For nearly four years, the Bank of England had treated the reduction of its balance sheet as a relatively straightforward process: allow some bonds to mature, sell others, and decide each year how quickly to continue.

That framework has now disappeared.

On September 17, the British central bank set, for the first time, a multi-year path designed to eliminate by September 2034 the government bonds it still holds for monetary-policy purposes. The portfolio will continue to shrink, but much more slowly, with fewer market sales and no liquidation of the longest-dated securities.

The decision therefore goes beyond a technical adjustment to quantitative tightening. It redefines what it means to exit fifteen years of quantitative easing.

The Bank of England is no longer trying to reconstruct the balance sheet it had before the 2008 financial crisis. It is now defining the balance sheet it intends to retain after QE.

In total, the Bank still holds around £488 billion of gilts acquired through its successive asset-purchase programmes.

Those securities will now be divided into three groups.

Around £222 billion of bonds maturing before the end of the programme will simply be held until redemption.

Another £146 billion of intermediate-maturity securities will be sold gradually, at an annual rate of £20 billion.

Finally, £120 billion of the longest-dated bonds will no longer be sold at all. They will remain on the Bank’s balance sheet to support the present and future issuance of banknotes.

This last category represents the deepest break with the previous framework.

These bonds were originally purchased through extraordinary monetary-policy programmes. Yet they will never be sold as part of quantitative tightening. In effect, their function changes: from assets created by QE, they become structural assets on the central bank’s balance sheet.

The Bank can therefore say simultaneously that it will reduce to zero the gilt portfolio held “for monetary-policy purposes” and that it will retain £120 billion of securities purchased during the quantitative-easing era.

There is no contradiction.

The perimeter has changed.

A Much Slower QT

The new architecture sharply reduces the pressure exerted by the Bank of England on the gilt market.

Once maturing bonds are included, the monetary-policy portfolio will shrink by an average of roughly £46 billion a year through 2034. Active sales will account for only £20 billion annually.

The contrast with previous years is significant.

After expanding its securities holdings to almost £895 billion at the peak of its QE programmes, the Bank began reducing the portfolio in 2022. The annual pace of quantitative tightening later reached £100 billion before being cut to £70 billion.

That system of annual decisions is now being abandoned.

In its place comes an eight-year timetable giving investors much greater visibility over the amount of debt the central bank intends to return to the market.

The Bank is also suspending gilt auctions until April 2027 and no longer plans to conduct active sales of very long-dated bonds.

That decision comes at a time when this segment of the British government-bond market has become particularly sensitive.

Long-term yields have risen sharply in recent years under the combined effects of inflation, heavy government issuance and structural changes in demand from institutional investors.

A few days before the Bank’s decision, the yield on the UK 30-year gilt had reached 5.96%, its highest level since 1998.

When the government is issuing large amounts of debt while the central bank is simultaneously selling bonds accumulated during QE, both institutions are effectively approaching the same market for buyers.

The government needs investors for newly issued bonds.

The Bank of England needs investors for the bonds it already owns.

The new framework reduces that overlap.

The Bank itself estimates that quantitative tightening explains only part of the increase in term premia since 2022. Its research attributes roughly 20 to 30 basis points to its own balance-sheet reduction programme, against a much larger increase driven mainly by global financial conditions, heavy sovereign issuance and structural changes in British demand.

But when a market becomes more fragile, a few dozen basis points can be enough to make the mechanism considerably more expensive, financially and politically.

The market reaction to the announcement illustrated the point: long-dated UK government bonds rallied after the decision, while their yields fell.

A More Porous Boundary Between Debt Management and Monetary Policy

The shift could go further still.

The Bank of England, the UK Treasury and the Debt Management Office are now examining whether traditional auctions could be replaced by direct sales to the government.

Under such a model, the Asset Purchase Facility would sell its gilts at market prices. The Treasury would then instruct the Debt Management Office, the agency responsible for financing the British state, to acquire them.

The Monetary Policy Committee would continue to determine how many bonds should leave the monetary-policy portfolio. But execution could become much more closely integrated with the broader management of public debt.

The mechanism has not yet been finally adopted. A decision is expected before April 2027.

Its logic is nevertheless clear.

Today, the Bank can sell an old long-dated gilt at the same time as the Debt Management Office is issuing a new bond to finance the government. Two streams of public debt can therefore reach investors simultaneously even though, economically, both originate from different parts of the British public sector.

