Scenario C: Treasury Market Liquidity, Dealer Risk Capacity and How a Funding Squeeze Develops

1 The scenario and why the Treasury market is the transmission point

The parent article in this series listed three labelled scenarios. Scenario A ran through diesel, farm and freight cash flow. Scenario B ran through credit, refinancing and Subchapter V elections. Scenario C is the markets scenario, and its subject is the place where stress stops being a story about one industry and starts being a story about the machinery every industry borrows against: the market for United States Treasury securities.

The parent described the channel in one paragraph. Stress reaches funding. A survey of market contacts conducted by the Federal Reserve Bank of New York lists market liquidity strains and volatility, and the basis trade, among risks over the next 12 to 18 months. In the scenario, Treasury market depth thins, margin and funding costs rise, and dealers cut risk capacity. The Federal Reserve, in its May 2026 financial stability report, documented that some measures of market liquidity deteriorated during heightened volatility and then recovered.

Two things about that paragraph matter before anything else is said about it. First, it is the parent article's own construction, and this article follows it rather than extending it. Second, it is a scenario. Nothing here is a forecast, nothing here carries a probability, and no reading of the numbers below tells you what will happen to the Treasury market. The purpose is narrower and more useful: to lay out the mechanism in enough detail that a reader can judge for themselves whether a squeeze is likely, and to state clearly what the official record already says about it.

Start with the observed fact that makes the channel live rather than hypothetical. The long end of the Treasury curve has repriced materially over the past two years. On September 16, 2024 the two-year yield was 3.56 percent, the ten-year 3.63 percent and the thirty-year 3.94 percent. On September 23, 2026, by the same constant maturity measures, the two-year was 4.85 percent, the ten-year 5.11 percent and the thirty-year 5.40 percent. Those are dated observations, not projections.

Higher yields are not themselves a stress event. What they change is the arithmetic of holding Treasury risk. A five-point move in yield on a long bond is a much larger move in price than the same move at a three percent yield, because duration cuts both ways. A dealer that warehouses a Treasury position, or a fund that holds one with borrowed money, is carrying more price risk per dollar of notional than it was two years ago at the same measured volatility. When that price risk is financed, the cost of carrying it becomes a live cash item rather than a rounding error.

That is why the Treasury market is the transmission point rather than one market among many. It is the collateral of the financial system and the global risk-free benchmark. Repo is transacted against it, derivatives are priced off it, and the discount rate in a great deal of corporate finance is a Treasury yield plus a spread. Illiquidity there does not stay there. It transmits through the cost of hedging, the cost of funding a position, and ultimately the cost of a loan to a firm that never purchased a bond.

There is a tension worth naming at the outset, because it is the whole reason this scenario is interesting. The Treasury market is the deepest and most heavily traded government bond market in the world, and it is also the market where the firms that intermediate it are least free to expand their balance sheets. Depth and resilience are not the same property. That distinction runs through every section that follows.

2 What the Fed reports, and what the survey of contacts lists

Before building the mechanism, it is worth being precise about what official documents do and do not say, because the categories get blurred in market commentary. The Federal Reserve's financial stability report is published twice a year and is structured as an assessment of vulnerabilities. In the May 2026 edition, as the parent article reported, hedge fund leverage is assessed as stable at record-high levels and concentrated in the largest funds, and margin calls were met without difficulty. The same report states that independent Treasury market intermediaries are limited in their capacity to absorb shocks. And it documents that some measures of market liquidity deteriorated during heightened volatility and then recovered.

Read that last sentence for what it is. It is the central bank describing an observed pattern across episodes rather than predicting an event. It is an expert interpretation of data, published on a schedule, and the parent article treated it that way. Nothing in the report, as documented, forecasts a funding squeeze.

The survey of market contacts run by the Federal Reserve Bank of New York is a different kind of document and must be attributed differently. It is a reported claim: a survey solicits the concerns of market participants, and the published list reflects those contacts' stated worries over the next 12 to 18 months rather than the institution's own outlook. Market liquidity strains and volatility appear on that list. The basis trade appears on that list. Reporting that fact is not the same as asserting that either will materialise, and this article does not assert it.

The two documents are most useful read together, because the credit market data lets you date the pattern the Fed describes. High-yield option-adjusted spread was 4.61 percent on April 7, 2025, the volatility spike of that year, and 2.73 percent on September 23, 2026. Investment grade OAS stood at 0.77 percent on the same September date, with single-B tier spreads at 1.59 percent. The widening and the subsequent compression are the Fed's deteriorated-then-recovered observation, dated to the day on both ends.

