At a Glance
- The BoE’s approach to QE has led to losses of almost 5% of GDP vs less than 2% in the US
- After the election, Treasury curiosity for a fresh approach becomes more likely
- We expect active bond sales to be scrapped – providing a boost to the gilt market
Back in 2021, when the Bank of England (BoE) announced the end of quantitative easing (QE) and the beginning of quantitative tightening (QT), few commentators understood how out of kilter the Bank’s balance sheet had become compared to central bank norms. The result has been losses on the BoE bond portfolios that dwarf those of other central banks.
To date, politicians have been wary of questioning the Bank but the election of a new government could serve to reenergise the Treasury. In our view, a new government is likely to desire a different approach and the obvious solution is to adopt central bank norms. If this happens it could be good news for gilt investors.
Under QE, the Bank of England became far more involved in government bond markets than other central banks
Adjusting for the relative stock of government debt, the BoE peaked at roughly twice the involvement in the government bond markets as the Federal Reserve. It ended up owning 37% of all gilts issued – against 19% for the Fed. At peak, the BoE’s holdings even dwarfed the ECB’s – remarkable since the ECB had the additional objective of supporting the Euro.
The BoE has been the most aggressive during both QE & QT eras
Share of total government debt held by the central bank
Source: Bloomberg as at 31 May 2024
Only the Bank of England is now actively selling government bond portfolios into the market
Now quantitative easing is in reverse and central banks are reducing their balance sheets. Most central banks are shrinking their portfolios passively – simply letting their bond holdings mature. By contrast, the BoE is actively selling into the market.
According to the Bank of England, this difference doesn’t matter. By contrast, we believe the effects are significant – both for taxpayers and investors.
Differences in both QE and QT are significant drivers on government borrowing costs
In our view, the BoE switching from being a large and persistent buyer to becoming a large and persistent seller is one of the primary reasons for gilts underperforming.
The impact of the sales is particularly evident in the interest costs on long term government bonds. This is compared to the Fed, which is not actively selling Treasuries, below.
Borrowing costs have gone up much more for the UK than the US since QT began
Government bond yields, %
Source: Bloomberg as at 31 May 2024
Since the UK started its QT programme, long-term yields in the UK have risen by 80-100bps relative to Treasuries. Notably, much of this rise took place before Q3 2022, excluding the period of the mini budget.
Our assessment is that active sales have added about 0.4% to the cost of government debt. This comes from absorbing a large amount of additional supply at once. Our opinion differs from the Bank’s own view which assesses the impact at ~10bps or less.
Our view is lent weight by the fact that even small announcements in the pace of QT have caused big shifts in the corresponding yield levels. Examples include the BoE December 2023 news that it would rebalance away from long-end sales or the May 2024 Fed statement slowing the pace of its balance sheet run off. Both saw significant positive reactions from the market.
Our view is also supported by the market impact from QT as witnessed during the BoE sales of corporate bonds, when spreads widened in the UK around twice as much as global benchmarks, again excluding the mini-budget period).
The bottom line is that QT has a greater impact than some think, making the September MPC meeting a potentially significant event.
An opportunity to change tack
Officially the Bank of England is keeping its options open. Governor Bailey and Deputy Governor Ramsden have both indicated that scaling back QT in September is not a given.
By contrast, we think the scale of the losses will force a change.
This will be a change. So far, the Bank has explicitly avoided allowing the losses to drive policy. Deputy Governor Ramsden has reiterated: “The MPC does not take into account financial risk or profit when taking monetary policy decisions, including about the gilt portfolio1”.
In the Bank’s view, QE and any resulting losses were necessary to carry out their inflation fighting mandate and such losses are for the Treasury to deal with, due to the famous indemnity negotiated back in 2009. While critics highlight the Bank has other objectives beyond inflation fighting, including being careful stewards of public money2, after the mini budget, the Treasury did not want to press the issue.
An election has the potential to change the status quo, whichever party wins. Neither will want to govern with the headwind of QE bills of £5-10bn every few months. A new approach will be required. The Bank could offer up a partial solution: cancelling the active sales and slowing the pace of QT. In any case, optically, changing tack like this is easier once interest rates cuts starts. Active sales tighten monetary conditions while rate cuts which loosen them.
