Balance sheet operations and central bank policymaking architecture

Author: Charlie Bean

Published: 6 February 2023

The great financial crisis of 2007-8 not only resulted in a recasting and tightening of the regulation of financial intermediaries, but also the introduction of a new arrow into the policymakers’ quiver in the form of macroprudential policies. The purpose of such policies is to moderate building financial stability risks during normal economic times so as reduce their likelihood and impact in the event of crystallization. It therefore constitutes a preventative counterpart to central banks’ traditional role as the lender of last resort during a financial crisis.

Balance sheet by Pascal Skwara, Getty Images

Many of the instruments of macroprudential policy are not new. Lender-focused instruments such as bank capital requirements have long been part of the prudential framework but varying them with a view to managing systemic financial risk, as with the counter-cyclical capital buffer (CCyB), is more novel. By the same token, borrower-focused instruments such as restrictions on the terms of household borrowing have been deployed historically as a tool to control aggregate demand, though in recent years they have been displaced by variations in interest rates. But exploiting their potential to moderate or mitigate systemic financial risks represents a new departure.

Going hand in hand with this new policy armoury, has been the creation of new machinery governing its deployment. Arrangements differ across jurisdictions and often reflect pre-existing institutional structures as much as deliberate design. Where the central bank is responsible for banking regulation and supervision, it is often in the lead. An IMF study (IMF, 2011), for instance, found central banks carried prime responsibility in almost a half of the 47 jurisdictions under investigation. In the remainder, responsibility fell either to another regulatory agency, the finance ministry, or – in about a fifth of cases – a committee or council comprising representatives drawn from several bodies or agencies (the case for both the United States and the European Union). The problem with arrangements involving multiple actors is that it is prone to inhibiting nimble and decisive policy action, especially if the aim is to act by consensus.

Where the central bank has prime responsibility for macroprudential policy, a subsidiary question is whether it should be undertaken by the same body that has responsibility for monetary policy or whether monetary policy and macroprudential policy should be operated by separate committees. On the one hand, a single committee facilitates coordination and reduces the risk that the two policies operate at cross-purposes. On the other hand, having separate committees permits the appointment of members with specialised knowledge that is appropriate to one task or the other, namely knowledge of monetary and macroeconomics for the monetary policy task, and of financial markets and institutions for setting macroprudential policy.

Here, the UK government, after consultation with the Bank, adopted a two-committee model, with the new Financial Policy Committee (FPC) incorporating many (though not all) of the features of the tried-and-tested Monetary Policy Committee (MPC). But having some members of the Bank executive in common and the Governor chairing both committees, not to mention the scope for joint briefing and meetings when required, facilitates co-ordination when needed. That is reinforced in the Chancellor’s remit letter to each committee which asks them “to have regard to each other’s actions, to ensure coordination between monetary and macroprudential policy” and directs them to “reflect, in any statements on its decisions, the minutes of its meetings and its…Reports, how it has had regard to the policy actions of the” other. Wording along these lines has been present ever since the March 2014 editions of the remit letters, i.e. not that long after the creation of the FPC.

This arrangement works fine in ‘normal’ times when Bank Rate is well above its effective lower bound and the FPC’s macroprudential instruments are efficacious in moderating the risks to financial stability. Indeed, in principle, each committee does not need to be instructed to have regard to actions of the other, even when each instrument affects both the monetary policy objective (inflation) and the financial stability objective (e.g. ‘GDP at risk’). That is because convergence to an appropriately co-ordinated equilibrium should occur naturally. This will be the case provided: (a) the policy instruments are assigned to each committees according to their comparative advantage, with the MPC in control of the instrument(s) that have a comparatively stronger effect on aggregate demand, while the FPC has charge of the instrument(s) that have a comparatively stronger effect on financial stability risks; and (b) each committee takes the instrument setting of the other as a given.

Such an assignment of instruments to committees according to their comparative advantage is the case at present: Bank Rate rather than macroprudential instruments is better suited to controlling aggregate demand; and macroprudential instruments rather than Bank Rate are better suited to controlling financial stability risks. (For analytical underpinnings to this analysis, see Bean, 2021.) That said, there is no reason to object to the instruction to each committee that it should “have regard” to the actions of the other.