Closer coordination would allow the DMO to manage that supply more consistently with market demand.

But it also raises a delicate institutional question.

Quantitative tightening belongs to monetary policy.

Debt issuance and debt management belong to fiscal policy.

The more closely their operational mechanisms converge, the more explicitly the distinction between the two must be preserved.

That is why the Bank has emphasised one principle: the Monetary Policy Committee will retain control over the size of the monetary-policy portfolio regardless of the mechanism ultimately used to dispose of the bonds.

Normalisation Without Going Back

The change is all the more striking because interest-rate policy itself is not becoming accommodative.

On the same day, the Bank of England kept its policy rate at 3.75% by a six-to-three vote. The three dissenting members wanted to raise it to 4%, reflecting inflation risks that the committee now sees as increasingly tilted to the upside.

The message matters.

A central bank can judge it necessary to keep interest rates high, or even consider raising them, while simultaneously concluding that rapidly shrinking its balance sheet is no longer desirable.

The two instruments do not perform exactly the same function.

The policy rate remains the primary instrument for adjusting monetary conditions.

The balance sheet is increasingly becoming an infrastructure to be managed over a much longer horizon.

That distinction marks an evolution in the debate over how to exit QE.

When central banks began buying government bonds on a massive scale after 2008, the programmes were presented as exceptional. The implicit assumption was that future normalisation would eventually erase their footprint.

The British experience now suggests something different.

Assets can be sold.

The reserves created through QE can decline.

Emergency programmes can end.

But the monetary system that emerges from them does not necessarily resemble the one that existed before.

The Bank of England will retain a substantial stock of government bonds on a lasting basis to support banknote issuance. It will also accept that the disappearance of the remaining monetary-policy portfolio may take until 2034.

The balance sheet inherited from the crisis is therefore becoming, in part, the ordinary balance sheet of the post-crisis era.

The Cost of the Exit

The shift also reflects a fiscal reality that has become increasingly visible.

Many of the bonds bought during the era of extremely low interest rates lost market value as yields subsequently rose. Selling them crystallises those losses inside the Asset Purchase Facility, whose financial results are guaranteed by the UK Treasury.

Slowing the reduction of the portfolio does not necessarily eliminate the economic cost of QE. It does, however, alter the timing at which some losses are realised and reduces the need to sell long-dated securities in an unfavourable market environment.

The issue has become politically sensitive in the United Kingdom because transfers between the Treasury and the Bank now make very tangible a mechanism that appeared almost abstract when interest rates were close to zero.

It would nevertheless be misleading to interpret the new strategy as a purely fiscal manoeuvre.

The Bank says its approach is primarily based on market functioning and on the objective of reducing its portfolio predictably without causing excessive disruption.

It also retains a safeguard clause.

The timetable can be changed if the MPC concludes that movements in the policy rate are no longer sufficient to achieve the inflation target, or if financial markets become severely dislocated.

In other words, the path is designed to be durable, but it is not irreversible.

The British Precedent

The Bank of England had been one of the major central banks most willing to conduct outright sales of the government bonds accumulated during QE.

It is now confronting the practical limit of that strategy: reducing a balance sheet worth hundreds of billions of pounds cannot be separated from the market that must absorb the assets.

The greater the supply of government debt, the more selective long-term investors become and the more volatile yields are, the more important the boundary between monetary normalisation and debt-supply management becomes.

The September 17 decision therefore does not mean that London is abandoning the exit from quantitative easing.

It means that the exit itself has been redefined.

The objective is no longer to sell every bond purchased during the crises in order to recreate the balance sheet of the pre-2009 world.

It is to distinguish between what genuinely needs to disappear and what has now become part of the central bank’s permanent architecture.

By 2034, the monetary-policy portfolio created by QE is supposed to have disappeared.

But part of its legacy will remain.

And that may be the real conclusion of fifteen years of extraordinary monetary policy: when an exception lasts long enough, exiting it does not necessarily take the system back to where it started.

It creates a new equilibrium.

Main sources: Bank of England, Monetary Policy Summary and Minutes, September 17, 2026; Bank of England, Asset Purchase Facility: Gilt Sales — Market Notice, September 17, 2026; Reuters, September 17, 2026.