Paul Fisher: Microprudential, macroprudential and monetary policy - conflict, compromise or co-ordination?
Speech by Mr Paul Fisher, Executive Director for Supervisory Risk Specialists & Regulatory Operations and for Insurance Supervision of the Bank of England, at Richmond University, London, 1 October 2014.
The author is also chair of the ifs School of Finance. Any views aired in this speech are personal and should not be taken as a policy statement by the Bank of England. Nevertheless, the speaker is grateful to colleagues at the Bank for helpful comments, and to Andrew Hughes Hallett for his advice on policy coordination issues over many years, including comments on the content of this speech.
Thank you for inviting me once again to speak at Richmond University. In the last year I have moved within the Bank of England from the Markets Directorate to the Prudential Regulation Authority (PRA) and from the Monetary Policy Committee (MPC) to the PRA Board. The PRA Board is one of our three formal policy committees, and makes the highest level supervisory decisions for those firms authorised by the PRA. That includes insurance companies, banks, building societies, credit unions and major investment firms. I have been attending meetings since June and will become a voting member in due course.
Having previously also served on the interim Financial Policy Committee (FPC) for 2 years, and the MPC for 5 years, I want to use this lecture today to set out how the three policy committees of the Bank work together.
I have seen and heard assertions in the press and other commentary that the three committees with their different objectives will inevitably be in conflict. Whilst there are many difficult policy challenges for each of the committees, and different judgements can be reached by reasonable people on any policy question, I will argue today that conflict is not inherent in the structure in the sense of having to choose between competing objectives. There should not be any compromise in terms of achieving those objectives over the medium term, even if the actual setting of each policy lever needs to take others into account. The key to realising this outcome is effective co-ordination.
Much of the economics literature around policy co-ordination traditionally focussed on two problems: the coordination of monetary and fiscal policy, especially in the case of an independent central bank; and international macroeconomic policy co-ordination1. In the past decade or so there has been rather more on inter-institutional coordination2. I will be drawing on that body of established work and applying it to the challenge of co-ordination faced by the Bank's three policy committees: the MPC, the FPC and the PRA Board. First, what are their objectives?
- The MPC's objective is to deliver price stability - defined by the Government's inflation target of 2% - and, subject to that, to support the Government's economic objectives including those for growth and employment.
- The FPC's objective is to protect and enhance the stability of the financial system of the UK - and, subject to that, to support the Government's economic objectives including those for growth and employment.
- The PRA has two statutory objectives: to promote the safety and soundness of deposit takers, insurers and major investment firms and, specifically for insurers, to contribute to the securing of an appropriate degree of protection for policyholders. It also has a secondary objective, to facilitate effective competition in the markets for services provided by PRA-authorised firms.
The PRA Board
The PRA Board is the least well known of the 3 committees and so I want to say a few words about how it operates.
I have been attending as an observer since I moved to the PRA on 1 June. The Board is constituted in a similar way to the other committees in that it is chaired by the Governor and has both internal executive members and externally appointed independent members. On the basis of current announcements it will have a total of 5 internal members3, and 6 independent members4 including the CEO of the Financial Conduct Authority (FCA). It meets approximately twice a month for formal meetings.
In summary, the Board is responsible for meeting the PRA's statutory objectives, including taking all the major supervisory decisions for the 1700 or so firms and groups it authorizes; formulates regulatory policy to meet those objectives; and has a range of strategic and management responsibilities including delegating some matters, principally to the CEO of the PRA or to formal PRA committees. Matters reserved to the Board and its delegations can be found in documentation on the Bank's website via a link in this text.5 In what follows I shall treat the PRA Board and the PRA as one policy making body, referring to one or the other as best fits the context.
Before moving on, it is worth completing the picture by noting that the PRA works alongside the FCA, creating what is known as a "twin peaks" regulatory structure in the UK. The FCA is a separate institution and not part of the Bank of England. It is responsible for promoting effective competition, ensuring that relevant markets function well, and for the conduct regulation of all financial services firms including all those authorised and regulated by the PRA. This includes acting to prevent market abuse and ensuring that consumers get a fair deal from financial firms. The FCA also operates the prudential regulation of financial services firms not supervised by the PRA, such as asset managers and independent financial advisers. In consequence it regulates a much larger number of firms than the PRA - over 50,000. But the FCA and PRA have a joint MoU and co-operate closely on a range of matters, most especially for those firms which are regulated by both ie by the PRA for prudential matters and by the FCA for conduct.