Two qualifications belong with those numbers. First, a spread is a price and a price is an observation, not a diagnosis; the FRED series record what traded, and the Fed's report interprets what it means. Second, the compression is not evidence that the channel closed. It is evidence that the last episode ended, which is precisely the point the black-swan test turns on later in this article. Spreads at these levels also say something about the starting position: there is no present stress priced in credit, which is what a scenario has to overcome.

3 Depth, not volume: how a thin book behaves

The scenario's first visible stage is a deterioration in depth, and depth is routinely confused with volume. They are different measurements and they can move in opposite directions.

Volume counts how much trades. Depth describes the cost and the size available at and near the best quoted prices. A market can print enormous volume while offering very little size to anyone who wants to transact immediately, because the trades themselves are consuming liquidity rather than supplying it. The bid-ask spread is the most legible symptom: when the spread between what buyers will pay and what sellers will accept widens, the round trip cost of a position rises for everyone who has to cross it.

In a thin book, three things deteriorate together. The quoted spread widens. The size available at the touch shrinks, so a trade that once cleared at the inside price now walks through several levels. And the price impact of a given order grows, so the same seller realises a worse average price than the same seller did a week earlier. None of these requires the market to stop functioning. Trades still print. That is exactly what makes the condition hard to see in headline activity statistics and exactly why the official language is careful: the Fed wrote that some measures of liquidity deteriorated, not that the market failed.

Why would depth thin in the scenario? Because depth is supplied by firms choosing to hold risk, and that choice depends on the expected return from doing so against the capital and balance-sheet cost of doing it. A volatility shock raises the first term's uncertainty and often the second term outright. When the return to market making falls relative to its cost, the rational response is to quote wider and hold smaller positions. The book does not empty because anyone decided to withdraw from the market. It thins because each participant independently decided to do less.

This aggregation effect is worth stating plainly, because it is the difference between a story about villainy and a story about incentives. No participant in this scenario needs to behave badly. The market's shock absorber is thinner because the number of dollars willing to lean against a price move has fallen, and that is a mechanical consequence of higher uncertainty plus a higher cost of carrying inventory.

The consequence for a leveraged holder is asymmetric. A firm that needs to sell into a deep book pays a small friction. The same firm selling the same size into a thin book pays more, and because its collateral is marked at the observed price, the price move it created reduces the value the market assigns to the rest of its position. Depth, in other words, is not a comfort feature of a market. It is part of the risk position of everyone who holds the asset.

4 Dealer risk capacity as the binding constraint

The parent article's phrase is that dealers cut risk capacity. That is the mechanism's pivot, and it deserves to be unpacked as an economic activity rather than a mood.

A dealer in Treasury securities is not primarily a forecaster. It is an intermediary that buys from sellers and sells to buyers, earning a spread, and holding the residual position in between. That holding is financed with borrowed money and it consumes balance sheet. Intermediation is therefore a balance-sheet activity: the amount of risk a dealer can intermediate is bounded by the capital and funding it has available and by the internal and regulatory cost of putting that balance sheet to work.

When the cost of carrying inventory rises relative to the spread earned, the dealer's optimal response is to intermediate less. It quotes tighter size, holds positions for shorter periods, and becomes more selective about the counterparties and the trades it will warehouse. This is not a failure of nerve. It is the same logic that makes any firm reduce an activity whose margin has narrowed.

The May 2026 financial stability report, as the parent noted, records that independent Treasury market intermediaries are limited in their capacity to absorb shocks. That is an expert interpretation published by the central bank, and it describes a structural condition rather than an episode. The firms that stand between buyers and sellers in the world's deepest bond market carry a constraint that does not relax automatically when volume rises. In a stress episode, demand for intermediation rises at exactly the moment the constraint binds hardest.

There is a structural asymmetry in this that is easy to miss. The largest holders of Treasury securities are not the firms with the most balance-sheet freedom. Funds with mandates to hold Treasuries, foreign official holders and money managers hold enormous quantities, and their holdings are generally not financed on the same terms as a dealer's inventory. So when a holder needs to reduce a position, the set of firms able to absorb that position on short notice is far narrower than the set of firms holding the asset. Concentration on the buy side of a stressed trade is what turns a large market into a thin one.

