How the Bank of England has wasted billions — and why Australia shows the way forward

Christopher Mahon is a chartered financial analyst and fund manager with almost 30 years’ experience. This is a guest blog for NEF; views expressed are the author’s own and do not necessarily represent those of the New Economics Foundation.

The Bank of England ran up billions of avoidable losses on its quantitative easing (QE) era decisions. And it is compounding the cost today.

During crises including the Covid-19 crisis and Brexit, the Bank bought up government bonds to support the economy via its QE programme. But it did so inefficiently, generating losses far larger than other central banks. Today, in selling those same bonds via quantitative tightening (QT), it is running up further avoidable costs.

It need not be this way, but the Bank insists that all monetary policy, including QE and QT, should be guided solely by the inflation target, not by financial cost or risk.

Disregarding these wider elements has come at a price. Firstly, the Bank lost over 5% of GDP, roughly twice the eurozone figure and four times the US. Secondly, the Bank is currently pressing ahead with gilt sales despite accepting that these sales are pushing up the government’s borrowing costs.

The best central banks take a broader view beyond thinking purely about their inflation target. The Federal Reserve always weighs costs, risks and unintended consequences before acting. Australia, also unhappy with its central bank, has taken note. It has rewritten its framework so that unconventional tools must pass explicit tests covering benefits, costs, risks, and exit plans.

At a minimum, Britain should adopt similar language. Given the Bank’s resistance to change, the Chancellor should consider going further: pausing gilt sales while an independent review examines how best to shrink the balance sheet.

QE losses: How big are they? And how do they compare?

The Office for Budget Responsibility’s most recent estimate puts the cumulative net lifetime loss of the Asset Purchase Facility (APF) (the relevant part of the Bank’s balance sheet) at an eye-popping £164bn.

That does not mean QE was a mistake. It helped support the economy through some of its darkest moments. But a bill of this size makes comparisons unavoidable: could the same support have been delivered at a lower cost? Compared to its peers, the Bank fares poorly.

Figure 1: Losses on government bond QE programmes outstrip other central banks

Cumulative losses on government bond QE programmes since they began (2009 – 2024), valuation losses measured on a mark-to-market basis, plus interest paid out on reserves, % GDP

Not all QE programmes were created equal

Table 1 highlights four choices that help explain why the UK outcome was worse than the Fed’s.

Table 1: Comparison of Bank of England vs Federal Reserve approach to QE

Long dated bonds: a text book ​‘original sin’

Ultra-long gilts, maturing more than 25 years ahead, produced the worst taxpayer outcomes, with some losing 70% or more of their value.

In buying these bonds, the Bank failed to tailor QE to Britain’s credit channels. Much UK borrowing, like mortgages, is priced off short- and medium-term bonds. By contrast, mortgages in the US are anchored to 30-year rates.

Part of the rationale for QE is to stimulate the economy by bringing down the rates households and businesses pay. So, targeting the shorter maturities that UK borrowing is actually priced off would give more stimulus-per-pound spent.

And this extra effectiveness comes with less risk. Short-dated gilts are less sensitive to interest rate changes than long-dated ones. If rates moved against the Bank, the losses on its holdings would be smaller.

Other central banks adjusted their purchases to be more selective. The Reserve Bank of Australia restricted QE to shorter dated securities. The Fed shifted purchases towards specific market dislocations, concentrating on long dated bonds during 2012’s Operation Twist, later moving to short dated securities during the pandemic.

The Bank justifies purchases of long dated gilts by pointing out Britain has more long-dated bond issuance than other countries. But this does not make long-dated purchases inevitable. Greater issuance of inflation protected debt did not lead to the Bank purchasing any of that (another decision of great cost).

By allowing itself to think only about the impact on inflation, the Bank failed to be efficient and focused with its balance sheet.

From QE to QT

Today, the Bank is taking the unusual route of actively selling bonds as part of the process to unwind quantitative easing. This additional supply of bonds to the market pushes up the cost of borrowing to the government – seen as an increase in yields.

While newspaper headlines have highlighted the issue for years, the Bank’s recognition of the issue came slowly.

In September 2022, a poll of specialist investors estimated that gilt sales would lift yields by 25 basis points. It took eight months for the Bank to give its own assessment: just 0 – 10 basis points.

That delay and lowball estimate raised an awkward question: was the Bank marking its own homework?

The incurious MPC

As time went by, independent estimates generally found a larger effect on government borrowing costs than the Bank. This included a high-profile paper co-authored by former monetary policy committee (MPC) member Kristin Forbes, which put the impact as high as 0.70%, around three times larger than the Bank’s estimate at that time.

Rather than re-poll investors to see if these independent estimates reflected a consensus, the Bank stopped asking investors the survey question. Instead, it defended its low estimates of its impact by citing US research, even though the Fed was not undertaking similar sales of bonds.

Today, four years later, the MPC’s midpoint estimate has risen to 25 basis points, finally matching the market consensus from 2022.

Where today’s consensus might be is hard to gauge. Investors have not been given the opportunity to state their owns views on the matter.

At best, this slow adjustment suggests the Bank’s analysis lagged the market by four years. At worst, dropping the question looks like a decision not to hear an inconvenient answer.

The impact on public finances

If gilt sales raise yields, they also raise the government’s borrowing costs.

The government funds its deficit with new bond issuance from the Debt Management Office (DMO). This issuance competes with Bank sales. More supply means higher yields — locked in for the life of each bond.

Since active QT began in November 2022, the DMO has issued around £600bn of new medium- and long-term debt.

On the Bank’s 0.25% estimate, active QT adds roughly £25bn to lifetime borrowing costs. At 0.4%, a mid-range independent estimate, the cost rises to about £40bn.