But while the BoE is not committing to scale back active QT, in the background pressure is building. The scale of the losses relative to other central banks is raising questions. Maybe it is time to adopt central bank norms.
The scale of the losses so far
Due to the differences in reporting and accounting, it is hard to compare the QE losses between central banks. For example, other central banks still made losses, but now treat those as a deferred liability. By contrast, the UK Treasury uniquely has to fund the losses with cash.
Outline estimated losses on QE government bond programmes, %GDP
Approx. mark to market estimates of asset returns + cost of finance, 2009-2024
Source: Columbia Threadneedle as at 31 May 2024
Losses on government bond programmes are much higher for the BoE as:
- The BoE bought the largest amount of government bonds, with ownership peaking at 37% of all gilts, vs 19% for the Fed’s ownership of US Treasuries
- The BoE owned longer maturity bonds which fell more when rates went up. This again is a design difference unique to the BoE. While the BoE adopted a fairly static maturity profile of 12-14 years, other central banks were focused towards shorter maturities more likely to benefit private sector borrowings. Additionally, the Fed actively managed their profile to between 6-9 years, crucially reducing maturities before rates went up4.
Some of the longer maturities that were the focus of the BoE give rise to significant individual losses. As an example, during QE in May 2020 the BoE bought a series of 2061 gilts at a price of £101. Under QT it has been selling the same securities at prices as low as £28, a loss of over 70%5 in under 3 years.
- The BoE chose not to include inflation protected bonds which fared much better – unlike the Fed which included such securities in their purchases under QE.
The opportunity for fiscal savings presents an opportunity for the next Chancellor
In April, a little noticed letter from the Chancellor to Andrew Bailey reminds the BoE that “minimising cost and risk” is in fact part of the BoE’s remit6. The Chancellor also reiterated the BoE include “value for money” as a consideration in decisions about the pace and timing of QT. After the election, expect this point to be a focus. Value for money can only really mean one thing: the current approach is scaled back.
It is too late to salvage most of the losses but we think scaling back active QT could still reduce overall costs for the taxpayer -albeit modestly. These savings would come from 3 sources:
- Lower interest costs: We estimate that active QT is elevating interest costs by 0.4%. Should active QT be stopped, we expect interest costs on government debt to come down. Last year for example the government raised £237bn of gilts. The interest saving could be worth £1bn per year.
- Avoiding the market indigestion effect involved in active QT: Active QT costs taxpayers due to the market having to absorb so much additionally supply at once. The 0.4% impact on yields means the proceeds from the 50bn sales programme are about 3% less than their natural fair value. This has cost taxpayers about £1.5 to £2bn per year. If active QT continues next year, the impact will might naturally be smaller7 but could still be worth ~£500m.
- Don’t sell now – wait for gilt prices to recover: The biggest – but more speculative – saving could come from waiting. Today’s gilt prices are especially low, having had to absorb both inflation and rate hikes. Loss on individual bonds have been as much as 70%. There could be little harm from holding the positions for longer, and waiting for prices to improve, as rates are cut. Waiting for rates to come down by 0.5%-1% would have seen this year’s £50bn active sales programme generate £2-4bn in proceeds. Next year’s smaller active sales programme would still benefit to the tune of £0.5-1bn in proceeds. Our calculations show active QT is costing taxpayers an additional £5-7bn in 2024 and is on course to cost another £2-2.5bn in 2025.
All this adds up. The savings means the BoE’s current approach –different in nature to central bank norms and costly to taxpayers – is likely be curtailed.
Could the BoE bond sales mark the bottom of the gilt market?
We think the BoE will have to end active QT in September. The Bank may present this as natural at the commencement of a rate cutting cycle. In reality their hand is likely to be forced by the costs of BoE decisions relative to other central banks. This is creating pressure on the Bank to fall in line with the Fed and the ECB, who designed their programmes very differently. During QE they did not let their risks multiply as much. Now, during QT they are now not actively selling their government debt. Even at this late hour, a return to these central bank norms could be worth £2-3 billion next year to the Treasury.
We believe the disappearance of the biggest seller of gilts from the market should spur significant outperformance of UK government bonds. History rhymes with the past. The UK central bank actively selling their huge holdings of gilts may well come to be seen as marking the nadir of the rate cycle, just as in the early 2000’s when Bank sales of gold marked the nadir of the precious metal cycle.