Policy co-ordination basics
The theoretical conditions under which one can achieve a number of different economic policy objectives at the same time was long ago embedded in the Tinbergen counting rule6, named after the great Dutch Nobel prize winning economist. A necessary condition is that there should be at least as many independent and effective policy instruments as there are independent policy targets. By an instrument I mean any policy setting which can be changed: such as Bank Rate or the required amount of capital to be held by firms. To be satisfied, the Tinbergen rule requires that the policy tools being used must each have an impact on at least one of the objectives in sufficient size to make a substantial contribution towards achieving it. And that the pattern of effects of one tool on all objectives cannot be perfectly reproduced by changing any other tools singly or in combination ie the policy tools are each unique in their precise pattern of effects.
Additional conditions are necessary if the objectives are assigned to different policy making bodies. So I will add a number of other sufficient conditions to cover this case.
- Each policy tool should be assigned to meeting a target that it is relatively effective upon (ie to exploit comparative advantage). This may seem obvious, but is less so if one is uncertain about a particular policy's effectiveness. If a policy is not very effective on its primary objective, one of the consequences could be unnecessarily large policy changes.
- The authorities need to be able to co-ordinate or communicate in some way about what they are doing and why. Otherwise one can show that there is a risk in some circumstances of a dynamic being created which is destabilising (eg where one authority tries to offset the effects of another in turn).
- And perhaps most importantly, the targets need to be achievable & coherent as a set.
Let me explore this last point briefly by taking an extreme case. If the MPC were to be set a target of absolute zero unemployment, that would be physically impossible and however hard they tried, it would not be achieved. If they did try hard nevertheless, it would probably lead to an over-expansion of the economy that threatened both financial stability and prudential safety and soundness. In a less extreme case one could imagine an MPC target for money supply growth which, given time-varying shocks to the velocity of money, might not be consistent with the other objectives.
Applying these conditions to the UK authorities
So the first question I want to ask is whether the MPC, FPC and PRA Board have been set an achievable and coherent set of targets? Is it possible to achieve CPI inflation at 2%, financial stability, and safety and soundness of firms simultaneously7? I assert that there is no inherent absolute conflict - these are different elements that can be associated with a single economic equilibrium.8 In the absence of events, random shocks or other disturbances, it would be possible to achieve that triple objective simultaneously. If that were not possible, there would be a truly deep conflict, and the only solution would be to go back to government and ask for the remits to be changed so that coherency was achieved. I see no arguments to support that.
But even though the different policy objectives can be achieved simultaneously, that does not mean that each committee can ignore the others. That would only be true in a special case when use of each committee's policy tools affected its own objectives and not those of others ie if the policy effects were perfectly independent. In practice independence is inevitably partial: each policy tool is likely to affect all the objectives, albeit in different degrees. In this general case, the choices of one committee will affect the choices of the others. An example of this was when the MPC, in giving it's forward guidance on monetary policy, included a financial stability "knockout" - acknowledging explicitly that there might be circumstances when monetary policy settings had to reflect the risks to financial stability. That interaction or "spillover" is not the same as having a conflict between the objectives. Policy spillovers will be reflected in adjustments to the precise settings of policy levers at particular moments in time, but should not compromise achievement of the ultimate objectives. But to keep on course for multiple objectives certainly requires co-operation to reflect these policy interactions.
One aspect of this co-ordination is a common understanding of the economy. For example, the gap between actual and potential output levels could have an important bearing on policy settings for each committee. If the three committees had different views about the size of that gap or just different economic models, then they could set inconsistent policies. That would be a conflict of judgements - not a conflict of objectives. The presence in the Bank of a single forecasting function and cross-use of its outputs, helps to avoid that particular trap. Any single central forecast would still most likely be wrong of course, but the committees will learn about changes to the outlook consistently, allowing consistent policy adjustments.
Another angle to cooperation arises from the "long and variable lags" in the economy - and not just in the monetary transmission process. Each committee can face very long and uncertain lags between deciding to take a policy action and the maximal impact on its objective. That makes it unlikely that any one objective can be achieved precisely at all points in time.
As well as predictable lags, policy outcomes are subject to all sorts of random disturbances that constantly require changes to policy settings. Suppose, however, a committee actually tried to obtain its objectives at every moment in time regardless. That would likely necessitate wild swings in policy settings: imagine the MPC trying to set Bank Rate so that 2% inflation was achieved exactly every month! Not only would it quickly prove impossible, attempting the feat would drastically reduce the chances of the other committees achieving their objectives. So a degree of humility is required by each committee, for the sake of coordination as well as the credibility of each.
