NextFin News - Hedge funds have built one of the most lopsided bearish bets against US Treasuries in two decades of records, a net short position of roughly $824 billion that will pay off only if the Federal Reserve delivers a rate hike at its September 15-16 meeting. After August's hotter-than-expected inflation print, fed funds futures now price about a 90% chance of a quarter-point increase, up from 37% just a week earlier. The bond market has crowded into a trade where the expected outcome leaves little room to win and a hold decision could force a violent squeeze.
The setup is a wager on a single institution's credibility, expressed at record size. It is also a reminder that in derivatives markets, being right about the event is not the same as being paid for the position.
The Positioning: A Record Short Built on One Assumption
Leveraged money is short about $823.6 billion of Treasury futures notional, according to the Commodity Futures Trading Commission's Traders in Financial Futures report for the week ended August 25, reconstructed from contract counts. On the benchmark 10-year note, leveraged funds sat at a net short of 42.4% of positioning as of early August, with asset managers on the other side at 49.3% long. The two sides of the trade are unambiguous: speculators are betting yields go higher, and real-money accounts are absorbing the risk.
The scale matters because of what sits behind it. The Treasury cash-futures basis trade — in which funds buy cash bonds, short the corresponding futures, and finance the position in repo — has grown to roughly $830 billion, about double its early-2020 peak. When futures are heavily shorted by leveraged funds while asset managers hold the cash side, the short is not an isolated macro view; it is a leveraged node in the plumbing of a $30 trillion market. A move against the crowded side forces re-hedging that can amplify the very price action the shorts are trying to monetize.
The timing sharpens the asymmetry. The 10-year Treasury yield closed at 4.98% on September 14, up about 0.94 percentage point over the past year and 0.25 percentage point in the past month alone. The 30-year bond printed as high as 5.25% in late August, a level not seen in nearly two decades. If September closes near current levels, the Treasury sell-off will stretch to seven consecutive months of losses — a streak that, in the historical record, has extended to a seventh month only twice before.
Record shorts plus a seven-month losing streak is the classic signature of a crowded trade at an extreme. The bond market is not merely expressing a view on inflation. It is making a binary wager on next week's FOMC decision, and it has priced that decision close to certainty.
The Fed Bet: What the Market Has Priced In
The short trade is, at its core, a leveraged expression of a single expectation: that Fed Chair Kevin Warsh will deliver the first rate move of 2026 at the September meeting. The pricing is explicit and has moved fast. CME FedWatch showed fed funds futures pricing a greater than 70% probability of a rate hike at the September meeting as of September 10, up from 37% one week earlier. After the August consumer-price report, that probability climbed to about 90%. A 25-basis-point increase would lift the federal funds target range from 3.5%-3.75% to 3.75%-4.00%.
The repricing followed Warsh's first Jackson Hole symposium speech as Fed chair, in which he said inflation remains too high and that the central bank may need to raise rates in the coming months. In his own words, he and other officials who supported holding rates steady at the July 28-29 meeting "thought the wiser course was to await new information in the intermeeting period ... before deciding whether a change in interest rate policy was advisable." The intermeeting period then delivered an August consumer-price report that pushed the odds higher still: the all-items CPI rose 3.4% year over year, above the 3.3% consensus forecast, while core CPI slowed to 2.4% from 2.5%.
Deutsche Bank said Warsh's address "surprised us in its specificity about the economy and outlook and with its lean in a decidedly hawkish direction." The firm expects 50 basis points of hikes this year, at the September and December meetings.
The market reaction fit that diagnosis: the short end of the curve sold off while the long end rallied, a steepening response to a Fed chair who sees inflation as the primary problem.
But the pricing is not uniform across venues, and the gap is informative. On the Kalshi prediction market, a 25-basis-point hike carried a 53% probability against 46% for no change, with only 2% for anything larger. That spread between the futures-implied probability and a direct event market is itself a warning: the consensus is priced, but it is not unanimous, and the conviction is concentrated in leveraged accounts rather than broad-based conviction.
The Mechanism: Why a Priced-In Hike May Not Pay the Shorts
Here is the uncomfortable arithmetic for anyone short Treasuries. A rate decision is a discrete event with three possible market reactions, and only one of them rewards the crowded trade decisively.
If the Fed hikes 25 basis points exactly as expected, the short position captures, at most, the small slice of the move that was not already priced. With about 90% of the outcome embedded in futures, the surprise component is thin. Yields may grind higher on the statement's policy path or on Warsh's press conference, but the headline itself is old news by the time it prints. This is the already-priced conventional wisdom: the market has rewarded the bears in advance.
If the Fed holds, the reaction is not a small loss — it is a short squeeze. A crowded one-sided position unwinds through the same door everyone is trying to exit, and the move is measured in days, not basis points. The reference point for leveraged Treasury positioning is March 16, 2020, when the SOFR spread blew out from 12 basis points to 174 basis points in a single session as the basis trade collapsed. Funding conditions today are far from those levels — SOFR's 99th percentile spread has held in a 7-to-9 basis point band since mid-August — but that is precisely why the risk is hidden rather than absent: the dial reads calm until it does not.
If the Fed hikes 50 basis points, the shorts win decisively — but that outcome carries only about 2% implied probability. The trade is therefore short convexity: it collects small gains in the expected case and risks a large loss in the disagreement case. The asymmetry runs the other way from what the positioning suggests.