That is also why this scenario produces an overshoot rather than a revaluation. An overshoot is a price move that exceeds what any change in fundamentals would justify, explained by the fact that the sellers had to sell and the buyers were constrained. The overshoot is painful and self-correcting, which is why it is a squeeze and not a permanent repricing, and why the 2020 and 2025 episodes ended in recovery rather than impairment.

5 The basis trade and procyclical leverage

The basis trade appears by name on the list of concerns in the New York Fed's survey of market contacts, which is a fact about the survey. Understanding why it appears there requires explaining the trade in plain terms.

Cash Treasuries and Treasury futures are two instruments that reference the same underlying government obligation. They trade in different venues, are used by different participants and are subject to different demand pressures, so their prices do not always sit in the relationship that theory suggests. The basis is that discrepancy. A basis trade is a leveraged position that profits from expecting the gap to close, typically by holding the cash Treasury and an offsetting futures position, financed in the repo market.

Three features of that structure are what place it on a risk list. It is leveraged: the position is financed, so the return depends on the cost of funding as much as on the price gap. It is sensitive to margin: the position's economics can change materially when the haircut applied to the collateral rises or when the exchange's margin requirements on the futures leg move. And its unwinding is procyclical. When funding costs rise or margin requirements increase, positions that were profitable become marginal, and the response is to reduce them, which is the response every other leveraged holder is making at the same moment.

Procyclicality is the technical term for a familiar pattern: leverage that grows in calm conditions and contracts in turbulent ones. A trading strategy that is profitable when funding is cheap and volatility is low attracts capital in those conditions, which is when it is easiest to scale and least visible as a risk. When funding tightens and volatility rises, the strategy's holders reduce together, and their reduction is itself a source of price pressure in the market they are unwinding through. The strategy does not need to be reckless for this to happen. It needs only to be leveraged, crowded, and correlated with everyone else's reaction.

The crowding point is the one the survey of market contacts is best placed to observe and the one this article can only describe qualitatively. Market participants see the concentration of positioning in their own order flow in a way that a published statistic cannot capture. That is precisely why a survey of contacts is a reported claim worth reading, and precisely why it is a claim rather than a measurement.

It is also where the honest limit of the analysis sits. The true aggregate size and leverage of basis positions is not fully observable, because the positions are held privately and the reporting that exists covers particular vehicles rather than the whole complex. Any account of how large the trade is, or how much forced unwinding it could generate, is an estimate rather than a reading. That unknown is not a reason to dismiss the concern; it is a reason to hold the scenario loosely.

6 The funding and margin spiral

Everything above describes pressure. The mechanism that converts pressure into an event is the margin spiral, and it is worth walking through step by step because each step is individually reasonable.

A leveraged holder of Treasury securities borrows against collateral. The lender applies a haircut, meaning the holder must post collateral worth more than the loan, and the lender is entitled to require more collateral if the value falls or if the risk assessment changes. That is margin. Separately, a futures position carries margin set by the clearing house, which can be raised when volatility rises.

Now run the sequence. Volatility increases. Margin requirements rise, or the collateral is marked lower, or both. The holder must post additional funds or reduce the position. If it sells, it sells into the book described in section three, which is thinner than it was, so the sale moves the price more than it would have in calmer conditions. The larger price move triggers further margin adjustments for other holders with similar positions, who face the same choice. Each round of selling validates the volatility that caused the first round.

That is a feedback loop, and it is self-reinforcing until something interrupts it. The interrupts are real and have worked before: holders with unused capacity step in, the price reaches a level where the expected return from leaning against it attracts capital, or an official facility provides funding. In the scenario as constructed, the loop is a risk rather than a certainty, and the historical record examined below is that it has been interrupted.

Two amplification channels deserve mention because they operate off the Treasury market's own price. The first is funding market transmission: Treasury collateral is used in repo, so a deterioration in its liquidity and an increase in its price volatility feed directly into the terms on which cash is borrowed against it. The second is the cross-market channel. If dealers are simultaneously absorbing stress in Treasuries and in credit, their capacity to do either is reduced, and the scenario becomes harder to contain. How dealer capacity actually behaves when both markets are stressed at once is one of the genuine unknowns here, because the recent episodes did not test that combination at scale.

It is worth being explicit about the category of the last paragraph. The spiral is a causal inference from the structure of the instruments: margin rules, collateral marking and mark-to-market accounting imply that mechanism. The claim that it would operate in a future episode is scenario content, not an observation, and no probability attaches to it.

7 From the Treasury book to a small-business loan

The reason this scenario belongs in a series that began with diesel and bankruptcy filings is that a Treasury market condition reaches the real economy, and the bridge is the cost of funding and hedging.

