Will QE start costing the government money? - Capital Economics
UK Economics

Will QE start costing the government money?

UK Economics Update
Written by Vicky Redwood
Cancel X

There are fears that, by making the government’s debt servicing costs more vulnerable to short-term rises in interest rates, quantitative easing (QE) is storing up trouble for when Bank Rate rises. However, right from the onset of the scheme, it was clear that initial transfers from the Bank of England’s Asset Purchase Facility (APF) to the government would need to be followed later by payments in the opposite direction. The government could still avoid these payments if it wanted. But this is all a long way off in any case.

  • There are fears that, by making the government’s debt servicing costs more vulnerable to short-term rises in interest rates, quantitative easing (QE) is storing up trouble for when Bank Rate rises. However, right from the onset of the scheme, it was clear that initial transfers from the Bank of England’s Asset Purchase Facility (APF) to the government would need to be followed later by payments in the opposite direction. The government could still avoid these payments if it wanted. But this is all a long way off in any case.
  • As a reminder, Chancellor George Osborne decided in 2012 that the cash surpluses and deficits generated by the Bank of England’s QE facility, the APF, would be reflected in the public finances on an ongoing basis. This was instead of them accumulating at the APF, culminating in a one‐off profit or loss to the government once QE was fully unwound and the facility closed. (See here.)
  • Up until now, in what we might call ‘Phase I’ of QE, the government has enjoyed regular payments from the APF. This is because the income that the Bank of England receives from coupon payments on its holdings of government bonds purchased through QE has significantly outweighed the interest (at Bank Rate) that it pays on the commercial banks’ reserves that have been created as a result. (See Chart 1.) Essentially the Government has replaced some of its fixed-rate debt with floating-rate central bank reserves.
  • Chart 2 shows that payments from the APF to the government have been about £10bn p.a. (aside from the initial £40bn in 2013 when surpluses accumulated since 2009 were transferred in one go). So, government borrowing has been about £10bn lower a year than otherwise. Cumulatively, the transfers stand at £109bn.
  • However, at some point, in what we might call ‘Phase II’ of QE, the government will have to start making payments to the Bank of England. This reflects two factors. First, as Bank Rate rises, the interest paid on the reserves will rise, but the income on the Bank’s gilts will rise far more gradually. (Only 5% of its gilts mature in the next year and two thirds don’t mature for at least five years.) So we will eventually reach the point where the interest paid by the Bank of England on its reserves exceeds the income it receives from bond coupons. The MPC plans to raise its gilt holdings to £875bn. With Bank Rate at 0.1%, the interest on the associated reserves will equal just £0.8bn p.a. Say, for simplicity’s sake, that coupon receipts remain at last year’s £16.1bn. Once Bank Rate exceeds 1.85%, the Banks’ reserves payments will outweigh these receipts.
  • Second, the Bank of England’s capital losses will grow as more gilts mature or are sold. Most of the Bank’s gilts were bought “above par” or for a higher price than their redemption value (i.e. at the time of purchase, markets expected the gilts would pay a coupon rate which was high relative to the expected path of Bank Rate). So it was always likely that a shortfall would materialise when the gilts reached maturity or were sold. Note that, even if the Bank of England reinvests a maturing gilt, the amount reinvested is the purchase price of that gilt, with the APF having to make up any shortfall between the redemption payment and this amount. The ONS estimates that the redemption value of the Bank’s gilt holdings at the end of last year was £612bn, £113bn less than the £725bn spent on acquiring the gilts.

