Written evidence submitted by Professor David Miles (MON0054)

Longer Term structure of the Bank of England balance sheet:

At the evidence session with the Treasury Committee on March 1st there was a discussion of the criteria used by the Bank of England in judging the appropriate composition of its much expanded stock of assets held. This brief note sets out some longer term issues that arise.

The Bank of England’s balance sheet is massively larger than it was before the financial crisis; it has grown from around 5% of GDP to over 25% of GDP. Most of the growth reflects the acquisition of a portfolio of around £400 billion of assets (overwhelmingly gilts) bought under the policy of quantitative easing and which sit in the Asset Purchase Facility (APF). The counterpart liability to these purchases is the much expanded stock of reserves held by commercial banks at the Bank of England, nearly all of which earn interest at Bank Rate. Returns on the portfolio (coupons and realized capital gains at point of sale) in excess of Bank Rate will accrue to HMT; losses are indemnified. The Bank of England’s capital is not at risk. The Chancellor must give permission for further purchases to be undertaken. Decisions on the composition of assets and on the timing of any sales sits with the Monetary Policy Committee. Those decisions are made to achieve the aims of monetary policy as set out in the MPC’s remit.

At some point the appropriate stance of monetary policy will return to a more normal level. As we approach that point the stock of assets held for purely monetary policy purposes (the QE assets) will decline. Ultimately one would expect the APF to be run down, quite possibly to zero. But it is not likely that the size of the Bank of England balance sheet will go back to where it was before the financial crisis. It is likely that commercial banks will want to hold greatly more reserves than they did before the crisis. If that is so, and assuming the Bank was willing to accommodate the much larger demand for reserves, the Bank’s balance sheet could be some hundreds of billions of pounds greater than it was 10 years ago. One could view this process of adjustment – in aggregate terms – as assets being sold by the APF and bought by the Bank to sit on its balance sheet. But the composition of assets held, and the implications of gains or losses on them, would likely be very different.

The duration (or interest rate) mismatch between assets held in the APF and the reserves that fund them is very great. Assets in the APF have a duration of around 12 years; the reserves pay Bank rate so have a duration close to zero. This generates interest rate risk which is inevitable given the purpose of the fund – which is to influence the prices of a range of longer-maturity assets so as to boost demand in the economy to better achieve the aims of monetary policy. That would not be the same when assets are held simply as a by-product of providing reserves for banks that value the unmatched safety and liquidity of deposits with the central bank. And the asset allocation decision would not be taken by the MPC but by the Bank executive.

It is not likely that a high degree of interest rate (or duration) mis-match would be appropriate for assets held against reserves, rather than assets acquired for monetary policy purposes. Once assets move out of the APF any losses from mis-match risk would directly hit the Bank’s capital – which is tiny. The Court of the Bank would need to be content that any significant interest rate risk was justifiable. It is not clear that the aims of the Bank – as set out in various remits and in Parliamentary Acts – would be consistent with its taking substantial interest rate risk. So it might be that the assets held against reserves would become neutral in the sense of paying a return very close to Bank Rate – for example by being invested in short term Treasury Bills. Such a portfolio would also likely be neutral in another sense, in that it would be unlikely to generate much of a monetary policy effect (unlike holding a large quantity of longer term bonds). That would mean that the MPC need not feel constrained in any way by the size of the Bank’s balance sheet and so decisions on that balance sheet would no longer be a matter for the MPC.

Might there be a reason for the assets held against central bank reserves to have a different maturity than those reserves – generating potential gains but also creating risks of losses? One argument is that on average yields on longer dated government bonds might be greater than Bank Rate. Thus if the Bank held a portfolio of longer dated bonds against reserves paying Bank Rate they would, on average, generate a positive return that would be paid as a dividend to HMT. But there is no guarantee that Bank Rate stays below yields on longer dated bonds. And the Bank could not assume that the demand for reserves was so steady that it could simply hold bonds until they are redeemed. If the Bank accommodated the commercial banks’ demand for reserves it might need to sell bonds as demand for reserves declined. This generates risks of capital losses.

Nonetheless the prospect of average gains might mean that the government wanted the Bank to take on a mismatched portfolio and could stand ready to make good the hit to Bank of England capital when and if losses come. This would rather change the aims of the Bank of England. But it is not at all clear this would be best way to take advantage of a possible tendency for yields on longer dated bonds to exceed Bank Rate. There is a much simpler and more direct way for the government to adjust its funding to take advantage of such a situation. This is for the DMO to issue a much greater quantity of short debt (e.g. Treasury Bills) and a correspondingly smaller quantity of longer dated (and on average more expensive) bonds. That greater stock of Treasury Bills would then be held by the Bank against the reserves. This generates the same overall gain to the government as having the Bank of England hold longer dated gilts against reserves. Either way the government is effectively funding more of its debt at Bank Rate.

17 March 2017