In their commitments to hit their statutory primary objectives I would say that:
- The MPC should not be "inflation nutters".
- The FPC should not seek the "stability of the graveyard".
- The PRA Board should not seek to implement a "zero failure regime".
That is supported helpfully by the secondary objectives of each policy maker. One can interpret them in several ways, but to me they help put the primary objectives in an overall context. We want all three objectives to be met in an economy which is growing at its maximum sustainable potential. Excessive zeal in meeting any primary objective too quickly could act, not just to slow actual growth, but reduce sustainable potential and this is important to keep in mind.
Cooperation and coordination are generated in the framework by several structures. Essentially, these are about information sharing. The most important is via cross-committee membership: the Governor chairs all three committees, and there is further cross-membership of internal members. And there is also joint briefing meetings/staff support. The Strategic Plan of the Bank announced earlier this year, is designed in part to enable the different policy functions of the Bank to work more effectively together, reducing overlap and exploiting synergies.
A notable example of policy co-ordination across the Bank's committees has been on the question of housing. Rising house prices could generate rising debt levels over time, creating risks to monetary stability as well as financial stability and the safety and soundness of firms. But addressing this by trying to adjust Bank Rate would potentially adversely affect the whole of domestic demand at a time of economic recovery and low price inflation, and it could have needed a very large change in Bank Rate as it is not very effective in controlling house prices.
At the end of June, the Bank announced a set of macroprudential actions designed to mitigate the risks from the housing market (not to control house prices per se). The Bank was able to construct policies which would operate directly on the flow of debt arising as a result of the housing market. It was clear that the FPC should take the lead in using these tools, but drawing on a common macroeconomic assessment (provided by the MPC's forecast, with policy analysis coming jointly from across the Bank) and a micro assessment provided by the PRA. The FPC could then set the overall macroprudential parameters by a Recommendation, with the particular policies being implemented by supervisory processes consistent with the PRA's objectives, and by the rules of the FCA. Indeed it is worth noting at this point that the coordination circle does include the FCA - the CEO of which sits on both the FPC and the PRA Board. One could extend the three-committee analysis in this lecture to four.
That brings us to the question of the policy levers themselves. Is it the case that each committee has instruments which are both sufficiently effective and independent?
The MPC clearly has strong influence over the price and quantity of money. Bank Rate is a basic price of money: the rate paid to commercial banks on their reserve holdings at the Bank of England. The quantity of narrow money (those reserve holdings plus banknotes in issue) is also set by control of the Bank's balance sheet: either by adjusting the stock of purchased assets or by adjusting the stock of secured lending to the banking system. In respect of this latter policy lever, the Bank's executive and its board of directors (Court) are ultimately responsible for management of the Bank's balance sheet, but there are formal agreements in place such that if operations need to be conducted for monetary or financial stability motivations, then the appropriate committees are consulted and will ultimately will make the decision9.
Distinguishing the FPC and PRA's instruments takes us into more interesting territory. The FPC can make recommendations to pretty much anyone about anything that would help meet its objectives. That's very wide ranging but effectiveness depends on those recommendations being accepted and implemented. That can only be guaranteed where there is effective coordination between the FPC and those to whom the recommendations are made. PRA, FCA, Bank and HMT all fall into that category. The FPC's most enforceable powers (of direction) relate to setting the counter-cyclical capital buffer and sectoral capital requirements which they can direct the PRA to implement across all deposit takers. A leverage ratio is likely to be added in due course and the Government have also stated that they intend to grant other powers to the FPC to ensure that it has sufficient policies to address issues arising from the housing market.
Capital and liquidity requirements are the most basic tools for controlling the banking system in aggregate. Superficially, and taking directions and recommendations together, the FPC could therefore be interpreted as undertaking coordinated supervision. If that were the case, the question must arise whether the PRA has any independent instruments and whether the FPC and PRA Board arrangements mean that both can achieve their objectives. Specifically, does the FPC dominate the PRA? The answer of "no" comes in two parts.
