Central banks may be accidentally subsidizing government borrowing through their efforts to prevent a repeat of past market blowups, and policymakers are starting to worry that anticrisis lending facilities could even be interfering with their own monetary policy.
The source of the problem is the switch from central banks being the lender of last resort to, in 2008 and 2020, also being market makers of last resort, ensuring corporate—and government—debt markets keep functioning. During a crisis, support is often essential to prevent a downward spiral that destroys the financial system.
But backstopping markets removes a key risk and encourages more borrowing—especially for the hedge funds that now own trillions of dollars of U.S. Treasurys.
“Ironically, vulnerability is created by mechanisms that were introduced to reduce vulnerability,” said Huw Pill, the Bank of England’s chief economist, one of those growing concerned, in an interview. “So, it’s a bit like a whack-a-mole kind of story.”
Offering either an explicit or implied guarantee that government-funding markets will remain open and liquid means hedge funds have less risk of being unable to finance highly leveraged trades. This is particularly true for the overnight repurchase, or repo, market, where borrowers pledge bonds for cash. The result has been a huge expansion of two popular government bond trades, arbitraging Treasurys or British gilts against bond futures or swaps.
The scale is extraordinary: The Dallas Fed estimates hedge funds ended last year with $2.4 trillion of Treasurys, up from $600 billion a decade earlier. Because the profits on each trade are tiny, hedge funds have to leverage as much as 100 times to get worthwhile returns, creating new risks.
This might sound abstruse. But in 2020, it was the Treasury basis trade blowing up that forced the Fed to intervene. In 2025, signs of trouble in the swap trade pushed President Trump to retreat from his tariff plan.
Pill worries that the reassurance central-bank policy provides bleeds into monetary policy by boosting borrowing. This, in turn, keeps government-debt yields lower than they otherwise would be.
“There’s lots of gilts to be bought,” he says. “How do you support that buying of gilts? You make it attractive. How do you make it attractive? Well, there are some imperfections in the market. So those imperfections create profit opportunities, but they’re not very big. So how do you make them more meaningful? You allow leverage to build up.”
“That’s good for the government because it gets to sell the gilts at a lower [yield] than it otherwise would. It’s good for the financial sector because they’re able to extract these rents effectively. And it’s good for the central bank because the market seems to be liquid and functioning. But all of those things are true until they’re not true.”
When it goes wrong, the more leverage, the worse the problem. And the worse the problem, the more likely it becomes that central banks have to create yet more special tools to address it. That then spurs the next buildup of leverage.
Pill thinks more effort is needed to come up with a modern version of the Bagehot Doctrine. Walter Bagehot, the 19th-century editor of the Economist magazine, summed up the role of the central bank as being to lend to banks freely, against good collateral, at a penalty rate. Access to instant cash helps banks withstand runs. The fact the central bank is offering a backstop should make the run less likely, and shareholders are penalized, through the penalty rate, if it is used.
Tools for saving markets from drying up are more haphazard. In 2020 the Fed, BOE and others just bought lots of government debt to inject liquidity into markets. That worked because, even though quantitative easing is also a monetary policy tool, they also wanted easier money.
Unfortunately, that created what Pill described as a tinderbox, ignited by the energy crisis after Russia invaded Ukraine. The excess money creation from left over from emergency QE then fanned the flames of inflation. This made it much harder to calibrate monetary policy when central banks decided to tighten (although policymakers were also, in my view, far too slow to recognize inflation).
Pill points to the “temporary, targeted” BOE buying of gilts amid the forced selling by leveraged pension funds after Britain’s botched tax-cut plan in September 2022 as a successful model. At a time when the BOE was trying to tighten monetary policy, it intervened in a way that stopped the pension fund selling spiral and stabilized gilts. Yet the central bank maintained tight monetary policy.
Bagehot would recognize the goal: Reduce the encouragement to take risk, known as moral hazard, that offering guarantees in advance creates, but retain the ability to mount a rescue in a crisis.
Unfortunately, much of central banking is going backward on this. Moral hazard is increasing, even for banks. In the 2023 bank bailout, the Fed accepted less-than-full collateral, recognizing Treasury bonds at face value rather than their (much lower) market value.
The emergency rescue facility then became a funding facility that even healthy banks chose to tap—in effect easing monetary policy by the back door and prompting the Fed to tighten the terms before it ended. Something similar could be under way with Japan’s plans to use an emergency Fed loan facility to raise cash to prop up the yen without having to sell its hoard of Treasurys.
I don’t know how to break the cycle of crises needing rescues that lead to more leverage and new crises. And I’m concerned we’re firmly into the added-leverage phase of the latest cycle.
At least central bankers are still thinking about it, even if they don’t, so far, have good answers.