The second-order channel runs through the basis trade and the curve. A hawkish surprise forces basis traders to re-hedge, amplifying moves in both cash and futures. A dovish surprise — a hike paired with guidance that it is a one-off credibility move — would send yields lower even as the Fed tightens, because the market would price the end of the cycle. In other words, the Fed can hike and the bond market can still rally. That is the trap built into a short positioned for a simple directional move.
The deeper transmission channel is the term premium — the extra yield investors demand for holding long-dated debt instead of rolling short-term bills. That premium has been rising not because traders expect higher overnight rates, but because they are being paid to absorb a supply of Treasury debt that keeps growing while the Fed's balance sheet shrinks. A rate hike addresses the first problem and ignores the second. If the market reads the hike as an admission that inflation is more persistent than the Fed previously acknowledged, the term premium can widen even as the policy rate rises — lifting the long end while the short end is pinned by the decision itself. That is how a Fed hike can coexist with a bear-flattening that punishes the very shorts betting on higher yields.
The Counter-Thesis: The Crowd Is Usually Wrong at Extremes
The strongest argument against the short is not macroeconomic — it is positional, and it comes with a history. Raymond James noted that with investor sentiment already deeply bearish on bonds and speculative short positions in 10-year Treasuries near record highs, the move appears increasingly overextended. Their analysis concludes that yields may be closer to a peak than the start of a new, sustained leg higher, and that following periods of six or more months of selling, reversals are common — only two prior streaks reached a seventh month.
Matthew J. Maley, chief market strategist at Miller Tabak, went further, arguing "there remains no empirical basis for the rate hike." His view is that Warsh is talking up inflation so he can claim credit for taming it when headline measures inevitably come down, and that labor-market data has been weak while inflation has come in better than expected since the last FOMC meeting. The August print partially supports that read: while headline CPI surprised to the upside, core CPI continued to cool, and the Fed's preferred core PCE measure has been running near 3.4% year over year — above target, but on a downward path rather than an accelerating one.
There is also the question of what a hike actually accomplishes. Much of the inflation pressure in 2026 has been energy- and supply-driven, tied to conflict in the Middle East and disrupted shipping. As one portfolio manager put it during the summer's oil spike, the Fed "is handcuffed" on the supply side of inflation, and "you can't hike your way into producing more oil for the global economy." A rate increase that does not address the supply side of inflation buys credibility at the cost of growth, and the bond market may recognize that trade-off faster than the Fed does.
The falsifying signal for the counter-thesis is concrete: if core CPI prints at or above 0.3% month-over-month for two consecutive months, or if core PCE re-accelerates above 3.5% year over year, the "no empirical basis" argument fails and the hawkish case regains empirical footing. Until then, the shorts are betting on policy credibility, not on inflation data that has consistently surprised to the downside on the core measure.
Who Wins, Who Loses, and What to Watch
The decision splits the market into clear winners and exposed parties. If the Fed delivers the expected hike with a muted yield response, the beneficiaries are the short end of the curve and money-market funds, which capture higher yields on cash without a violent repricing elsewhere. The exposed are the crowded leveraged shorts, who are right but unrewarded, and duration holders who survive on the hope that this is the last hike rather than the first.
If the Fed holds, the winners are the short squeezers and the long-duration funds that have absorbed the selling — a rapid rally would mark to market the largest transfer of the cycle. The exposed are the leveraged shorts forced to cover at worse prices, and any strategy that assumed a one-directional path for yields into year-end.
The FOMC announces its decision on September 16, 2026, after a two-day meeting. Three signals will determine whether the record short was prescient or crowded folly.
First, the decision itself: a 25-basis-point hike confirms the consensus; a hold triggers the squeeze; a 50-basis-point hike validates the most aggressive bears. Second, the policy statement and Warsh's press conference: a hike paired with guidance that this is a one-off move would likely send yields lower, not higher, as the market prices the end of the tightening cycle. Third, the 10-year yield's reaction: if it fails to break decisively above 5.10% after a hike, the short thesis is exhausted — a classic "sell the fact" outcome.
The horizons point in different directions. Short-term, this is a positioning story: crowded shorts, extreme sentiment, and a binary catalyst. Medium-term, the fundamentals matter more — inflation prints, labor data, and whether the Fed's credibility gamble restores its inflation-fighting reputation without breaking growth. Long-term, the structural driver is not the Fed at all: it is the term premium demanded for holding US debt in an era of large fiscal deficits and a global supply of bonds that keeps growing. That structural leg is the one part of the bear case that survives any single meeting outcome.
The base case is a 25-basis-point hike with a muted yield response, leaving the shorts technically right but economically unrewarded. The upside case for bears is a 50-basis-point surprise that sends the 10-year through 5.25%. The downside case is a hold decision that forces leveraged shorts to cover into a rally, with the 10-year falling back toward 4.50%.
The bond market has bet more than $800 billion that the Fed will do exactly what everyone expects it to do. In markets, that is usually how money is lost, not made — because the trade crowded to the edge is the one most vulnerable to the single outcome nobody is positioned for: a Fed that delivers the hike, and nothing else.
Explore more exclusive insights at nextfin.ai.
