Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
From 2009 until the end of last year, net asset purchases by major central banks — the US Federal Reserve, European Central Bank, Bank of England, and Bank of Japan — totalled about $20tn. That figure must come down. The big question is how far, and how fast.
Following the financial crisis, central banks hoovered up bonds as part of quantitative easing programmes to stimulate demand-sapped economies. Then the pandemic hit, leading to a further bond-buying binge to calm markets. Central banks (with the exception of the BoJ) have been slimming their balance sheets this year via quantitative tightening: letting expiring bonds roll off their balance sheets, and in the case of the BoE, through sales.
When central banks buy bonds from banks, the latter receive a credit known as central bank reserves — the safest and most liquid financial assets. QT reverses the process, reducing liquidity in the system. Still, the Fed’s total asset holdings amount are equivalent to about 30 per cent of the US economy — just under $8tn — and the ECB’s, more than half the eurozone’s gross domestic product.
Maintaining too large a balance sheet leads to heightened financial instability — excess reserves distort the private market for liquidity provision, create dependence on the central bank, and, as Andrew Hauser, an executive director at the BoE, outlined in a recent speech, it can incentivise inappropriate risk-taking.
It can also raise operational and reputation risks for central banks. When interest rates rise, central banks suffer losses on their bond portfolios and pay out more interest on bank reserves created by QE.
“Many central banks are now facing big financial holes, which are politically uncomfortable” said Ricardo Reis, a professor at the London School of Economics. In July, the BoE forecast it would make a net loss of more than £150bn over the next decade as it unwinds QE. Although the cost is covered by treasuries, it is hardly good for the public image of the central banks. The aim of QE should be to calm markets or provide stimulus when rates are already low. If it is not unwound, central banks risk being seen as financing government deficits.
A more trimmed balance sheet also allows central banks to regain “valuable policy space in an environment in which the current large volume of excess liquidity is not needed”, as Isabel Schnabel, a member of the ECB’s executive board, noted in a speech in March. Rates may also need to be pushed higher than would be the case with smaller balance sheets, raising the chance of deeper recessions — particularly if they just ratchet higher with each crisis.
But the trillion-dollar problem facing central banks is how to shrink their footprint without sparking ructions. High government deficit forecasts, particularly in the US, point to an ample supply of government bond issuance down the line. Ongoing QT with bond sales only adds to that supply. This may push yields too high, and lead to something breaking in the economy — the Fed’s QT efforts in 2019 drove market convulsions. Calls to abandon QT are already mounting.
How far central banks should go depends on what is the optimum size of their balance sheets, or the preferred minimum range of reserves as the BoE calls it. “It should be large enough to satiate the demand for reserves,” Reis argues. This means central banks should not slim down to pre-global financial crisis levels — economies have grown and banks’ liquidity needs have risen (as demonstrated by the demands on Silicon Valley Bank following rapid deposit outflows that led to its collapse).
That has made calculating the precise level of the PMRR more difficult. In the US, the banking system’s lowest comfortable level of reserves has been estimated by analysts to be about $2.5tn, compared with more than $3tn currently. This suggests the end of QT is still distant, particularly when factoring in the Fed’s other liquidity facilities.
But there are several complications: can central banks cut rates on one hand while carrying out QT with the other? And for the ECB, QT is complicated by the need to defend “peripheral” sovereign bond yields, stopping them from widening too far from those for other eurozone debt. Given it holds a disproportionate amount of these bonds, QT sales could put pressure on them.
Central banks should, nonetheless, dip their toes, and aim to bring down their holdings to more appropriate levels over the long-run. It will not be an easy process — and the dieting will need to be calibrated, fitting in with monetary and financial policy risks. Perhaps, though, the difficulty of offloading assets will spur a rethink on how generous central banks ought to be with buying them in the future.
