In a September 2025 article, the then backbench former Labour Transport Secretary, Louise Haigh, but now First Minister to prime minister, Andy Burnham, argued that the UK should follow the European Central Bank (ECB) and Swiss National Bank practice and pay tiered or partial interest instead of full payment (renumeration) on central bank reserves.
This was because “subsidising the banking sector at full interest while cutting investment or welfare spending is economically incoherent” while moving to tiered reserve remuneration “would not compromise monetary policy”.
This year, the September 2026 Monetary Policy Committee decision (September 2026 MPC QT decision) to replace the 2022-26 annually decided Quantitative Tightening (QT) programme with a slowed multi-year fixed programme, will mean that all government bonds (gilts) held for monetary policy (MP) purposes are sold or redeemed at par by the end of 2034, rather than by 2071 at a slowed average annual £46bn rather than £87.5bn pace (see section one and Table one below for detail, explanation and context).
That decision combined with continuing elevated BR levels, given the interaction between asset sale and continuing high fully renumerated interest remuneration losses, will result in an increased £20bn-plus annual hit to the public finances, skewed and concentrated in the short term towards interest losses that count against the current budget fiscal rule.
The APF deficit specifically attributable to the gap between reserve interest renumeration paid and gilt interest received by the APF, as reported in the 2025-26 Annual APF Report (page 22) was c£12.5bn across its financial year, ending February 2026.
Last October, Time to recycle reserve interest income for public purposes (ASocialDemocraticFuture, October 2025) concluded that:
- The Bank’s mandate should be revised (probably best through the Chancellor’s annual remit letter to the Governor) to ensure that the Bank and Treasury jointly take responsibility for the future public finance implications that arise from the Bank’s monetary policy (MP) operations, most pertinently and presently with reference to the operations of the Asset Purchase Facility (APF) concerning active gilt sales (noting that Treasury representative attends MPC and their contribution – limited to public finance implications – should be minuted).
- QT, in words of the governor himself, has limited or tangential impact on the Bank’s MP operation. Active APF gilt sales should be halted or proceed only when future net public finance impacts are minimised on proper and due assessment (probably as best advised by UK Debt Office).
- Quantitative Easing (QE) should only be used in future when its expected impacts are clearly identified and justified relative to possible alternatives and should be implemented taking regard to its future impact on the public finances.
- The House of Commons Treasury Committee should step up its scrutiny and review capacity pertaining to the interactions and inter-relationships between fiscal and MP.
- The Bank itself should develop and improve its communications to explain its operation of MP in lay understandable language more honestly and transparently.
This post revisits Louise Haigh’s argument and recommendations one to two of the above in the light of the MPC September 2026 QT decision and subsequent analyses by Barclays Research (Barclays) for the Institute of Fiscal Studies (IFS) 2026 Green Budget.
Section One provides explanatory background to the technicalities already introduced.
Section Two outlines monetary policy (MP) operational independence versus public value for money (vfm) considerations.
Section Three summarises the public finance impacts and outlines possible alternative policy alternatives.
Section Four makes concluding observations, including that a more commercial and public vfm approach to gilt sales combined with a shift to partial rather than full renumeration of reserve interest could lessen future total losses in cash and net present terms, such as putting the Bank under the obligation to consider public vfm when making choices among options to which the MPC is otherwise indifferent (i.e. in terms of the implications for monetary conditions and, hence, the outlook for inflation relative to the MPC’s target.
However, John Healey’s September 2026 letter to the Bank’s governor does not suggest any willingness of the new Burnham-led government to go down that road.
While, the Bank’s effective (in contrast to a theoretical or protect the citadel) credibility in discharging its statutory independent remit should not be undermined, a strong case exists for fiscal and monetary policy coordination to be improved in a way conducive to the sustainable maintenance of bond market confidence.
That does, however, presuppose and requires the government to adopt and communicate an overarching macro-economic and public finance framework attached with a compelling and credible medium to long term vision.
It should encompass, not only more customised and transparent institutional arrangements pertaining to the operation of the Asset Purchase Facility (APF) but fiscal institutional reform designed to make public expenditure choices and management more efficient and sustainable.