Begin with the observed backdrop, which is important because it cuts against the scenario. The net percentage of banks tightening standards for commercial and industrial loans, from the Federal Reserve's senior loan officer survey, peaked at 33.9 percent in the third quarter of 2023. It fell to zero by the fourth quarter of 2024, rose to 18.5 percent in the second quarter of 2025, then eased to 5.3 percent in the first quarter of 2026 and 8.1 percent in the second, before returning to zero in the third quarter of 2026. Those are survey readings with dates. Bank credit is not restrictive at present, and the scenario therefore requires a turn from the current position rather than the continuation of a tightening already under way.

The transmission from the Treasury book to a borrower runs through three channels. The first is the benchmark itself: corporate and small-business lending rates are set as a spread over a risk-free curve, so a repricing at the long end lifts the base on which everything else is calculated. The second is hedging cost. A lender or an operator that wants to fix a rate for a period does so through instruments whose pricing depends on Treasury market liquidity and on swap spreads; wider frictions raise the price of certainty for a firm that is buying it. The third is the supply of credit from market-based lenders. When funding markets tighten and dealers cut risk capacity, the investors who buy loan and bond product require more compensation, and marginal borrowers are the first to find the market closed.

That third channel is the direct link to Scenario B. The operators in that scenario carry floating-rate equipment loans and refinancing schedules. In the scenario as constructed, a thinner Treasury book and higher funding costs raise the rate at which those loans reprice and narrow the set of investors willing to take the refinancing risk. Nothing about the mechanism requires a bank to fail or a lender to lose its appetite entirely. A modest increase in the required spread is enough to move a marginal credit from fundable to unfundable.

A single sentence should mark the boundary of what this section claims. The causal chain from a thin Treasury book to a declined loan application is an inference about how prices propagate, supported by the structure of the instruments involved. The scenario content is that it would happen at sufficient scale to matter for the population of borrowers in Scenario B. That is not observed, and no probability is assigned to it.

8 Why this is catalogued, not a black swan

Everything described above is on the record, and that fact is the analytical conclusion of this article rather than a footnote to it.

Apply the parent article's test. A black swan requires an event outside regular expectations, an extreme impact, and an explanation that appears only in hindsight. Scenario C fails the first condition decisively. The Federal Reserve's May 2026 financial stability report documents that some measures of market liquidity deteriorated during heightened volatility and then recovered, and records that independent Treasury market intermediaries are limited in capacity. The New York Fed's survey of market contacts reports market liquidity strains and volatility, and the basis trade, among concerns for the next 12 to 18 months. The mechanism has been described in official publications, the vulnerable instruments have been named, and the pattern has already played out more than once in recent memory. An event that participants and supervisors are actively watching for is, by definition, anticipated.

The historical comparison makes the point concretely. In the March 2020 episode, Treasury market liquidity deteriorated sharply and the Federal Reserve intervened on an unprecedented scale; the market recovered and continued to function. In the April 2025 episode, spread widening and volatility were pronounced and the market recovered again. Neither episode ended the Treasury market, and both are now the reference cases that shape how the next one would be analysed. That is the opposite of a surprise: the playbook exists, it has been used, and the recovery is the documented outcome.

So what is left open is not whether the channel exists. It plainly does. The open question is whether a future episode would be worse than the catalogued range. Three conditions would point toward worse. Sustained high rate volatility would keep the cost of warehousing risk elevated and prevent depth from rebuilding between shocks. Crowded leveraged positioning in strategies whose unwinding is correlated would convert a price move into a flow. And a binding constraint on dealer capacity, coinciding with stress in credit as well as Treasuries, would remove the shock absorber at the moment it is needed most. Against those, three factors cap the scenario: the depth of official attention to Treasury market resilience, spreads that currently sit at compressed levels and indicate no present stress, and a bank lending posture that is not restrictive.

The unknowns that remain are worth naming precisely, because they define the difference between a scenario and a measurement. The aggregate size and leverage of basis positions is not fully observable, so the potential size of a forced unwind cannot be read off a published series. How dealer capacity would behave under simultaneous Treasury and credit stress is untested in the recent record. Whether official tools would be used again, and in what form, is a policy question rather than a market one, and nothing in the published record commits to an answer. Each of those is an uncertainty about magnitude, not about the existence of the channel, which is exactly why the honest conclusion is the parent article's: this is a catalogued vulnerability that would not qualify as a black swan, and the question of severity is the one that stays open.