Matters become more problematic when the instrument(s) of one or both committees cease to be effective at achieving their mandated objectives. The mandate for the MPC has acknowledged this possibility ever since the March 2013 remit letter, which noted that:
“Circumstances may also arise in which attempts to keep inflation at the inflation target could exacerbate the development of imbalances that the Financial Policy Committee may judge to represent a potential risk to financial stability. The Financial Policy Committee’s macro-prudential tools are the first line of defence against such risks, but in these circumstances the Committee may wish to allow inflation to deviate from the target temporarily, consistent with its need to have regard to the policy actions of the Financial Policy Committee.”

This allows the MPC to ‘aim off’ the inflation target so as to mitigate the risks to financial stability should the FPC judge that their macroprudential instruments are not up to the task – also sometimes referred to as ‘leaning against the wind’. An example of this might be if systemically significant financial institutions outside the regulatory perimeter are building up excessively leveraged positions that can be discouraged through the adoption of a tighter monetary policy. However, it leaves the trade-off between the two objectives – achieving the inflation target and moderating financial instability risks – entirely in the hands of the members of the MPC; the FPC only has influence indirectly, when arguably it would be better as a joint decision.

Moreover, the set-up really should be symmetric, and here there is a lacuna in the present architecture. In principle, just as the FPC could judge that its macroprudential instruments are unable to mitigate the risks to financial stability effectively and thus wish the MPC for a while to set a tighter monetary policy than needed to achieve the inflation target, so the MPC could also judge that it is unable to loosen monetary policy enough to hit the inflation target and thus wish the FPC to relax its macroprudential stance in order to foster the expansion of more credit in order to boost aggregate demand.

This is by no means a nice academic point. Bank Rate has been at or near its effective lower bound for much of the past decade. And while the MPC has so far judged that asset purchases financed by reserve creation (i.e. quantitative easing) represent a viable alternative monetary instrument, that need not be the case in future. In particular, the downsides to the use of the tool have attracted increasing attention: increased wealth inequality; heightened risks to financial stability by encouraging a ‘search for yield’; and increased sensitivity of the public finances to changes in Bank Rate. So it is quite possible that the MPC may in future decide that quantitative easing no longer represents a viable monetary policy instrument (or the Chancellor of the Exchequer – who must explicitly consent to the MPC’s asset purchases – could choose to veto them for essentially the same reasons).

Consequently, for completeness if nothing else, it would seem sensible to go beyond the present general instruction to the FPC that it should “have regard to” the actions of the MPC by incorporating into the FPC remit letter a paragraph analogous to that quoted above from the March 2013 MPC remit letter and which allows the FPC to loosen macroprudential policy – for instance by lowering the CCyB – should that be felt necessary to generate a sufficiently high level of aggregate demand.

The interactions between monetary and macroprudential policies can nevertheless probably be managed adequately by separate committees so long as the instruments that each has responsibility for are distinct. But matters become murkier when essentially the same instrument is under the control (or appears to be under the control) of both committees. This is most likely to arise in the context of balance sheet operations, which may be a response to a variety of considerations.

Broadly speaking, there are five purposes to which central bank balance sheet operations might be directed (see Tucker and Cecchetti, 2021). They are: to affect the level of aggregate demand through asset purchases (monetary policy); to provide funds to solvent but temporarily illiquid financial firms (classic lender of last resort); to stabilize critical asset markets (buyer or market maker of last resort); to direct the allocation of credit to or from particular borrowers (in essence, a form of fiscal policy); and to provide emergency financing to the government (monetary finance).

A case of such confusion of powers was provided by the Bank’s programme of gilt purchases (financed by reserve issuance) announced on 28 September 2022 whose aim was to stabilize the very long-dated segment of the gilts market in the wake of the Government’s ‘mini-budget’ the preceding Friday. In that announcement, the Bank explained that the “Financial Policy Committee noted the risks to UK financial stability from dysfunction in the gilt market. It recommended that action be taken, and welcomed the Bank’s plans for temporary and targeted purchases in the gilt market on financial stability grounds at an urgent pace.” So the motivation was very clearly to act as a buyer of last resort for financial stability reasons.

The Bank went on to say that “the Monetary Policy Committee has been informed of these temporary and targeted financial stability operations. This is in line with the Concordat governing the MPC’s engagement with the Bank’s Executive regarding balance sheet operations. As set out in the Governor’s statement on Monday, the MPC will make a full assessment of recent macroeconomic developments at its next scheduled meeting and act accordingly.” So the MPC was informed about, though not asked to opine on, the operation, even though it was apparently indistinguishable from quantitative easing and in principle could have been expected to have a (small) macroeconomic effect.