1 Background
Quantitative Easing (QE) involved the large-scale purchase of gilts from the private sector to improve financial sector liquidity, confidence, and lending. It was undertaken across five phases between March 2009 and December 2021 (see Table One below).
Quantitative Tightening (QT) involves actively selling sales of gilts into the market before their redemption date) and passively redeeming gilts at their maturity, not replacing them.
Between 2022 and 2026 it was undertaken at the pace that an annual September MPC decision set for the following October to September year.
Successive QE and QT programmes, including the future profile set out in September 2026 MPV QT decision (set in more detail below), are catalogued in Table One.
A dedicated special purpose body, the Asset Purchase Facility (APF), owned by the Bank of England (BoE or the Bank), with its losses indemnified by the Treasury, was established to hold and manage the total £875bn government (gilts) it cumulatively purchased under successive QE it has undertaken by autumn 2021.
This stock of purchased gilts was of varying maturity (average 13-14 years duration until their redemption (repayment) date) and were supplemented by a very small amount of corporate bonds.
To back the APF stock of purchased gilts, the Bank issued central bank reserves to the commercial banks (CBs) remunerated at the prevailing bank rate (BR) decided by the Bank’s Monetary Policy Committee (MPC) in line with its inflation remit (see section two).
Privately held fixed-rate government debt (gilts pay a bi-annual interest payment (coupon) fixed when issued) were thereby swapped for publicly APF held floating-rate obligations on the reserves created consequent to the gilt purchases it made, variable with BR.
BR changes are thus a crucial determinant of the relative relationship of the flows in and out of the APF that determine whether it is making a surplus profit or deficit loss (see below).
The market value of the gilts held by the APF also vary inversely with BR, among other factors, including their maturity duration.
The swap worked in the favour of the APF and hence the public purse during the period when BR was at historically low levels and exceeded the fixed gilt coupon payments that the APF paid out.
The Treasury received quarterly remittances of profits reflecting that APF interest surplus. By the autumn of 2022, such remittances totalled around £124bn – a boon to the public finances.
The situation then flipped as BR steadily rose and as the Bank began QT through successive 2022-26 annually decided programmes.
In September 2026, the MPC decided to replace its 2022-26 annually decided programme of gilt sales with a multi-year fixed path plan to unwind (active and passive sales) all APF gilts held within the APF for monetary policy (MP) purposes by end 2034 (MPC September 2026 QT decision).
It comprises four main components:
- From April 2027, the unwind will take place at a slower pace compared to the annual average of the previous annual programmes: QT will be eased, in other words (see Table One for further detail and context).
- Of the total c£488bn stock of gilts still held in the APF in September 2026, £222bn with redemption dates earlier than 2035, no longer will be actively sold (that is before their redemption date): instead, they will be redeemed at their issue price (paid at par), as and when they mature (passive redemption or sale).
- Nor will now the £120bn of ultra long term gilts the APF holds (with 2049-71 redemption dates) be sold; instead, they will be retained to back current and future bank note issuance.
- Only the remaining £146bn worth of gilts, with redemption mid-maturity dates between 2035 and 2049, will continue to be actively sold.
The next two sub-sections recaps and looks at in a bit more detail, the two separate but related factors that drive APF profit and loss flows, that the Treasury received until autumn 2022 and subsequently now indemnifies by payments to the AFP, respectively.
Reserve interest remuneration
From the onset of QE until the autumn of 2022, total gilt interest (coupons) received by the APF on its purchased gilt holdings exceeded the interest that it paid to the Bank.
This was because BR (and thus the reserve interest remuneration rate), in the wake of the GFC and then the pandemic, remained close to zero when the average interest on gilts held within the APF was roughly two per cent, allowing APF surpluses to cumulate, as stated above close to £124bn by autumn 2022.
Their quarterly remittance back to the Treasury created greater fiscal space for successive coalition and conservative governments, whose failure to recycle this positive public finance flow into additional productive investment was an economic mistake for the UK that perhaps rivalled Brexit.
