Where quantitative tightening meets the taxpayer

At noon on September 17th the Bank of England did something easy to miss inside the noise of an interest-rate decision: its Monetary Policy Committee voted 9-0 to change how it shrinks its own balance sheet. Bank Rate stayed at 3.75%, and inflation, at 3.1%, sits stubbornly above the 2% target and rising on energy prices tied to the Middle East conflict. The quieter news was the plan for quantitative tightening (QT). The Bank will now unwind its £488bn of government bonds at an average pace of roughly £46bn a year, down from £70bn. It will stop selling the longest gilts altogether. And it will suspend all sales for six months, until April, while it consults the Treasury.

To the lay reader QT sounds like plumbing best left to central bankers. In fact it is where monetary policy and fiscal policy collide, and the new plan is the fullest admission yet of who has been paying for it.

The Bank is unique among big central banks in selling its bonds outright rather than waiting for them to mature, a quirk of the unusually long average maturity of British government debt. That means it must turn into cash, year by year, the gap between the depressed prices at which its crisis-era purchases were struck and the higher yields at which it now parts with the paper. That gap is a realised loss, and by law the Treasury underwrites it. The Treasury does not conjure the money from nowhere: it borrows it, selling more gilts into the same falling market. The Bank's own estimate is that QT has added about a quarter of a percentage point to gilt yields; some analysts put the effect at the very long end nearer three quarters of a point. Either way, the tightening had begun to feed on itself long before a global sell-off pushed British 30-year borrowing costs to their highest since 1998.

Hence the peculiar arithmetic of the new plan. Of the £488bn held, some £222bn of debt maturing by 2034 will now simply run off. Another £120bn of the longest bonds, maturing in 2049 or later, will remain on the balance sheet permanently to back banknotes; these are exactly the issues the Bank will stop selling. The residual, about £146bn, will be sold at roughly £20bn a year. Even here the detail tells: sales will be made directly to the Debt Management Office at market prices rather than through the Bank's own auctions, and none at all before the Bank has talked to the government.

Governor Andrew Bailey argues that the change adjusts the timing of losses, not their total size. In an accounting sense that is true. But timing is the politics. As Patricia Pino, an economist, puts it, the cost may be "neutral" across six decades, yet governments live on parliamentary timelines, and the immediate, billion-pound cash demands of realising losses are what a quiet squeeze was designed to soften. This is the point the Reform party, no friend to the Bank, has been hammering: that a monetary choice keeps showing up as a fiscal intervention, with the Treasury settling losses that turn yesterday's easy money into today's bill.

The institutional reading is more uncomfortable than the accounting one. The Bank insists the move is about risk and clarity, not easing. But a six-month pause "while it consults the Treasury," and sales handed to the DMO, shift the boundary between an independent central bank and the government it is meant to stand apart from. QT existed to make the system live with the consequences of the Bank's own crisis purchases, to force a re-run of the ledger without a friendly borrower in the room. Now the holder and the borrower are quietly coordinating how to unwind in a way that spares the public finances. The fiscal tail has found the monetary dog.

The lesson travels well beyond gilts, and well beyond Britain. The episode is a stress test of what happens when any central bank must shrink a large crisis-era balance sheet in a higher-rate world: the losses do not disappear, they are reassigned to the taxpayer, and they resurface as long-end borrowing costs. American investors hold the same question, mostly at one remove. The Federal Reserve—which raised rates the very previous day—has largely let its holdings run off passively, letting bonds mature without the act of crystallising a loss, whereas the Bank is selling into the open market and booking the damage in full. Whoever holds long-duration bonds or rate-sensitive equities is holding some version of the question: who bears the cost of turning easy money into tight money?

The 9-0 unanimity lends the plan the air of settled policy and the closure of an orthodox loop. It does something else. It registers the limit of an approach that could not shrink the balance sheet and repair the public finances at once. Central-bank independence was once defined as the power to say no to the government. This plan is the Bank effectively asking the government for help in stopping it from hurting both of them. Retreat in central banking is rarely announced as what it is; it arrives as a unanimous vote to consult. The next recession will require that balance sheet to grow again, and the gap between buying cheap and selling dear is now something every member has agreed to carry on the taxpayer's account, year after year, until the yield curve says otherwise. It will have its chance to say so.