Why do markets break at their strongest point? Because what looks strong is often just leverage disguised as liquidity. The dollar system is not only cash and bonds. It is a web of swaps, futures, and repo rolled day after day under the assumption that tomorrow will look like today. Hedge funds shifting from traditional basis trades into leveraged swaps is one more turn of the screw on a machine that depends on smooth rolling. It works—until it does not.
Much of the world’s dollar borrowing never shows up as loans or bonds. It lives inside foreign exchange swaps and forwards, where non US banks and shadow banks borrow dollars by swapping their local currency cash flows and pledging collateral. By global estimates, this hidden dollar obligation has swelled to tens of trillions. The notional size matters less than the structure: principal is due at maturity, and most positions rely on short dated rollovers. Accounting treats these like derivatives, not funding, so the true debt load is undercounted and misunderstood. Corporates hedge dollar revenues this way. Insurers hedge dollar assets this way. Asset managers park cash this way. In calm times, margin nets exposures and lowers apparent risk. In stress, the need to roll or post variation margin turns a quiet hedge into a scramble for dollars.
The classic hedge fund basis trade—buying cash Treasuries financed in repo and shorting futures—thrives on small, stable gaps. As those spreads compress or become more volatile, funds adapt. They pivot into interest rate swaps, cleared at central counterparties, and FX swaps or cross currency basis trades to harvest carry and duration away from balance sheet constraints. On paper, this looks safer: clearing reduces bilateral credit risk and netting lowers margin. In practice, it centralizes exposures at a few nodes and assumes that collateral remains cheap and rolls remain frictionless. The leverage does not leave the system; it moves to the venue with the least binding constraint. When that venue wobbles, the same trades become crowded exits in narrow corridors.
Recent research from the Federal Reserve highlighted a sharp build in hedge fund Treasury futures and repo positions, reflecting how nonbanks have become core liquidity providers in the world’s benchmark market. Later, the Office of Financial Research reported a notable decline in hedge fund repo borrowing in late 2024. That sounds like de risking. It may simply be migration. Less repo can mean more swaps or synthetic exposures in futures, options, and cleared derivatives where leverage is embedded and netted. Liquidity that depends on daily margin and sponsored netting is like a bridge tuned to a single frequency. It looks solid under expected loads. A change in wind—collateral haircuts, CCP margin hikes, or dealer balance sheet limits—can push it into resonance. Then liquidity vanishes all at once.
Leverage concentration is less a credit story than a game theory story. When a crowded strategy meets a shock, best response becomes clear: sell what is liquid, post what is demanded, and shrink. Everyone knows everyone else must do the same. That common knowledge drives a margin spiral. Central clearinghouses raise initial margins. Prime brokers tighten financing. Dealers protect scarce balance sheet. The price impact feeds back into measured risk and triggers more selling. In FX swaps, the risk is even blunter. If you cannot roll, you owe dollars you do not have. What was coded as a hedge becomes a payable. Participants understand this dynamic in theory. In real time, no one volunteers to be the last liquidity provider.
We have seen versions of this movie. In 1998, LTCM’s leveraged relative value and swap exposures blew up when correlations snapped and liquidity thinned. In September 2019, a modest cash drain collided with dealer constraints, and repo rates spiked until the central bank stepped in. In March 2020, the Treasury market—supposedly the deepest pool—buckled as levered basis positions met outflows and margin calls, forcing sales of off the run securities. In 2022, UK pension funds hedging rate risk with leveraged swaps faced collateral calls that cascaded into gilt sales. Different markets, same pattern: leverage stacked on supposedly safe assets, financed short, unwound fast. Stability begets risk taking; risk taking removes buffers; a small shock finds the fault line.
Risk systems love recent history. Value at risk compresses as realized volatility falls. Funding assumptions bake in market depth that exists only when it is not needed. Capital and liquidity regulations made banks sturdier, but they also pushed activity into nonbanks and made dealer balance sheet more expensive. Central clearing cleared bilateral credit, but it did not make collateral elastic. Accounting keeps FX swap principal off balance sheets, so headline leverage looks lower than functional leverage. Probability models built on tidy distributions are miscalibrated for plumbing events, where price and funding shocks arrive together. In those moments, correlation goes to one, haircuts gap wider, and the parameter that matters is time to liquidity, not expected return.
A system that gains from disorder has slack, redundancy, and modesty about what it knows. The dollar funding web has moved in the opposite direction. Every efficiency—netting, just in time rolls, collateral upgrades—pulls slack out. Central bank backstops are powerful, but they are not a business model. They cannot be pre committed to every market and every instrument without inviting larger risks. Investors talk about carry as if it were rent on a sturdy house. In truth, a lot of carry is selling insurance on the absence of accidents in the plumbing. The premium looks free until the claim arrives. A little less leverage, a little more term funding, and a little more unnetted cash on hand are not glamorous. They are what make a system absorb shocks instead of amplify them.
Prices still talk. Watch the cross currency basis for stress in offshore dollar demand. Watch the spread between implied dollar funding in FX swaps and overnight index swaps. Watch on the run versus off the run Treasuries, futures basis tweaks, and specials in repo that hint at collateral scarcity. Track changes in CCP margin schedules and clearinghouse skin in the game. Read the reports on hedge fund financing, not just their returns, and compare repo borrowing trends with futures open interest and swap volumes to see where leverage migrates. Dealer inventories, settlement fails, and the take up of central bank facilities tell you when balance sheet is tight. None of these are predictions. They are early warnings that the machine is running with less slack. The question is not whether leverage is present. It is whether it can roll when the music changes.