As BR rose in response to inflationary pressures driven by post-Covid supply chain frictions, compounded in the UK by the Truss debacle, BR remunerated reserve interest broke through the break-even level determining whether gilt interest (coupons) received by the APF on its purchased gilt holdings exceeds (surplus) or falls short of the interest (deficit) that it pays to the Bank.
Consequently, as the disclosed in the 2025 APF annual report (page 22) the APF deficit resulting from remunerated reserve interest outflows exceeding gilt interest received reached an eye-watering c£44.7bn across the two financial years, ending February 2025, reflecting annual interest outflows exceeding £30bn.
Given these emerging and massive outflows, Louise Haigh’s proposal to move from full to partial or tiered reserves remuneration was certainly not off the wall.
Paul Tucker, ex-deputy governor of the BoE, in an 2022 IFS Green Budget chapter (Tucker 2022) almost pre-echoing her, had earlier pointed out, according again to the Bank’s then own figures, that if partial remuneration of reserves had been implemented across the then coming 2023-25 period, between £30bn and £45bn could have been saved for the public purse,
That was, he noted, roughly nine per cent of the annual public spending on health, on education, and on defence – a prediction not out of kilter consistent with the official outturn figures cited above.
According to Tucker, “it was, or ought to have been” clear to both the Treasury and the Bank that the combination of QE and paying interest on the additional £440bn additional reserves created as the result of the CovidQE5 gilt purchases (see Table One), presented a clear future risk to public finance sustainability.
As an aside, the government by issuing long-term bonds to the market could have alternatively locked-in historically unusually low long forward rates – below one per cent – to fund pandemic-related deficit funding.
Instead, the APF, in effect, borrowed at a floating rate from the Bank to purchase QE5 Covid-related gilt purchases.
This meant that the Treasury had taken on the public finance risk that BR over the life of APF held gilts would rise above their long forward fixed coupon rate – a risk that always risked becoming acute with any return to inflation as, in fact, did and continues to occur, and should have been foreseen.
Partial or tiered renumeration does not to end reserve interest renumeration completely, but only on the proportion of the reserves required to be maintained for monetary policy (MP) purposes.
Instructionally, as way of wider past policy context, the Bank did not renumerate reserves until 2006. Rather, it implemented MP by keeping them ‘scarce’ deliberately to maintain a mechanism to transmit BR decisions across the financial system.
Since 2022, however, the BoE has been transitioning to an operating system where the quantity of reserves is determined by demand for them from mainly the clearing or commercial banks (CBs) at any prevailing BR, rather the reserves supply matched to ADP gilt holdings formally held for MP purposes.
That transition will complete when AFP gilt holdings fall to a level consistent with CB demand.
Then, the Bank will simply lend CBs the future reserves that they require through the already established ‘repo’ arrangements – that point, where such repo-lending backs reserves rather than APF gilt holdings reflected in the BoE balance sheet, is much closer now than it was in 2022.
Bailey and his Bank colleagues have also continued to point out the possible difficulties of correctly dynamically specifying the proportion of reserves that should be remunerated at BR or not: that is without possibly undermining MP purposes, largely dependent, as they are, on the prime BR transmission route, while it still relies on changes to interest payable by the BoE on CB reserves transmitting to other short term interest rates.
In sum, on one hand, until 2029-30 the AFP outflow requiring Treasury indemnity support counting against the current budget rule, as modelled by Barclays Research in its 2026 Green Budget chapter, attributable to reserve interest renumeration, is almost certain to continue above £12bn annually, underscoring the case for a move to partial renumeration; on the other, on the face of it, technical changes on the Bank’s reserves operating system appear to militate against such a move.
QT gilt sales capital losses
The other side of the renumerated reserve interest equation and its public finance impact are APF asset sales and the capital losses.
Since autumn 2022 such loses have been crystallised by QT, selling gilts at market prices less than their APF purchase price (simply put, buying high, selling low).
Some, if not most, purchased APR gilts were originally issued when interest rates were lower than prevailing post-2022 levels; as rates then rose, the capital value of such gilts (mark to market price) fell to align their effective interest rate to the prevailing rate payable (yield) on newly issued and traded gilts of the same duration maturity.
