Bond Market Liquidity and Dealer Balance Sheets
Treasury market liquidity does not come from an exchange order book — it comes from dealers willing to warehouse bonds on their own balance sheets, and post-crisis capital rules mean that willingness now shrinks exactly when the market needs it most.
Prerequisites: Repo and Reverse Repo, The Money Market and the Short End of the Curve
The US Treasury market is the deepest bond market in the world, and yet when a wave of selling hits it does not absorb the flow smoothly the way a textbook describes — bid-ask spreads widen, trade sizes shrink, and prices gap. The reason is structural: almost every Treasury trade a client wants to do goes through a dealer's own balance sheet first. The dealer buys the bond from a seller, holds it in inventory, and looks for a buyer later. That warehousing capacity is not infinite, it is not free, and since the financial crisis it has been directly capped by regulation — which means market liquidity now has a regulatory ceiling that tightens exactly during the episodes when the market needs the most capacity.
Dealers provide liquidity by holding bonds on their own balance sheet between a seller and the eventual buyer. Since Basel III, that capacity is capped by leverage ratio rules that treat a low-risk Treasury the same as a risky loan for sizing purposes — so a flood of selling that dealers would once have absorbed can instead blow out bid-ask spreads, because taking on more inventory costs regulatory capital, not just balance sheet space.
Why a bond, not an order book, needs a dealer at all
An equity can trade directly between two anonymous participants on an exchange because a liquid stock has continuous natural two-way flow. A given Treasury CUSIP does not — most of the time there are more natural sellers than buyers, or vice versa, at any instant, especially away from the handful of most recently issued, most liquid "on-the-run" bonds. A dealer bridges that gap by taking the other side itself, buying when clients want to sell and selling from inventory when clients want to buy, earning the bid-ask spread as compensation for the risk of holding bonds whose price can move against it before a matching client shows up.
The regulatory constraint: the supplementary leverage ratio
The Supplementary Leverage Ratio (SLR), introduced after 2008, requires large banks to hold capital against total assets — including Treasuries — without any risk-weighting adjustment for the fact that a Treasury is close to risk-free. A commercial loan and a Treasury note consume the same leverage-ratio capital per dollar of balance sheet, even though only one of them can plausibly default. The practical effect: at quarter-end, when banks are measured and reported on this ratio, dealers pull back Treasury inventory to make their balance sheets look leaner on the reporting date, and bid-ask spreads widen measurably around those dates.
Worked example: spread widening under stress
In calm conditions, an off-the-run 10-year note might trade with a bid-ask spread of about 1/32 of a point (roughly 3 cents per $100 face, or $3,125 round-trip on $10 million face). During the March 2020 "dash for cash," when everyone from foreign central banks to leveraged relative-value funds tried to sell Treasuries simultaneously, spreads on some off-the-run issues widened to 4/32 or more — over 12 cents per $100 face:
on $10 million face, i.e. $12,500 round-trip.
Four times the calm-market cost, on the single deepest bond market in the world, because every dealer was simultaneously being asked to absorb inventory it had shrinking regulatory room to hold — the exact moment demand for liquidity spiked was the moment supply of it contracted.
Where dealer balance sheets already show cracks
Repo rates spiking in September 2019 and the Treasury market seizing up in March 2020 both trace back to this same mechanism: dealers ran out of regulatory room to intermediate exactly when the demand to trade spiked. The Fed's response in both cases was to step in directly — as a repo counterparty in 2019, as an outright buyer of Treasuries in 2020 — because private dealer balance sheets alone could not clear the market at reasonable prices.
"The Treasury market is the deepest and most liquid in the world" is true on average and can be badly wrong in a crisis. Liquidity in Treasuries is not a fixed property of the asset — it is a function of how much balance sheet dealers are willing and regulatorily able to commit at that moment, and that quantity is procyclical: it shrinks precisely when stress makes everyone want to trade at once.
Where it shows up
Every relative-value and basis trade in Treasuries (see The Implied Repo Rate and Net Basis) depends on dealers being able to warehouse the position; when balance sheet is scarce, basis trades widen and relative-value residuals persist longer because the arbitrage capital needed to close them is capacity-constrained too. Regulators have since proposed central clearing mandates and SLR exemptions for Treasuries specifically to loosen this bottleneck, an ongoing and unresolved debate in market structure policy.
Key terms
- Dealer inventory — bonds a dealer holds on its own balance sheet between buying from one client and selling to another.
- Supplementary Leverage Ratio (SLR) — a capital rule requiring banks to hold capital against total assets, including low-risk Treasuries, without risk-weighting.
- Quarter-end effect — the widening of dealer spreads around regulatory reporting dates as banks temporarily shrink balance sheets.
- Procyclical liquidity — the tendency for market-making capacity to shrink exactly when demand for liquidity rises.
Related concepts
Practice in interviews
Further reading
- Duffie, Still the World's Safe Haven? Redesigning the U.S. Treasury Market After the COVID-19 Crisis
- Federal Reserve Bank of New York, Staff Reports on Treasury Market Liquidity