First, beyond the formal powers of direction, the FPC can make recommendations on a wide ranging basis. Only the PRA and FCA may be obliged to respond10, but in principle, the FPC could seek to involve itself in much wider matters than those over which the PRA has legal powers. Second, the FPC cannot involve itself in matters relating to specific firms. For example, ultimately only the PRA Board can sign off on what levels of capital an individual deposit taker must hold, adding their independent judgements to the minima which are set by a combination of international standards and any FPC decisions. Those judgements reflect the specific risks being faced by individual firms and there are a range of factors, any of which may cause a capital requirement to be made. The PRA's judgemental approach to supervision is set out in its Approach Documents for Insurance and Deposit Takers which are published on our website.
In exercising its authority, following its published approaches, the PRA also makes judgements around a broad set of firm-specific factors that the FPC would not be involved in: risks not captured in the formal capital definitions, senior appointments, approving specific transactions, risk and governance controls etc. In short the FPC makes decisions to ensure financial stability in the system as a whole and the PRA acts to support the safety and soundness of individual firms.
But the PRA approach to supervision also involves other considerations. I have already mentioned that we must not operate a "zero failure" regime. That was in the context of not taking extreme decisions in over-zealous pursuit of our objectives. But if investors and other market participants believe that supervisory effectiveness leads to a situation in which a firm cannot fail, then they will likely take inappropriate exposures to those firms and the market will become distorted. In particular if a firm is likely to be bailed out by the public sector then that reduction in risk means that firms will have to pay less for their liabilities - giving them an implicit subsidy. That has been discussed at length in a number of Bank publications, but a key point for today is the Bank's continuing commitment to end the "Too Big to Fail" problem and, specifically, to seek to establish realistic recovery and resolution plans for the firms it authorises. That consideration also affects supervisory decisions on legal entity structures, transactions, senior persons and so on.
In short, the FPC and PRA have many different, independent policy decisions to make. Ensuring both financial stability and safety and soundness of firms, requires both policy sets.
The issue is sometimes raised as to how systemic stability and safety and soundness interact - especially during or after a financial downturn, when defensive actions by financial firms may be pro-cyclical. This is perhaps the period when cooperation and communication between the committees comes to the fore most strikingly, and forward guidance on policy can be most effective and necessary. The FPC and MPC will want to make sure that the regulatory system is not pro-cyclical - they will want to avoid anything that reduces the capacity for credit creation when the wider economy needs most support from its financial sector. One might worry that supervisors would see the general relaxing of requirements - the use of capital and liquidity buffers - as weakening the safety and soundness of their firms. In fact, the PRA also needs to recognise the relevance of the economy-wide situation to its objectives. Weakness in the real economy in a general downturn is one of the biggest threats to the safety and soundness of firms, as it can lead to substantial impairment of asset values. More than this, the consideration of capital and liquidity is more subtle than many people seem to appreciate.
It is often asserted that requiring firms to hold more capital automatically reduces the capacity to lend. That is not the case. Capital is a source of funding and sits on the liability side of the balance sheet. What matters is how a firm goes about improving its capital ratios. Reducing lending is only one way of adjusting a capital ratio up. As well as raising new capital, retaining more cash, rather than paying dividends or staff bonuses, can also contribute. And it turns out that, at the start of the financial crisis, the larger banks had many non-core assets11 on their balance sheets, beyond lending to households or non-financial companies. In the past few years, capital positions have been rebuilt in many firms by running down or selling on these other, non-core assets, rather than seeking to reduce direct lending to households and businesses.
The regulatory action most likely to affect the capacity to lend is that of increasing liquidity requirements - liquid assets obviously do compete with other assets on the balance sheet including less liquid loans. But in a proximate sense, most firms fail not at the point when they have run out of capital, but when they run out of cash (although the latter may be a forward indicator of the former). Firms holding enough liquidity in reserve is a crucial defence in maintaining a robust financial system and hence financial and monetary stability.
The key to solving this policy dilemma is to make sure that firms hold enough liquidity (and capital) in the good times so that, when a downturn hits, supervisors can take a relaxed view about firms using those holdings as buffers to absorb the macro shock. That was not the position post 2008/09 when banks had not been holding enough liquidity or capital in the good times. It is an aim of the new regulatory frameworks put in place post-crisis, that banks should hold bigger liquidity buffers in normal times.
The Bank of England's balance sheet can also be used actively to backstop liquidity for the system as a whole, via the banking system, and the new operating frameworks put in place over the past 5 years will allow that to happen more quickly and easily.