Indeed, the official gilt yield series catalogues that the 30-year gilt yield was less than two per cent between July 2016 and May 2022, falling to 0.6% in May 2020.
A similar gilt issued in September 2026 was a c5.34% yield.
The May 2020 bond would need to fall in market value nearly ninefold to pay the same effective interest rate.
That is an extreme simplified example, but the director of the NIESR in a September 2026 blog (Aikman blog) reported that the QT programme of annual gilt sales, which was conducted since 2021 to September 2026, involved the APF selling gilts of £112bn nominal value (value if held until redemption) for cash proceeds of roughly two thirds of their original APF purchase value (c£115bn).
Partially halting active sales (proportion) on the longest dated gilts and leaving the shortest date gilts to mature and redeem at par will certainly lessen crystallised capital losses on the longer dated gilts.
These currently carrying in some cases up to 70% ‘mark to market’ capital losses relative to their issue and par price, although some of the shorter dated gilts may have been purchased at below their par price, which will mean that they still incur capital losses when they are passively redeemed (maturity losses).
As Adam Lawrence in his blogs on QT has consistently argued, selling long-dated gilts into a weak market with few natural buyers was always an expensive way to undertake QT in public vfm terms.
The Debt Management Office (DMO), subject to confirmation, will arrange from April 2027 onwards for the APF 2035-49 gilts that the MPC has decided still should be sold to be cancelled and then replaced with shorter duration gilts (maturity date less than fifteen years, likely in the five-to-10-year range, subject to DMO finalisation and confirmation) of the same value: a maturity swap, neutral in consolidated public sector accounting terms.
These shorter duration replacement gilts will be issued and sold into the market outside the APF.
The government will then bear interest risk sensitive to BR on the shorter dated replacement gilts issued to replace more sensitive to BR.
It will also take on refinancing risk, where and if the market becomes unwilling to purchase the replacement debt.
The replacement gilt interest payable to the new private gilt purchasers/holders will then count as a general government revenue cost, no longer subject to the APF indemnity, while the APF no longer will receive the interest previously payable on the cancelled gilts.
The future slower pace of QT will also mean as a counterpart, as the previous sub-section noted, that reserves (and hence AFP remunerated interest paid on a more slowly reducing level of reserves) will likewise reduce less quickly than they previously were expected to.
2. Monetary policy (MP) operational independence v APF public value for money (vfm)
The Bank’s statutory prime and overriding remit is to maintain price stability. This is calibrated to an operational MP target to maintain a symmetrical (not too low or high) inflation rate of 2%, measured by the 12-month increase in the Consumer Price Index. A supporting but subsidiary objective is to support the government’s economic policy, including its objectives for growth and employment.
Separate and subject to its statutory remit, the chancellor in successive exchange letters with the governor, has stated that the Bank should minimise the lifetime cost and risk of the APF (public vfm) that the Treasury indemnifies in full.
In practice, that has meant the volume and composition of APF asset purchases (QE) and sales (QT) has been set by the MPC in line with its primary price stability remit, at least ostensibly; it is only the implementation of MPC decisions relating to APF public vfm, process liaison with the Treasury and DMO and reporting, that is expressly delegated to the Bank Executive.
It was discussions between the Bank’s Executive and the Treasury and Debt Management Office (DMO) that decided implementation of the September MPC QT decision (subject to finalisation and confirmation in April 2027) would occur through selling the 2035-2049 APF gilts directly to the DMO, rather than into the secondary gilt market.
These new arrangements, should, according to the governor’s letter (see previous reference), should maximise value for money by minimising cost and risk over the lifetime of the APF, “but subject to achieving the MPC’s chosen unwind target and in line with the MPC’s key principles”.
The pace and composition of QT unwinding, therefore, will be delivered, as the September MPC set out (as Table One and section one defines), whether the new DMO procurement route is confirmed or not in April.
The September 2026 MPC minutes reported that the benefits of a fixed intended pace of unwind and its gradual predictability was considered by members to outweigh any resulting loss of flexibility and would allow “the MPC to take QT into account better when setting BR as the active instrument of monetary policy while also limit(ing) the impact of the remainder of QT on gilt yields beyond the initial announcement effect” (M36 and 37).
