A safe asset that needs defending is not safe. The world’s benchmark collateral wobbled after a political move against a Federal Reserve governor, and that wobble revealed more than a headline. It exposed the system’s reliance on soft constraints, short memories, and cheap leverage. When the bedrock of pricing starts to trade like a speculative instrument, you are not looking at a one-off. You are looking at a structure that has been optimized for fair weather, and that now meets a harder climate.
Long Treasuries sold off while short yields fell. The curve steepened to near a three-year high. The trigger was an attempt to remove Fed Governor Lisa Cook over alleged misconduct, which markets read as a challenge to central bank independence. If the Fed’s autonomy can be bent by political pressure, then rates can be cut sooner, but future inflation risk rises. That mix flattens the front end and lifts the long end. The dollar slipped, a small but telling 0.3 percent move, because currencies price credibility as much as carry. You can see the game theory at work. When the rule-maker looks uncertain, the players update fast. Term premium, missing for a decade, rushes back. Bonds reprice the cost of time inconsistency in a matter of hours.
There is a name for this. Fiscal dominance is when the needs of the Treasury shape the decisions of the central bank. The more debt outstanding, the higher the political incentive to keep rates down and the interest bill contained. History shows how this ends. Before the 1951 Accord, the Fed capped yields to finance war debt. In the 1970s, political pressure blended with policy drift, and the result was stagflation. The core problem is commitment. In a repeated game, credibility is collateral. Once it is marked down, participants demand a higher premium to hold duration. That premium shows up in a steeper curve and a weaker currency. You do not need a formal policy change to get there. Market participants price the probability of interference. Even a small rise in that probability has a large effect because it touches the discount rate that governs everything else.
The market’s plumbing makes these swings larger. Hedge funds have again built large relative value positions in Treasuries and futures, with the gross scale around hundreds of billions of dollars. The classic basis trade is long cash bonds financed in repo and short futures. When yields rise or policy risk spikes, funding haircuts can rise and margins change. The position is capital efficient until it is not. In April 2025, inflation fear and erratic policy signals triggered a sell-off in Treasuries that forced an orderly but sizable unwind in these positions, estimated near 800 billion dollars in size. Today, funds are still net short over a trillion dollars in Treasury futures by some measures. The leverage lives in the shadows of dealer balance sheets and short-term funding. This is not villainy, it is engineering. But the structure has negative convexity. Small shocks force de-risking, which widens spreads, which forces more de-risking. The feedback loop is built in.
The policy answer on offer is to relax bank capital rules, including the supplementary leverage ratio, to lure more balance sheet into Treasuries. It sounds stabilizing. It may do the opposite. Commercial banks learned the 2023 Silicon Valley Bank lesson on duration the hard way. Neutral risk weights do not neutralize interest rate risk. They will not rush to load up on long bonds while rate volatility is high. Broker-dealers, who actually intermediate repo and futures-related flows, hold a small slice of Treasuries but feel SLR’s bite on financing. Looser ratios would lower their cost of funding hedge funds’ basis trades. More balance sheet would chase the same spread and magnify the same procyclical leverage. In calm periods, it looks like liquidity. In stress, it becomes a fire hose that runs dry. If the goal is stability, subsidizing short-term leverage to support long-term debt is a poor design choice.
Treasury market depth is real until it is needed most. We have seen this movie. The 1994 bond rout crushed leveraged bets that assumed smooth mean reversion. In March 2020, Treasuries dislocated as the dash for cash overwhelmed balance sheets, and the Fed had to act as market-maker of last resort. In 2023, banks sat on large unrealized losses after rates rose, and the fastest deposit flight in the digital era forced asset sales. In April 2025, a political shock met a pile of leveraged basis positions, and yields lurched. The pattern is a sandpile. Grains accumulate, stability looks fine, and then a minor shock causes a slide because the slope was already critical. Probabilities in these systems are not normal. Correlations go to one as principal trading firms and dealers widen spreads or stand back. VaR models shrink in calm and explode in stress. The risk is not volatility; it is liquidity that vanishes when everyone runs the same play.
Investors still anchor to the idea that Treasuries are cash-like. A survey shows only 35 percent of retail investors feel confident in the economy and 55 percent see inflation as the main risk, yet more than half did not change their portfolios. Many still plan to add money in the next year. This is not rational confidence. It is status quo bias and underweighting regime change. Banks made the same error before 2023, holding long duration “safe” assets against runnable funding. The safe asset is safe if you can hold it. If you must finance it, roll it, or mark it to market under stress, the safety is theater. Markets also cling to a monetary backstop that may be politically harder to deploy if inflation risk is back. The dollar’s slip during a governance scare was a small signal that the world notices. Credibility is earned slowly and repriced fast.
Economists, banks, and lawmakers can denounce attacks on central bank independence. Statements do not rebuild term premium once it is gone. Markets watch actions and constraints. If the Fed is seen as an agent of fiscal needs, even occasionally, then every auction, every issuance choice, and every policy meeting becomes a coordination game with poor equilibria. The solution is dull and hard: keep the boundary clear between debt management and monetary policy, and make that boundary costly to cross. Treasury can extend the average maturity of issuance to reduce refinancing risk rather than chasing the cheapest point on the curve. The Fed can defend its autonomy with rule-like behavior that narrows the space for ad hoc moves. These steps do not excite markets. They make the system less interesting. That is the point.
If you want a market that does not need prayers, redesign the incentives. Make margining and haircuts more countercyclical across repo and futures so leverage shrinks in euphoria and expands in stress. Expand standing facilities that provide elastic, priced liquidity against Treasuries to reduce forced sales, with clear limits to avoid subsidy. Central clearing can help visibility, but without guardrails it can also concentrate risk. Encourage more balance sheet that is patient and unlevered to absorb shocks, even if it costs more in normal times. Above all, treat Fed independence as the actual collateral behind the dollar and the curve. When institutions are strong, term premium is low because the rules are credible. When the rules wobble, you can write all the SLR relief you want; spreads will still gap because the foundation moved.
A Treasury market that is truly safe would not be this sensitive to a personnel fight, a tweet, or a funding spread. The lesson is not to fear Treasuries. It is to respect how tightly we have coupled political risk, leverage, and pricing into the same narrow channel. In engineering, tight coupling without slack fails at the weakest link. The same is true here. Add slack. Guard the rules. Price the tails.
