Overview
This report contains information on the Bank of England’s Asset Purchase Facility (APF) for 2025 Q3, describing operations from 1 July 2025 to 30 September 2025. It also contains information about how cash flows between the APF and HM Treasury (HMT) might evolve over time, as well as estimates of government debt issuance cost savings due to quantitative easing (QE). More information on what the APF is and what it does is available in our Market Operations Guide. A short timeline describing the history of the APF is provided as background at the end of the report.
APF operations in the past quarter
This section contains details of UK government bond (gilt) operations conducted for monetary policy purposes during 2025 Q3. It also includes information on gilts lent to the Debt Management Office (DMO).
At its September 2024 meeting, the Monetary Policy Committee (MPC) voted to reduce the stock of gilts held in the APF by £100 billion over the 12-month period from October 2024 to September 2025, comprising both maturing gilts and sales.
Over 2025 Q3, in line with the MPC’s September 2024 decision, the Bank continued with the sale of the APF’s stock of gilts held for monetary policy purposes. A total of three gilt sales operations were run during July 2025. These sales led to a reduction in the stock of gilts held for monetary policy purposes of £3.6 billion. Gilt maturities over the quarter led to a further reduction in the stock of gilts held for monetary policy purposes of £28.3 billion. As of 24 September 2025, the stock of gilts held for monetary policy purposes stood at £558 billion.
At its September 2025 meeting, the MPC voted to reduce the stock of gilts held in the APF by a further £70 billion over the 12-month period from October 2025 to September 2026. In September 2025, the Bank published the schedule for gilt sales in 2025 Q4. As part of implementing the MPC’s stock decision, for the year starting 2025 Q4, the Bank will aim to sell fewer long maturity sector gilts than gilts at other maturities, such that approximately 40% of the MPC’s target is met by selling short maturity gilts, 40% by medium maturity gilts and 20% by long maturity gilts, measured in initial proceeds terms.
Summary of holdings
Table A: Summary of stocks in the APF in 2025 Q3 (a) (£ millions)
Week ending | Gilts (b) |
|---|---|
2025 Q2 (c) | 590,018 |
2 July 2025 | 590,018 |
9 July 2025 | 589,141 |
16 July 2025 | 589,141 |
23 July 2025 | 587,495 |
30 July to 3 September 2025 | 586,401 |
10 September to end-2025 Q3 | 558,069 |
Footnotes
- Source: Bank of England.
- (a) The outstanding amount in each facility is reported on a settlement date basis.
- (b) The overall stock of APF gilt purchases for monetary policy purposes, net of sales and redemptions, valued at initial purchase price.
- (c) 2025 Q2 measured as the amount outstanding as of 25 June 2025.
Chart 1 shows the cumulative net value of APF transactions between the establishment of the APF in 2009 and the end of 2025 Q3.
Chart 1 is separated into two panels with different scales. Gilt purchases and the Term Funding Scheme (TFS) – which from 2016 to 2019 was on the APF balance sheet before its transfer to the Bank’s balance sheet – are on the left panel.footnote [1] The legacy corporate bond and commercial paper schemes that have been operated via the APF balance sheet are shown on the right panel.
Gilt lending arrangement with the DMO
Gilts purchased for monetary policy purposes via the APF continue to be made available for on-lending to the market through a gilt lending arrangement with the DMO. The average daily aggregate value of gilts lent by the APF to the DMO during the three months to 30 September 2025 was £8.0 billion. Chart 2 sets out the average daily value of APF gilts lent to the DMO via the gilt lending agreement over the past two years.
Chart 2: Average daily aggregate value of lending of APF gilts to the DMO
Cash-flow arrangements between the APF and HM Treasury
In line with the indemnification of the APF by HMT, the assets held in the APF generate a range of cash flows which – alongside interest costs and the gains or losses made at maturity or sale – drive consequent cash transfers between HMT and the APF.
Between 2009 and 2022, the APF’s activities generated positive net cash flows from the APF to HMT, peaking at a cumulative £123.9 billion at end-September 2022.