If the Bank held the bonds acquired during QE to maturity, rather than selling, it would avoid competing with the DMO – and avoid these extra costs. But in allowing itself only to think about inflation, the Bank disregards the financial impact to the government.

What can we do: Learning from Australia

Britain is not the first country to be unhappy the conduct of its central bank. In 2021, after Australia’s messy collapse of ​“yield curve control” – an attempt to hold bond yields at a fixed level – a wide-ranging review produced firmer rules for deploying and unwinding additional monetary tools.

Similar to the UK, the question was not whether QE had been justified, but whether it could have been designed better and with less taxpayer risk.

Ground Rules for QE — Australian style

The Reserve Bank of Australia (RBA) and the federal government periodically publish a single ​“Common Understanding” – which plays a similar role to the Bank’s remit letters.

And here’s what those new rules say about additional monetary policy:

“The Monetary Policy Board will communicate a framework to guide the use of additional monetary tools, including the benefits, costs and risks associated with available tools. It will draw on a range of inputs, including international experience, independent expert assessments and lessons from the Reserve Bank’s use of additional monetary tools.

“The framework will also detail the decision-making approach that the Board intends to adopt when it believes that there may be a need to use or discontinue alternative monetary tools. This includes how monetary policy will work together with other arms of policy. In addition, the framework will set out considerations at the outset about how these additional tools might be exited under different scenarios.”

What this would mean for the Bank of England 

The Australian wording has lots of useful language for the next iteration of the Bank remit letter:

“…benefits, costs and risk…”: This is the crux of the issue. Unlike the Bank, the Fed devoted considerable attention to unintended consequences before any action. chairs Bernanke, Yellen, and Powell all allowed costs, risks and trade-offs to inform their decision making. Perhaps this is a reason why the Fed incurred far smaller losses than the Bank.

By contrast, the Bank’s monetary policy committee regularly point out that financial costs and risk are not considered and that decisions are taken ​“solely” to meet the inflation target.

“…lessons…”: Good institutions learn from the past. Despite losses of £164bn, the Bank has said little about what it would do differently next time.

“…independent expert assessments…”: This would encourage the Bank to keep testing its view against investor assessments, rather than dropping inconvenient questions.

“…working with other arms of policy…”: Today, that means not pressuring government borrowing costs unnecessarily. And for financial stability, not adding to bond market volatility.

“…exited under different scenarios…”: The failure to consider risks or undertake scenario testing sits uneasily with the Bank’s leadership role for prudential regulation and risk management.

Other Australian reforms hold promise for the UK too. For example, Australia also introduced a skills matrix to broaden the RBA’s market expertise.

The wider lesson from Australia: change need not weaken independence.

Why value for money is not enough 

A gentle nudge for value for money has been tried before, without success.

In 2024, the Treasury select committee pushed for greater consideration of value for money in APF decisions. Jeremy Hunt subsequently reminded the Bank that the APF remained subject to ​“the requirement to protect value for money”.

Andrew Bailey’s response, however, qualified this to operational matters only – such as coordinating gilt auctions with the DMO. The Bank successfully rebuffed the idea that the taxpayer consequences of MPC choices should inform policy.

As a subtle push failed, the remit letter needs to go further: detailed RBA-style language.

Adapting to Bank specific issues: efficiency, proportionality, frequency

Incorporating RBA wording into the remit letter is a good start. But additional wording will be needed to tackle UK specific issues:

Efficient: Use the public balance sheet selectively, reflecting Britain’s own transmission channels (such as the differences in mortgage structures in the UK vs US) rather than copying overseas programmes.

Proportionate: The Bank went from running the world’s most aggressive QE platform to running the most aggressive QT platform. Neither extreme was obviously necessary.

Reviewed frequently: Infrequent review has at times meant predictability strayed into inflexibility. The final round of QE locked the Bank into bond purchases until December 2021. Yet by May 2021, CPI inflation was already above target and rising fast. More frequent reassessment may have led to QE being withdrawn earlier, without the extra costs from stimulus when inflation was already above target.

Scrapping the indemnity helps but more is needed

Even with this extra language, there is no guarantee. The Bank may resist again, as it did after Hunt’s intervention. Some commentators therefore push for the scrapping of the financial indemnity the Bank was given by the Treasury when QE was first started.

It is true the indemnity appears to have reinforced the Bank’s belief that financial consequences belonged elsewhere. Removing it would send the opposite signal: the Bank is accountable for the risks it takes. So far, so good.

But the change would probably require deferred-asset accounting, as used by the Fed. If the Bank sidelined value-for-money when losses involved cash transfers, there is no guarantee it would treat deferred liabilities any more seriously.

And in terms of fiscal rules, deferred liabilities still count as part of public sector consolidated balance sheet – although could cost less than issuing gilts to cover Bank losses.

Pause sales and commission an independent review 

A firmer option is for the Chancellor to commission an independent review of how the Bank’s balance sheet should be reduced and to pause active gilt sales while it reports.

A respected former MPC member could lead a focused and swift review, drawing on the different approaches taken by independent central banks.

A temporary, clearly defined pause would not challenge operational independence. And reference could be made to reviews in Australia or to the new ​“taskforces” announced by new Fed chair Kevin Warsh, as examples of similar appraisals conducted elsewhere. 

Conclusion

Every stage of QE and QT involved choices. Some supported the economy; others exposed taxpayers to avoidable risks and costs.

Losses of £164bn justify a harder look at the current framework. Australia shows that reform need not threaten central bank independence.

It is time for a new approach. Over to you, Chancellor Healey.

Image: iStock