Chart 1: BoE Interest Paid & Received Per Month (£bn)

Chart 2: BoE Transfers to HM Treasury (£bn)

Sources: Refinitiv, Capital Economics

Sources: Refinitiv, Capital Economics

  • As a result of all this, former Bank Governor Mervyn King said in 2012 that ”under reasonable assumptions it is likely that the majority of any transfer of funds to the government will eventually need to be reversed”. In 2012, the OBR estimated that the APF would lower government borrowing from 2012/13 to 2016/17, then raise it from 2017/18 to 2022/23 as QE was unwound. In the event, ‘Phase I’ of the APF has lasted a lot longer than envisaged, given that Bank Rate has stayed unexpectedly low and QE has grown, not shrunk.
  • The eventual exact net profit or loss once (if!) QE is wound up will depend on several factors in the meantime, including the path of Bank Rate, when QE is unwound and the behaviour of gilt prices. There is clearly huge uncertainty around all of them. There is a good chance that the government will make a net gain on the APF, given that Bank Rate has stayed lower for longer than was generally priced into markets when the gilts were bought and the yield curve has correspondingly shifted downwards. But a loss would be possible if, for example, the announcement of an unwinding of QE prompted bond yields to rise sharply.
  • Note that even if the scheme made a net loss, that would not mean that QE has been of no help to the government. It has still lowered gilt yields and therefore the cost of new borrowing. Moreover, it has removed the risk of a sell-off on that part of the government’s bonds that are now held by the Bank. In any case, QE has generally not been (at least not openly!) about directly helping the government anyway.
  • So, to recap so far, the prospect of the APF turning in losses in the future is neither surprising nor anything to worry about. Nonetheless, suppose that the government were keen to avoid these “losses” in ‘Phase II’ of QE. This might simply be because it wanted to save the money. But it might also be worried about how it would look, something that has perhaps been vindicated by recent alarmist opinion pieces in the press. Indeed, the Bank’s Independent Evaluation Office warned earlier this year that “there could be reputational risks if any large future cash-flow reversals took any stakeholders by surprise.”
  • So what could be done? Some suggest the Bank could just sell the gilts, extinguishing the reserves. The Bank would no longer have to pay interest on any extra reserves, while the government would revert to paying interest to private sector bond holders. However, because the Bank bought its bonds above par, it would still lock in a loss (as bond prices fall towards their redemption value as they mature). This “loss” is just the flipside of the benefit that buying above par has given the Bank in the way of high coupon payments relative to Bank Rate. So, the government would still find itself making payments to the APF.
  • In any case, selling its bonds now would seriously risk derailing the economic recovery. After all, the Bank is currently buying more bonds to support the economy. It could wait until the economy were stronger, but by then, yields would have risen and the losses that it would have to book would be bigger.
  • Ex-MPC member Charles Goodhart has suggested another solution, namely to stop paying interest on the reserves. However, the ample reserves in the system would push short-term interest rates to zero and lead to the central bank losing control over short-term rates. Goodhart argues that the Bank could prevent the expansion in the money supply leading to an explosion of credit growth by required capital or liquidity ratios. But in our view, that would still leave the problem of the Bank being unable to raise interest rates on reserves to deal with inflationary pressure from the build-up of bank deposits of households and firms.
  • One way around this could be to implement reserve requirements for commercial banks and to cease paying interest on these required reserves. This would prevent banks from using some of their reserves to boost lending, reduce the interest bill incurred by the Bank of England and yet still allow it to tighten policy by raising the interest rate on “excess” reserves. This would not be costless, though. It would be an effective tax on commercial banks i.e. a form of financial repression. (See here.) 
  • Finally, the government could just refrain from transferring any money to the Bank. The Bank could hold a “deferred” asset on its balance sheet. Or it could just recognise the losses, eroding its £4.5bn of capital (if the current framework whereby the government tops up the Bank’s capital were changed). Although the Bank could eventually end up in negative equity, this is not an issue for central banks as it is for companies.
  • So, fears about the threat to fiscal sustainability from the impact of rising Bank Rate on the APF are overdone. But perhaps the key point is that all of this is a long way off. We think that Bank Rate won’t even start to rise until 2026 and won’t reach 1% until 2030. So, for the next few years at least, the government should continue to enjoy the financial benefit of the QE programme. Indeed, the OBR expects a further £60bn to be transferred in total from the APF to the government over the next five years.

Vicky Redwood, Senior Economic Adviser, victoria.redwood@capitaleconomics.com
Paul Dales, Chief UK Economist, paul.dales@capitaleconomics.com