In 2012, it was agreed to transfer the APF’s net income to HMT on a regular basis, as well as the net income that had accumulated since the facility’s inception in 2009. When this arrangement was put in place, it was recognised that reverse payments from HMT to the APF were likely to be needed in the future as Bank Rate increased and as the APF’s gilt holdings were eventually unwound by the MPC.footnote [2] The first such quarterly transfer from HMT to the APF occurred in October 2022 and payments have been made on a quarterly basis thereafter.
A Quarterly Bulletin article in May 2022 explained the mechanics of cash flows and provided an illustrative projection into the future based on prevailing market conditions and the MPC’s policy at the time.
Future APF cash flows are highly uncertain and are sensitive to a number of factors, including changes in Bank Rate. First, Bank Rate affects the interest payment the APF must make on its loan from the Bank – a rising Bank Rate means there is a smaller or negative surplus of income once interest on the Bank of England loan is paid. Second, Bank Rate affects the level of the yield curve which will have an impact on the price received when gilts are sold from the APF to the private sector.footnote [3]
In line with previous APF Quarterly Reports, Chart 3 below provides an updated summary of actual cash flows to date and illustrative projections into the future. Reflecting the considerable uncertainty around future cash flows, the projections are based on a set of scenarios for the MPC’s approach to unwind, reflecting the MPC’s annual review process, and the path for Bank Rate. These illustrative projections are highly sensitive to the assumptions used and are in no way reflective of the MPC’s future considerations regarding APF unwind.
Since the current value of cash flows further into the future is generally lower than the value of cash flows in the nearer term, the net present value (NPV) for past and future projected cash flows are calculated for each of the scenarios to facilitate a comparison between them. Depending on the assumed path for Bank Rate, the illustrative cumulative lifetime NPVs of cash flows in the scenarios considered fall in the range between -£60 billion and -£120 billion. As set out in Box A, fiscal savings are estimated to offset these cumulative cash flows in a range of £50 billion to £125 billion.
In all scenarios, the stock of gilts is assumed to reduce by a total of £70 billion in the year to September 2026, through a combination of maturities and sales, in line with the MPC’s preferred approach to unwind over this period.
In the first scenario (‘Scenario 1A’), from October 2026 onwards APF unwind is assumed to continue at £70 billion per year, in line with the MPC’s approach in the year to September 2026, until the portfolio is fully unwound. In the second scenario (‘Scenario 2A’), the annual pace of unwind from October 2026 onwards is assumed to be the average pace of sales since 2022 (£32 billion per year) alongside scheduled maturities, until the portfolio is fully unwound. For both Scenarios 1A and 2A, it is assumed that Bank Rate follows the market path, as of 30 September 2025.
In line with the approach of previous APF Quarterly Reports, Chart 3 also includes scenarios in which Bank Rate falls gradually over the coming three years back to a level equal to an estimate of the equilibrium interest rate, produced by staff in 2018 and as described in Box 6 of the August 2018 Inflation Report, and then remains at that level for the remaining life of the APF. These two illustrative scenarios – one for each of the two assumptions of the pace of APF unwind as outlined above – are shown as Scenarios 1B and 2B. They are designed to demonstrate the sensitivity of cash flows to different interest rate environments.footnote [4]
Projected annual net cash flows are indicated by the purple bars in Chart 3 for Scenario 1A, while the gold and pink lines show the cumulative cash flows in Scenarios 1A and 2A respectively. The difference in the pace of unwind across the scenarios has a minimal impact on the respective net cumulative lifetime NPV. In both Scenarios 1A and 1B the NPV is approximately -£120 billion, as indicated by the red diamond in Chart 3.
Cumulative cash flows for Scenarios 1B and 2B are indicated by the azure and yellow lines respectively. In both Scenarios 1B and 2B, the NPV is approximately -£60 billion, as indicated by the peach diamond in Chart 3.
The Bank has previously noted that an assessment of lifetime cash flows associated with QE and quantitative tightening (QT) does not amount to an assessment of the overall fiscal, let alone economic, impacts of QE. By lowering interest rates, QE supported the economy and improved the terms at which the Government could raise debt.
