In May 2026, Mark Dowding, chief investment officer at BlueBay Asset Management, ran a stress test the market did not request. The subject: Federal Reserve credibility. The conclusion: potential collapse.
Dowding's read is specific. Under incoming Chair Kevin Warsh, the Fed is preparing to abandon forward guidance — the communication layer that, since 2008, pre-committed policy paths so investors can price the future. The trail is stark. "In the era of forward guidance," Dowding argued, "the Fed enjoyed high trust." Remove the guidance: information vacuum. Doubts compound. Confidence evaporates.
This is not a stylistic disagreement. The U.S. Treasury market is global finance's settlement layer. It runs on one assumption: the Fed knows what it is doing, and says what it knows. Behind it sits debt at record levels, growing at an alarming speed. Deleting the oracle layer from the world's largest collateral database is not a communication update. It is a protocol change.
Dowding's warning deserves a forensic read. Not because he answers every question, but because he names the right variable: trust.
Context: A Rules-Oriented Chair Meets a Rule-Dependent Market
Warsh is not new to the machine. He was a Fed governor during the 2008 crisis, inside the building during the 2013 taper tantrum, and has spent his career questioning quantitative easing. A rules-based hawk by instinct, he runs against the expectation-management orthodoxy that took root after the crisis. Warsh favors discretion over predictability. His instinct is to act, then let markets interpret. The market's instinct is to be told in advance.
That puts him in tension with a market built on predictability. Modern monetary policy transmits through expectations, not just the policy rate. When the Fed tells investors where rates are going, the market aligns: borrowing costs adjust, duration is managed, inflation expectations anchor near 2 percent. The Fed runs a protocol with forward guidance as its consensus mechanism — millions of nodes staying aligned because the communication layer is honest.
Abandon it, and alignment becomes trial and error. The market reverts from trust-but-verify to trust-without-anchor. It does so while the federal debt bleeds — tracing the silent bleed from 2017's broken logic, when tax cuts passed without a lasting funding mechanism and every subsequent year layered compounding interest onto the structural gap.
The Transmission Breakdown
The logic is mechanical. Remove forward guidance → policy path predictability drops → investors demand a higher term premium on long-dated bonds → long-end yields rise → duration repricing hits every asset priced off the 10-year. Mortgage rates follow. Equity discount rates follow. Federal interest expense follows.
That last point is what Dowding implies but does not state: the U.S. debt is now a destabilizing variable in the Fed's own transmission chain. The federal government is the largest borrower in every market it touches. Its financing needs are inelastic — it rolls over trillions annually regardless of price, and the roll cost compounds each quarter. Every rate rise is both a policy tool and a budget shock. This is fiscal dominance in its cleanest form: the borrower's affordability becomes the lender's policy constraint. A Fed that cannot predict its own path cannot manage the fiscal monster it helps fund.
The chain of custody breaks at the same point every time: expectation. When the market aligns with a path, the path stabilizes. When the path vanishes, the market resorts to speculation — and speculation is volatility.
The Ledger of Distrust: Term Premium and Auction Bids
Forensics reveal the truth markets try to bury. The 10-year term premium — compensation for holding long duration — sat near zero for much of the post-2010 period. It has since crossed positive and trended toward 50 basis points. The market now charges the Treasury for the privilege of holding its debt. That is a lending pool slowly raising its borrow rate before the collateral check fails.
The next signal is auction demand. The bid-to-cover ratio — bids against each dollar of issuance — historically runs between 2.3 and 2.6 times. A sustained break below 2.0 is the validation event: the moment the market stops pricing default risk and starts refusing to fund the state. DeFi calls that a bank run on a stablecoin reserve. Confidence evaporates at a threshold, not along a curve. The market is not a linear instrument; it is a threshold machine.
What the Bulls Got Right
The contrarian case is not weak. Dowding treats forward guidance as synonymous with credibility, but credibility is built on delivery, not prediction. The 2021 "transitory inflation" episode proved it: Powell's confident guidance broke publicly and painfully. The damage came not from an unclear path, but from a clear path that was wrong. Rigid promises become liabilities when regimes shift. The code never lies — only the auditors do.
Nor has the debt triggered crisis. Treasury auctions still clear. Buyers show up — they just demand a higher premium. The system is pricing risk, not snapping. Dowding's frame lacks a threshold quantification: how much debt risk already sits inside the term premium, and how much would a genuine information vacuum add? If the market has spent two years repricing fiscal risk, Warsh's communication shift may be a small additional shock, not the trigger for a nonlinear collapse.
I have seen this pattern before. During the 2017 ICO cycle, I audited token contracts and watched projects lose credibility not when documentation was unclear, but when promised logic failed execution. The over-promisers died faster than the projects that said nothing and shipped. Warsh's discretion model — fewer promises, more action — is not an abandonment of credibility. It is a different theory of it.
Takeaway: Watch the Thresholds
The confidence function is binary; it does not decay, it flips. Luna's death was a math error, not a market crash — a peg that relied on arbitrage without the liquidity to close the gap. The Fed's risk is a parallel math error: believing an announced path is the same as a stable path — or, alternatively, that its absence creates no gap at all.
P0 signals, in order: Warsh's first FOMC press conference, the 10-year term premium trend, and the next two Treasury auctions. Break 50 basis points on the premium, print below 2.0 on bid-to-cover twice, and the settlement layer has changed character. A Fed that cannot be priced is the costliest unquantifiable variable in the world's largest market. Patterns emerge only when emotion is stripped away — the emotion here is hope, and the pattern is hard.