least because expanded issuance is probably a permanent feature of the Treasury market.