Unlike APF cash flows, which are directly observable, there is significant uncertainty around any estimate of the wider economic and fiscal impact of QE. However, it is likely that the fiscal benefits of QE significantly, or fully, offset the net lifetime transfers from HMT to the APF under the indemnity. Box A illustrates one way to estimate part of this impact – the fiscal savings from lower government debt issuance costs as a result of QE – and suggests that these benefits amount to £50 billion to £125 billion in NPV terms. As set out in Box A, this analysis does not take into account the wider benefits of QE in supporting broader economic activity and resolving market dysfunction, both of which would have had significant fiscal benefits relative to counterfactual scenarios of no QE. While many of these benefits accrued in the past, a significant proportion of the fiscal savings is still to be realised, given the long maturity at which the Government issued debt during the period that QE pushed down on yields.
As emphasised in this report, future APF cash flows are highly uncertain, and the ultimate balance between fiscal costs and benefits will shift over time as the APF remains exposed to significant interest rate risk. While QE and QT have been deployed to meet the MPC’s monetary policy objectives, rather than for fiscal reasons, it is important to recognise all the channels through which QE and QT have fiscal implications.
Footnotes
- Sources: Bloomberg Finance L.P. for market rates as at 30 September 2025 and Bank calculations for data in relation to APF cash flows.
- (a) The stock of assets used for the projection of cumulative cash flows is based on holdings as at 30 September 2025, consistent with the holdings reported in Table A.
- (b) In all scenarios, the stock of gilts is assumed to reduce by a total of £70 billion in the year to September 2026 in line with the MPC’s September 2025 decision.
- (c) NPV calculations based on end-September 2025 data.
- (d) The methodology supporting these quarterly projections is subject to regular review.
Box A: Estimating government debt issuance cost savings due to QE
The Bank publishes projections of potential paths of cash transfers between the APF and HMT in the APF Quarterly Report to support public transparency. However, there are several channels by which QE/QT have fiscal implications outside of the cash flows between the APF and HMT. An important channel is how QE allowed the Government to issue debt at lower interest rates, reducing the cost of financing. This box estimates the magnitude of this channel. Another important channel not explored here is how QE supported economic activity,footnote [5] thereby protecting jobs and supporting tax revenues.
Various rounds of QE since 2009 lowered gilt yields, which meant that the Government issued debt at a lower cost than otherwise. During these periods of QE, the Government issued long maturity debt,footnote [6] meaning that the fiscal benefits were locked in for a long time.
Using government debt auction data, it is possible to calculate the average cost of debt financing for the Government and estimate the increased cost that would have occurred in the absence of QE by applying QE yield impact and persistence assumptions (Chart A). The assumed QE yield impacts draw from estimates in the 2022 Q1 Quarterly Bulletin article. The persistence of the yield impact is varied to create two scenarios: (a) full impact for one year, followed by two years of tapered impact; and (b) full impact for four years, followed by two years of tapered impact.footnote [7]
The lower issuance costs due to QE translate to an estimated £50 billion to £125 billion of fiscal savings in NPV terms (Chart B). Only about half of this benefit has been realised to date, given the UK’s long debt maturity structure. Therefore, many gilts remain outstanding at coupons lower than they would have been absent QE.
These results are based on several underlying assumptions. First, there is a large degree of uncertainty surrounding the persistence of yield impacts, so a range is provided based on two scenarios that are consistent with available studies, detailed in footnote 7. Second, the QE yield impacts used are based on yield movements on the days of the announcement – which may understate the yield impact as anticipated effects have not been incorporated.footnote [8] Third, only the impact on conventional gilt issuance has been incorporated, but QE also affected other asset prices including inflation-indexed gilts.footnote [9] Fourth, we have not assumed a change in the maturity structure of issuance due to QE.footnote [10]
This analysis incorporates some small upward pressure on the cost of issuance in recent years from QT.footnote [11] QT is designed to have a lower impact on yields and the economy, so the impact of QT is not the reverse of the impact from QE. As Bank Rate is no longer constrained at the effective lower bound, the implications of higher gilt yields from QT can be offset by a lower Bank Rate as needed to achieve the 2% inflation target. Additionally, the QT impact is reduced by sales being gradual and predictable, and conducted only in appropriate market conditions.footnote [12] This means that the effect of QT on financial conditions, the economy, and the cost of issuance is less than the impact that QE had, by design.
Chart A: Average issuance yield of outstanding debt, with two scenarios showing the estimated average issuance yield impact in the absence of QE
Footnotes
- Sources: DMO and Bank calculations based on end-September 2025 data.
