The BoE’s QE programme has all the features of a carry trade – borrowing cash to fund volatile positions in long-dated government bonds.
During QE the BoE became far more exposed to the bond market than other central banks with the result that losses, currently at ~ 4% of GDP, are quadruple those of the Federal Reserve.
Now active QT means the BoE is continuing to sell its positions at the worst point in the interest cycle and has missed out on a potential £2.5bn from the recent rally in bonds.
We call for active QT to be stopped at the September MPC meeting.
Last month’s collapse of the Yen carry trade made headlines around the world. Investors were shaken but within days markets started to recover.
But when such carry trade reversals are big enough to impact government finances, it becomes a more serious matter altogether. The Bank of England’s (BoE) quantitative easing (QE) programme is a case in point. The latest official estimates assign a nominal loss of some £130-140bn from the BoE’s QE programme – all of which must be paid for by taxpayers1.
To recap the mechanics of these losses: under QE, the BoE borrowed cash to buy government bonds. This all worked well when cash rates were near zero and locking in a 1% yield on a 30-year gilt looked attractive.
Borrow cheaply. Invest in higher yielding assets. It’s a classic carry trade. And the BoE did so on a scale unsurpassed by other central banks, even extending QE into 30 and 50-year debt instruments.
But fast forward to 2024, with cash rates at 5%+, and the same 1% yield locked in for 50 years isn’t quite such a good deal. In fact, it is costing a fortune as a result of the “negative carry” on those low yielding bonds. This explains the huge losses being racked up by the Bank’s QE programme.
As we’ll explain later on, the BoE expanded QE way beyond what other central banks were prepared to do and the result is the losses at the BoE dwarf other central banks – around 4x larger than the Federal Reserve’s.
What to do when a carry trade goes wrong
When carry trades go wrong investors can do one of two things:
(A) liquidate at the bottom of the market. Normally this happens when an investor has no other option (e.g. your counterparty demands it). In this scenario your own sales combine with sales from other forced sellers. The combined sales will drive down prices further and frequently take place during the period of maximum market turmoil. As an unwind strategy it is, in general, best avoided.
(B) hold on for better prices – an approach that can be taken if the investor has balance sheet control.
Other central banks took the latter approach to unwinding QE. They haven’t sold their bonds and are simply let their holdings shrink passively as each bond reaches maturity. Uniquely, the Bank of England went for option (A) selling a chunky £50bn per year of gilts into the market, at the point of maximum fear of both inflation and further rate hikes. They have termed this policy “active QT” (quantitative tightening) and no other central bank is attempting the same. More simply it is a policy of “buying high selling low”.
Unwinding at the bottom
This pattern of buying high & selling low is illustrated in the chart below. We have used the performance of the FT Actuaries All Gilt Total Return index – a broad gilt index that is similar in profile to the Bank’s own holdings.
The Great Unwind: BoE Purchases and Sales
UK Gilts Total Return Index
With annual totals of BoE Gilt operations overlayed
Source: Bloomberg, Bank of England as at 6 September 2024. Based on FTSE Actuaries UK Conventional Gilts Total Return Index, rebased to 100 in Dec 2021.
Now, with inflation cooling, interest rate cuts, and the recent equity market turbulence, gilt prices are marching upwards. Suddenly, the Bank of England gilt’s sales, at the low prices of the past two years, are looking rather unwise.
Had the Bank of England simply held to maturity (like other banks) the value of their holdings would now be around £2.5bn higher (net of financing costs). And potentially, gilts should continue to bounce back if the Bank of England keeps on track to cut rates further.
So it is likely that the decision to sell their holdings will look more regrettable as time passes.
On top of this, we believe selling £50bn of gilts each year at a time when fear about inflation has been at the most severe is pushing down on the prices of gilts, worsening the overall losses. It also raises the cost of debt to the government – by around 0.4% per year or around £1bn in extra interest costs on our estimates (see here).
The most costly QE programme in the world: QE programmes compared
Comparing losses between the different central banks is not easy. While each institution publishes data, it does so on a different accounting basis, often with substantial delays in reporting, sometimes in a piecemeal fashion.
To get around this, in our measurement, below, we have simplified matters. We focused on the QE government bond holdings. We know what type of bonds each Bank bought, on what dates, and know the performance of those bonds since then. We use a mark-to-market estimation and, while not the same as each central bank’s own accounting standard, it has the advantage of treating each central bank in a comparable fashion. And we can run the report each day.
This approach means we can estimate the performance of each central banks QE programme, all on a like to like basis, and see the impact of the latest market moves as they happen – without having to wait for official reports.
The results are not favourable for UK taxpayers. Bank of England losses are quadruple that of the Fed.
Estimated losses on QE programmes, % GDP
(Using mark to market valuations, includes cost of finance, cumulative 2009-24)
Source: Columbia Threadneedle, data through 6th Sept 2024
The losses are the worst for the UK because of certain BoE decisions:
- The BoE chose to buy the largest amount of government bonds
BoE ownership peaked at 37% of all gilts, vs 19% for the Fed’s ownership of all US Treasuries - The BoE chose to own far longer maturity bonds which fell more when rates went up.
Under QE, the BoE purchases included 30 and 50-year gilts that locked in the low interest rates for much longer. Other central banks focused towards shorter maturities, that were in turn believed to be more relevant for providing stimulus to the private sector. The BoE never changed tack and instead maintained its longer profile throughout the 12-year period of QE. Others such as the Fed adjusted to ever shorter maturities as time went by. - The BoE locked in the losses by selling at the worst point in the interest rate cycle. Some of the 30-50 year gilts the BoE bought have been sold with losses of ~ 70%2 in just three years. Other banks have held to maturity and seen their QE holdings rally in recent months.
- The BoE chose not to include inflation protected bonds which fared much better than traditional gilts- unlike the Fed which included such securities in their purchases under QE.
It is too late to salvage most of the losses. But even at this late hour, stopping active QT and the BoE gilt sales could help improve the picture modestly.
The incurious MPC
So far, there is little recognition by the Bank of how their choices have differed to other central banks or how these decisions have worsened the financial outcomes.
QE is little talked about by the Monetary Policy Committee (MPC). By contrast, the various members of the MPC are each willing to make speeches highlighting their different views on inflation or interest rates. But when it comes to the QE programme, few questions are asked, and few differences aired.
QE decisions are nearly always approved by unanimous votes (unlike with interest rate decisions). Yet these unanimous decisions with little public debate have led to losses that dwarf other central banks.
We will never know if alternative approaches, more akin to other central banks, were considered.
Where to from here?
For some time, we have been warning that the Bank of England’s decision to sell gilts would mark the low point of the gilt cycle – just as the gold sales of the early 2000’s marked the low point of the gold cycle. Similar preannouncements of sales are used, depressing prices. Similar disinterest is expressed by the BoE in the prices achieved and/or the scale of losses crystalised.
This might be about to be change. In one of his final acts as Chancellor, Jeremy Hunt, reconfirmed (despite protestations from the Bank) that taxpayer value should be a consideration in decisions regarding the Bank’s QE programme. In September, the MPC is due to re-appraise its approach to QE. We have estimated their approach to unwinding has added (all in) around £5-7bn in costs this year alone3.
With the recent surge in gilt prices, and the prospect of interest rate cuts next year, the “selling at the bottom” nature of active QT should now be evident to all. In our view, it is time to stop the policy. Instead, hold until maturity should be the mantra for the MPC.