The issue was further exacerbated by the fact that the MPC at its previous policy meeting only a few days earlier had announced its intention to commence quantitative tightening by selling off some of the gilts held in the Asset Purchase Facility. While a material change in circumstances can always be used to justify a change in policy and there is no doubt that the Bank’s support action was appropriate, to the outside world it looked rather like one part of the Bank acting directly against what another part was intent on doing. So the optics were not great.

Those optics could have been improved had the Bank’s long gilt purchases been sterilised by simultaneously selling off an identical quantity of short gilts so as to leave the quantity of reserves unaffected. Given the relatively small size of the Bank’s long gilt purchases, the macroeconomic consequences of this additional step would have negligible. But it might have helped to separate the necessary short-term financial stability operation from the longer-term strategy of the MPC of gradually unwinding the almost £860 billion of bond holdings acquired under its quantitative easing programme.

Even this is something of a fudge, in that it only really makes sense if the transmission mechanism operates exclusively through the quantity of high-powered money. In practice, few commentators take such a narrow view about the transmission mechanism. Indeed, a reserves-focussed view is incompatible with the combination of steady nominal demand growth over the past decade (pandemic aside) and the explosion of central bank balance sheets over the same period as a result of quantitative easing.

Such tensions between the use of the central bank’s balance sheet for monetary policy and for financial stability purposes seem likely to continue to arise in the future and in other jurisdictions. A prime example is the euro area, where the European Central Bank has essentially operated two quantitative easing programs for primarily monetary policy purposes (that is, to meet its inflation objective): the Asset Purchase Program (APP); and the Pandemic Emergency Purchase Program (PEPP). But it has also introduced two asset purchase facilities that can be used for essentially financial stability reasons: Outright Monetary Transactions (OMT); and the Transmission Protection Instrument (TPI). Both these tools allow for purchases that are focussed on specific jurisdictions (providing certain conditions are met).

The first of these was never deployed, though its mere announcement proved highly effective during the worst of the euro-area debt crisis. But in the current environment of excessively high inflation, one can certainly envision circumstances in which the ECB Governing Council might want not only to continue to raise policy rates, but also to sell bonds bought under the aegis of the APP and PEPP; such sales would presumably be suitably balanced across countries, maturities, etc, reflecting the holdings within the respective programmes.

It is entirely plausible, however, that the resulting increase in market interest rates will be greater for more highly indebted sovereigns, possibly even generating considerable instability in those markets. And that is precisely what the TPI was introduced to prevent. But the consequence would then be APP/PEPP bond sales taking place alongside purchases of bonds of more highly indebted sovereigns under the TPI. The net effect of the two operations is thus to change the geographical composition of the ECB’s assets as well as its overall size. There is nothing wrong with the economics of this, of course, but considerable care would be needed around the associated communications.

What does this mean for the policy-making architecture? By itself, the simultaneous use of the central bank’s balance sheet for multiple purposes argues for a single policy committee in order to ensure that decisions are not taken at cross purposes and that communication with market participants, the media and the wider public is as coherent as possible. It probably also helps to have separate facilities with clearly defined purposes along the lines of the ECB model, so that when asset purchases and/or sales are taking place for more than one reason any apparent conflicts can be more readily explained and justified (such as general sales of sovereign bonds from the APP/PEPP coupled with simultaneous purchases of the bonds of some of those sovereigns under the TPI).

There are, though, the costs associated with having a single committee. In particular, unless that committee is quite large, it makes it harder to accommodate the specialised markets knowledge appropriate to macroprudential and financial stability concerns alongside having a variety of people with the macroeconomic skill set that is suited to monetary policy. In my experience, relatively small committees comprising participants all with a background that enables them to contribute effectively are vastly preferable to large committees where a significant fraction of members lack sufficient expertise to contribute. In this regard, I believe the present Bank of England set-up has generally worked well.

There is, however, one aspect of the UK arrangements that could perhaps be tightened up. As it stands, the management of the Bank’s balance sheet is the responsibility of the Bank Executive, although the arrangements for that are somewhat ad hoc, with the Governor carrying the responsibility of working with whichever members of the Executive he deems most appropriate to the problem at hand, and then answerable to the Bank’s Court of Directors. In practice, this will usually involve consultation with the Deputy Governors (known internally as Govco) together with whichever executive directors (s)he deems appropriate. Formalising this into an established committee with responsibility for managing the balance sheet, as well as laying out its relationship to the MPC, FPC and the PRA board would seem a worthwhile step. This is all within the power of Court to initiate and should not require any legislation or amendments to the Bank of England Act.