It reaffirmed the core three principles underpinning QT that sales are conducted so to: maintain BR as the prime active determinant of MP; not to disrupt the functioning of financial markets; and in a gradual and predictable manner.
The minutes went on to uphold the overarching principle that the MPC should decide the stock of gilts held for MP purposes, something that “remained a key feature of the UK’s economic framework and essential for the effective delivery of monetary policy” (M39).
A September 2026 speech by Deputy Governor Ramsden, a very senior Treasury and Bank insider involved strategically and operationally across all successive QE and QT phases, and a current MPC member, underlined that “to ensure monetary independence, and to prevent any blurring of fiscal and monetary responsibilities”, gilt sales, including their pace and frequency, should be pre-defined and non-discretionary decisions by the Bank.
He advised that a secondary QT purpose to sell gilts into the market could be to increase long term interest rates through influencing term premia (relative spread between short and long term gilt rates), so reducing pressure to increase BR.
Or, as the NIESR Aikman blog put it, QT implementation choices that materially affect gilt yields and financial conditions should not fall outside the MPC’s formal authority and accountability.
Such formulations, however, do not totally fit with the longstanding official summary stated Bank view that unlike QE – used as a MP mechanism to reduce interest rates and therefore support inflation – the aim of QT is not to affect interest rates or inflation nor to signal MP intentions: it is, rather, to provide balance sheet headroom (normalisation) to undertake QE again in future, should that be needed to achieve the Bank’s inflation target.
The transmission mechanisms of QE and its overall efficacy on the real economy are not well understood or certain and that of QT very much less so, as the Bank itself recognises.
Indeed, a February 2024 Treasury Committee report (para 69) noted that the Bank and MPC do not consider QT to be an active tool of monetary policy and think that it has minimal economic and financial impacts, concluding that “That being so, while it is right that MPC members should have monetary policy and the inflation target foremost in their thinking and decision-making, the Bank and Treasury should explore how criteria on value for money and the spending power of the Treasury could be included in decisions about the ongoing pace and timing of QT”.
It further noted that maintaining the APF gilt portfolio at its then present level (i.e. neither passive nor active QT) would slow the losses and even reverse in the very long term, with a significantly lower lifetime net present value of loss, insofar that “additional bonds may be bought below par so that gains are realised at maturity, which gradually translates into a positive effect on the overall cashflows” (points relevant to the policy options that the next section, especially Slowing, pausing or ending APF gilt sales combined with institutional change, discusses.
In that light, it does seem that the MPC September 2026 QT decision is over- emphasising the importance of its independence to determine the pace and composition QT for MP purposes and its inflation remit.
The MPCs decision to slow QT and make it more certain rather than condition contingent makes more sense if its prime purpose is to underscore that QE is temporary and to normalise the Bank’s balance sheet, thereby protecting the potential long term monetary ‘firepower’ of future QE rounds if that was deemed necessary, than it does from an on-going operational MP perspective.
The next section examines whether wider public policy considerations are being given sufficient weight relative to the Bank’s apparently over-zealous defence of its operational independence in conducting QT as an integral part of its MP remit.
3 Public finance and policy implications
According to Barclays Research APF March 2026 estimates, made prior to the September 2026 MC QT decision, total APF deficit (interest and sales) losses would have averaged close to £20bn until 2030-31 (footnote four).
Following that decision, it now estimates that total APF annual losses will be around five billion higher across the 2026-28 period than they otherwise would have been but then decrease from 2028-29 onwards.
Their distribution, however, will skew towards interest losses in the short term.
The APF deficit specifically attributable to the gap between reserve interest renumeration paid and gilt interest received by the APP, as reported in the 2025-26 Annual APF Report (page 22) was c£12.5bn across its financial year, ending February 2026.
Annual APR interest losses, according to Barclays Research (see figure 4.2) estimates, are likely to continue to hover around £12 to £13bn between 2026-28 before slowly reducing into the five billion to ten billion range thereafter: almost double the levels that the Office of Budget Responsibility (OBR) estimated back in March, as is reported in Table Two.
The OBR can be expected accordingly revised them upwards when its budget-related outlook is published later this week (week ending October 30, 2026).