Chart B: Debt issuance saving from QE for two scenarios, in present value terms
Footnotes
- Sources: DMO and Bank calculations based on end-September 2025 data.
The fiscal benefit from lower debt issuance costs is locked in at the time of issuance. In contrast, APF cash flows are highly uncertain – small changes in interest rates can substantially alter the projected lifetime cash flows from the APF.
Chart C shows debt issuance cost savings alongside the APF lifetime cash flows based on current market interest rates and scenarios where interest rates are higher or lower. It also shows the net effect of these two components.
The fiscal savings from lower debt issuance costs are estimated at £50 billion–£125 billion, while the lifetime net cumulative cash flows of the APF is approximately -£120 billion, based on market interest rates as of end-September 2025. The combined effect is in the range of -£70 billion to +£5 billion. In scenarios where market interest rates decrease by 250 basis points or increase by 250 basis points, the combined position ranges from -£135 billion to +£95 billion, driven by the interest rate sensitivity of the APF.
The figures in Chart C are in net present value terms, so the lifetime APF gain or loss projections are little changed depending on the pace of QT unwind. Different unwind strategies impact when losses are incurred but do not necessarily change the lifetime profit or loss accumulated in the APF. Gilt sales, for example, incur upfront costs due to realising losses but have the benefit of reducing lifetime net interest costs from carrying gilts on the APF’s portfolio.
While the unwind strategy of the APF does not necessarily alter lifetime profit or loss, it does affect the exposure to interest rate risk. A longer holding period means greater interest rate risk because there is a higher probability of substantial interest rate changes. A strategy that unwinds APF holdings more quickly reduces interest rate risk faster. The range of possible outcomes – both upside and downside – is much narrower if asset holdings are reduced more quickly, as illustrated in Chart D.
Chart C: Discounted debt issuance cost savings and APF lifetime cash flows, with interest rate scenarios
Footnotes
- Source: Bank calculations based on end-September 2025 data.
Chart D: Illustrative APF cash flows, in present value terms
Footnotes
- Source: Bank of England.
In summary, by lowering interest rates, QE supported the economy and improved the terms at which the Government could raise debt. Estimates suggest that reduced debt issuance costs due to QE have resulted in fiscal savings of £50 billion–£125 billion. A large part of this benefit is still yet to be realised given the UK’s long debt maturity structure. As previously mentioned, the lower cost of government debt issuance is in addition to other fiscal benefits associated with a stronger economy, such as stronger job growth and tax revenues.
APF history and background
Below is a summary of some of the key milestones in the history of the APF since its establishment in 2009. The APF sits in a wholly-owned Bank of England subsidiary company – The Bank of England Asset Purchase Facility Fund Limited (BEAPFF).
- 19 January 2009 Chancellor’s Statement announcing the intention to set up an asset purchase programme.
- 29 January 2009 Establishment of the APF Fund (exchange of letters between the Bank and HMT).
- 9 November 2012 HMT announces the transfer of gilt coupon payments to the Exchequer (exchange of letters between the Bank and HMT).
- 3 August 2016 MPC agrees to expand the APF by launching a Term Funding Scheme (TFS) and a Corporate Bond Purchase Scheme (CBPS) (exchange of letters between the Bank and HMT).
- 21 June 2018 Bank and HMT agree new capital and income framework codified by a new Memorandum of Understanding.
- 21 January 2019 TFS drawings (and collateral) transferred from the APF to the Bank of England’s balance sheet.
- 19 March 2020 MPC agrees to expand the APF with a £200 billion increase to the stock of UK gilts and sterling non-financial investment-grade corporate bonds to reach £645 billion. This was followed by the MPC deciding to expand the APF with a £100 billion increase in June 2020, and then a further £150 billion in November 2020, bringing the total stock of asset purchases to £895 billion.
- 2 February 2022 MPC votes to begin to reduce the stock of UK gilt purchases by ceasing to reinvest maturing assets, and the stock of sterling non-financial investment-grade corporate bond purchases by ceasing to reinvest maturing assets and by a programme of corporate bond sales.
- 21 September 2022 MPC votes to begin sales of the stock of gilts held in the APF. Gilt sales subsequently began on 1 November 2022 with an unwind pace of £80 billion by September 2023.