How will this all co-ordinate in practice? When the Bank, HMT and FSA took joint action in 2012 to reignite the recovery via the Funding for Lending Scheme - reviewed in my two previous talks to you - the FSA, as predecessor to the PRA and having two members on the interim FPC, contributed by relaxing liquidity requirements and making it clear that liquidity buffers were there to be used. This was a crucial element of the package.
Coherence vs agreement
I believe I have set out so far a reasoned view that the three Bank policy Committees - the MPC, FPC and PRA have a coherent set of objectives, structures, appropriate policy tools and coordination to ensure that there is not any inherent absolute conflict in the framework. And the strategic changes internally at the Bank announced earlier this year could be seen as being a reinforcement to that structure by maximising the synergies across the Bank, including the PRA.
But coherence doesn't always mean agreement. There will no doubt be disagreements at some point. Indeed, the appointment of independent members to the policy committees is intended to ensure challenge and the absence of "group-think". Dissenting views on the MPC are not only an established part of the culture but actively encouraged by the Treasury Committee in Parliament.
The FPC and PRA Board, in contrast, generally work by consensus, although there are provisions for recording dissent and votes if necessary. But disagreements over judgements between individuals are normal - that's why we have decision making by committee in the first place. What we have not seen is any reason why the structure should be regarded as internally conflicted. How likely is it that there will be dissent across committees? There are mitigants in place to reduce the chances of that. Cross-chairmanship and membership from Bank executives should enable us to keep to coherent policies even when there are differences of opinion. And I say that - despite evidence from operation of the MPC where there has generally been as much dissent recorded between the Bank executive members as by the independent members - because we have a collective institutional need to make the whole framework function in the best interests of the UK economy. And another structural way of addressing divergent views is to have joint meetings and briefings on the issues which have most cross-over, which help bring the different externally appointed members together. That has already happened successfully on several issues, most notably housing.
It is important that we give the new structure time to work. First, because we will learn a lot. If there are flaws - and it would be reasonable to think there will be some - we need a bit of time to identify them and iron them out. Macroprudential tools are being used in the UK for the first time. And the PRA's new approach to supervision is still being embedded. Second, the regulatory reform agenda is still being decided and implemented internationally - insurance regulation in particular has a long way to go to set truly global standards. Measures to tackle Too-Big-to-Fail are still being finalised. So we are still some way from having a stable international regime within which to operate. Third, policy changes take time to have their effect. Indeed, the experience of monetary policy over the past few years demonstrates the need for patience.
As well as addressing the issues around policy coordination, I hope I have also drawn your attention today to some aspects of how the PRA and its Board work. These are not secrets, and there is a wealth of information already published. But prudential supervision tends not to attract the same public attention as monetary policy unless something goes wrong. The work of the PRA and its Board is nevertheless a vital and distinct part of the new system and we need to be as open and transparent about our role, policies and operating procedures as we can be.
1 See for example, R Bryant (1995) International Coordination of National Stabilization Policies. Brookings, Washington, and the many references therein.
2 See for example some of the material in N Acocella, G Di Bartolomeo and A J Hughes Hallett (2013), The Theory of Economic Policy in a Strategic Context, Cambridge University Press, Cambridge.
3 In addition to Mark Carney as Governor, the proposed internal members from May 2015 will be Andrew Bailey as Deputy Governor and CEO of the PRA; Nemat Shafik as Deputy Governor for Banking and Markets; Jon Cunliffe as Deputy Governor for Financial Stability; Paul Fisher as Deputy Head and Executive Director at the PRA.
4 As I speak the independent members are Sandy Boss, Iain Cornish, Charles Randall and Martin Wheatley as CEO of the FCA. Mark Yallop is due to join on 1/12/14 and David Belsham on 1/5/15.
6 See for example, AJ Hughes Hallett (1989) 'Econometrics and the Theory of Economic Policy: The Tinbergen-Theil Contributions 40 Years On.' in Oxford Economic Papers, vol. 41, issue 1, pages 189-214.
7 With apologies to my insurance colleagues I am going to leave on one side the objective of policy holder protection. That should not affect this discussion.
8 There is an interesting longer term issue of whether stability - on any one objective, let alone all three - necessarily breeds the conditions for subsequent instability, given that humans take more risks when conditions are thought to be stable and safe. I am not going to address that issue in this lecture.
9 These arrangements with MPC and FPC have been published.
10 Recommendations to PRA or FCA could be made on a "comply or explain" basis. That is not the normal approach, which is more consensual.
11 Derivatives, prime brokerage, secondary market trading activities are all examples of what I mean here.