The disparity underscores the uncertainty of distribution flows within and out of the APF and their sensitivity to BR movements and APF operational decisions, including the pace and composition of gilt sales.
The precise temporal (over time) impact and distribution of APF losses between the different fiscal rules is even less certain and clear.
Both Barclays and the NIESR expect the September MPC QT decision to increase pressure on the current budget fiscal rule.
Barclays Research estimate that current budget ‘headroom’ will be reduced by £3.4bn while it should slightly help the supplementary debt rule.
The essential fact that should not be lost sight off is that APF losses (to be met by Treasury transfer under the terms of its indemnity) must be accommodated by either higher public borrowing or taxes or lower spending.
Put another way, if APF losses could be reduced, greater fiscal space would be created to avoid cuts in public services driven by short-term fiscal pressure rather than sustainable reform drivers, or to undertake long term spend-to-save investment programmes, or to reduce the most economically harmful taxes, or to reduce borrowing.
The pressing question is how the adverse public finance impacts of the APF can be best ameliorated without effectively, in contrast to theoretically, jeopardising BoE monetary independence.
Below some of the main alternative policy options are briefly defined.
Reducing the public cost of reserve interest renumeration
The case for only partially renumerating BoE reserves created as a by-product of QE, whether though a shift to a tiered structure applied by the European Central Bank (ECB) and others, or a dedicated levy, as proposed by the Institute of Public Policy and Research (IPPR) publication Fixing the Leak in August 2025, has been made by many, including by Louise Haigh in her September 2025 previously referenced article.
The recent uplift in the actual and expected short-term cost of fully renumerating reserves puts it into renewed sharp profile.
Moving to a tiered or partial remuneration of reserves is more customised to cause and remedy, establishing also a workable precedent for the future, if QE had to be or was repeated.
However, the Bank’s shift to a demand-driven repo reserves process, as section one explained, which is now well progressed, muddies the waters (as section one touched on) for such a change in renumeration treatment, possibly limiting it in terms of both time and proceeds generated.
Also, the argument can and is made that the resulting loss of reserve interest income would be passed on by the CBs affected to their customers through higher charges or reduced savings interest payments.
That, however, is by no means certain given the oligarchic competitive structure of the banking industry.
Tucker predicted in his 2022 IFS Green Budget paper (see previous reference) that as BR rises, full remuneration of reserves might not be passed through to customers but go to equity holders (and managers) instead.
Introducing a tiered-reserves scheme accordingly “would undo a transfer to bankers and shareholders rather than impose a tax on banking intermediation” (section 7.6).
Recent reported rising levels of reported CB profits, as BR has risen and remained elevated, has tended to spotlight that point.
Tucker also did point out, however, that if the supply of banking services is assumed more elastic than demand – since banks’ behaviour is motivated by a drive for profits not by need – CBs could seek to mitigate the resulting hit to their revenues by, indeed, passing on the reduced level of reserves renumeration to borrowers or to depositors.
An associated risk of some disintermediation of deposits into the less regulated non-bank sector not subject to reserve requirements also should be considered.
Another possible argument against such treatment, as the last section touched on, is that belatedly switching to partial rather than full renumeration of reserves income within a continuing longstanding (25 year) QE/QT programme (on the basis that QT is the reverse counterpart of QE) would undermine the future credibility of any future QE operating regime and its sustainability.
Yet the Bank could announce that any future QE programme would be subject to a partial or tiered reserve renumeration if new emerging economic conditions required the supplementation of its new demand-derived repo approach (see section one).
The issue is more whether the Bank wants or is prepared to do that and within what institutional process it occurs.
The alternative public revenue-raising option is more general increased taxes or reduced reliefs on the profits the wider domestic and international banking sector make in the UK.
This seems on current indications to be the most likely option for adoption in the October 2026 budget.
It avoids any interaction with the Bank’s MP remit and is possibly more amenable to a political in contrast to a technocratic narrative.
On the other hand, it could prove difficult/problematic to achieve comparable proceeds to ending full interest renumeration (proceeds from increased bank taxation of around one to two billion have been suggested), at least without possible unintended implementation consequences, such as the threatened exit of some international bank operations, depending on the magnitude and duration of the increased rates.