- 28 September 2022 The Bank announced it would undertake purchases of UK government bonds under its financial stability mandate. Purchases concluded as planned on 14 October. Sales of this portfolio began on 29 November 2022 and were concluded on 12 January 2023.
- 6 June 2023 the Bank announced that it had completed its sales of sterling non-financial investment-grade corporate bonds.
- 20 September 2023 MPC votes to unwind the APF by £100 billion over the period from October 2023 to September 2024.
- April 2024 APF corporate bond portfolio reached full maturity. From this point onwards, the stock of corporate bonds held in the APF was zero.
- 18 September 2024 MPC votes to unwind the APF by £100 billion over the period from October 2024 to September 2025.
- 17 September 2025 MPC votes to unwind the APF by £70 billion over the period from October 2025 to September 2026.
Links to additional information related to the APF
- Exchange of letters between the Bank and HMT, 17 February and 3 March 2009.
- Asset Purchase Facility Annual Report 2024/25.
- Asset Purchase Facility Quarterly Report – 2025 Q1.
- Asset Purchase Facility Quarterly Report – 2025 Q2.
- Exchange of letters between the Governor and the Chancellor on the Asset Purchase Facility – November 2025.
Next publication date: 10 February 2026
ISSN 2041-1936
The Bank launched the Term Funding scheme with additional incentives for Small and Medium-sized Enterprises (TFSME) during April 2020. The TFSME does not appear in this report because it is operated from the Bank’s balance sheet, rather than the APF.
The details were set out in an exchange of letters in November 2012 between the Governor and the Chancellor. This was also explained in a May 2022 Quarterly Bulletin article and reconfirmed in an exchange of letters between the Governor and the Chancellor in September 2022.
Further information regarding how changes in Bank Rate impact the cash flows of the APF can be found in the Bank’s May 2022 Quarterly Bulletin article: QE at the Bank of England: a perspective on its functioning and effectiveness.
Box A of the February 2025 Monetary Policy Report noted that the estimated equilibrium rate was likely to have increased modestly since 2018, but that there was significant uncertainty around the range of estimates at any point and the extent of any increase.
Further information regarding economic stimulus from QE can be found in the Bank’s May 2022 Quarterly Bulletin article: QE at the Bank of England: a perspective on its functioning and effectiveness.
Importantly, the UK Government has issued more long-term debt than those of other major economies. The average term of outstanding UK Government debt is around 14 years, compared to around six to seven years in, for instance, the US, Germany and Canada.
Specifically, we use the following yield impacts for each QE round: QE1 (March 2009): -65 basis points; QE2 (October 2011): 0 basis points; QE3 (July 2012): -11 basis points; QE4 (June 2016): -15 basis points; QE5 (March 2020): -44 basis points. Each with two persistence assumptions: (a) First year 100% impact, second year 50%, third year 25% (consistent with Joyce and Tong (2012) and Braun et al (2023)); (b) and Four years 100% impact, fifth year 50%, sixth year 25% (consistent with Bridges and Thomas (2012)).
As discussed in Bank of England Staff Working Paper No. 980, QE was often not a surprise at the point of announcement.
The impact of QE on inflation-indexed gilts is less certain, but if an equivalent yield impact is incorporated onto inflation-indexed gilt issuance costs, the estimated fiscal savings would increase by £15 billion–£40 billion.
To the extent that QE allowed for longer duration issuance (or prevented a shortening), it meant that lower financing costs were locked in for longer, particularly following the Covid crisis.
The assumed QT yield impact draws from estimates in the Monetary Policy Reports (MPR) for the: August 2023 MPR (10–15 basis points), August 2024 MPR (10–20 basis points) and August 2025 MPR (15–25 basis points) Monetary Policy Reports. In this analysis, a QT yield impact on the issuance yield of auctions is assumed at: QT year 1 (February 2022–August 2023): +12.5 basis points; QT year 2 (August 2023–August 2024): +15 basis points; and QT year 3 (August 2024–October 2025): +20 basis points. An additional year of a +20 basis points QT impact at the same issuance levels as the past year would add about £4 billion to the cost of debt issuance.
Quantitative tightening: the story so far − speech by Dave Ramsden.