Truth told, incidence outcomes are uncertain but as Dan Neidle in a recent Tax Policy Associates post suggests, probably would involve shareholders (front-loaded) and customers ultimately sharing the hit over different timescales.
Nor would it be so well targeted towards those financial institutions securing windfall gains from full reserves renumeration, nor set a sustainable template for any future iterations of the QE process.
Neither approach would address directly capital losses on APF gilt sales.
Slowing, pausing or ending APF gilt sales combined with institutional change
Some private sector analysts, such as Adam Lawrence, warned for a long time that QT practice, taking little or no account of both commercial and public value-for-money considerations, would lead to a massive lifetime cumulative net loss for the APF.
It, the OBR estimated in March 2026, could reach £164bn by the end of 2036. The actual outcome is uncertain, depending, as earlier noted, on future BR and gilt yields, as well as the pace and composition of future gilt sales.
According to their reading, although the September 2026 MPC QT decision was a significant step in the right direction, insofar that the longest ultra-long gilts will no longer be sold and held instead to maturity as backing for currency issue.
The 2035-49 gilts, including some carrying substantial capital losses, however, will still be sold into the market.
As an alternative scenario to the September 2026 MPC QT fixed plan, such continuing losses could be possibly avoided by the APF selling such gilts into the market only when it can make a profit or at least to avoid a loss, in sharp and painful contrast to selling long-dated gilts already depressed in price into an over-supplied market.
It could be possibly combined with carefully targeted measures of fiscal cooperation, involving, in effect, working with the DMO to buy long term gilts cheap and sell them expensive, as market conditions arise, as did occur in 2022.
Mahon, in a September 2026 New Economics Foundation (NEF) paper, highlighted that APF cumulative valuation losses measured on a mark-to-market basis, plus interest paid out on reserves, exceeded five per cent of gdp by 2024, double that of the Federal Reserve or the ECB.
Noting the Bank’s resistance to change, he recommended that the chancellor should commission an independent review of how the Bank’s balance sheet should be reduced.
This should take account of the different institutional arrangements that other countries, such as Australia, have adopted to reconcile monetary policy and public vfm considerations.
Meanwhile, active gilt sales should be paused until the review reported and its conclusions are reflected in the annual chancellor to governor remit letter.
Tucker in 2022 had earlier proposed the Bank should be put under the obligation when making choices among options to which the MPC is otherwise indifferent (i.e. in terms of the implications for monetary conditions and, hence, the outlook for inflation relative to the MPC’s target) to consider public vfm.
He pointed out then had an obligation of that kind existed, public resources could have been saved without impairing monetary-system stability, suggesting that a carefully drafted version of such an obligation might usefully be added, together with codified hurdles for monetary financing, when the MPC remit is updated in the annual governor to chancellor exchange of letters.
The official Bank view set out by Governor Bailey and others is that crystallised gilt sale losses represent simply the future cashflow foregone by such sales discounted at today’s rates, making the timing of the losses largely irrelevant, except in accounting terms.
Indeed, Ramsden in his September 2026 speech (see previous reference), reiterated that the lifetime costs of the APF unwind are neutral, whether sales start with the longest or shortest maturities, and went on to suggest that the Bank Executive should sell the £146bn of APF gilts designated for sales to the DMO in reverse order, starting with the longest maturity gilts, as that would reduce interest rate risk on the Bank’s balance sheet the most quickly.
But public tax and spending decisions and fiscal rule compliance are not calibrated in lifetime present value terms, certainly with reference to the primary current budget rule.
And, as Barclays Research points out, a choice to spend £20bn on APR indemnity transfer now or later is a similar choice as to whether to bear finance public spending through current taxation or borrowing funded later by taxation.
Even more pertinently, as discussed above and in the preceding sub-section, a more commercial and public vfm approach to gilt sales combined with a shift to partial rather than full renumeration of reserve interest could reduce total losses in cash and net present terms.
Moving the APF back into the BoE balance sheet
Islam, Khan and Meaning, 2024 proposed lifting the APF Treasury indemnity altogether, putting the remaining APF gilt holdings onto the Bank’s own balance sheet, more consistent with United States Federal Reserve and other international practice.
Losses would then have to be borne on the Bank’s own balance sheet rather than by the Treasury and public purse directly.
Current losses would be offset in an accounting sense by expected deferred income from the Bank’s future seigniorage activity, derived from its power to create currency and reserves, allowing it to smooth realised losses over a longer timeframe.
Presumably, asset sales would proceed in accordance with the September 2026 MPC QT decision. If so, the Bank’s balance sheet would bear the financial cost of that and any future QT decisions.
Against that some argue that a loss-making central bank could prove more vulnerable to political influence. Stiff resistance from the Bank to such a change can be expected.
4 Concluding comments
The technical interlocking complexity of many the issues covered discourages the drawing of definitive conclusions, other than given their economic and distributional implications, maximum effort is devoted to translating technical issues into informed public choices.
In that light, the following summary observations, drawn from the preceding sections, are offered:
1. The September 2026 MPC QT decision combined with continuing elevated BR levels, given the interaction between asset sale and continuing high fully renumerated interest remuneration losses, will result in an increased annual £20bn-plus hit to the public finances, skewed and concentrated in the short term towards interest losses counting against the current budget fiscal rule, notwithstanding that it should help to limit APF total losses across the longer term, all other things remaining the same, notably future bank rate (BR) movements.
That, however, will be at cost of the transfer of short-term interest rate risk directly to the government.
2. The ostensible emphasis given by that decision to the operational independence of the MPC to set and determine the pace of QT appears overblown and inconsistent relative to the limited and questionable and continued uncertain monetary policy (MP) and wider economic impacts of QT.
The decision appears, in practice, to have taken some limited account of public vfm considerations, whether in response to the behind-the-scenes Treasury pressure and/or political/informed public media concern.
It partially ended and slowed the selling of longer dated APF gilts below their purchase price into a weak market when market gilt prices was depressed as a consequence of BR rises, crystallising high-multiple capital losses; on the other hand, as per (1), in the absence of a shift to partial/tiered interest renumeration of reserves, it will redistribute and concentrate public finance cost flows towards the short term.
3. Lack of Bank understanding or sensitivity to the real-world public finance implications of APF operations, beyond the technical procurement process of selling gilts efficiently, almost certainly driven by (2), remains apparent, however.
This is demonstrated by its doubled down commitment to the lifetime net present value-based interpretation that the timing of losses is immaterial – an assessment giving no quarter to possible mechanisms that could reduce the overall net public finance cost of QT, both in short-term and lifetime terms.
4. A more commercial and public vfm approach to gilt sales combined with a concurrent shift to partial rather than full renumeration of reserve interest could potentially reduce total APF lifetime losses in cash and net present terms.
Such a treatment would appear better targeted towards those financial institutions securing windfall gains from full reserves renumeration compared to the more general taxes the 2026 budget is likely to introduce and would also set a sustainable template for any future iterations of the QE process, if found necessary.
It would provide an opportunity for the Bank to announce that any future QE programme, if new emerging economic conditions required the supplementation of its new demand-derived repo approach (see section one), would be subject to partial or tiered reserve renumeration.
5. Such a shift, however, indicates changes in institutional arrangements pertaining concerning along the lines that Tucker 2022 and Mahon 2026 proposed, including that the Bank should be put under the obligation when making choices among options to which the MPC is otherwise indifferent (i.e. in terms of the implications for monetary conditions and, hence, the outlook for inflation relative to the MPC’s target) to consider public vfm arising from APF operations. .
6. When all is said and done, the Bank’s effective (in contrast to a theoretical) credibility in discharging its statutory independent remit should not be undermined.
In this website’s view, an imperative and scope exists, however, to improve fiscal and monetary policy coordination in way conducive to the sustainable maintenance of bond market confidence.
That, however, presupposes and requires the government to adopt and communicate an overarching macro-economic and public finance framework attached with a compelling and credible medium to long term vision that encompasses not only (5) but focused fiscal institutional reform designed to make public expenditure choices and management more efficient and sustainable.