Monetary policy in a world with macroprudential policy

Speech by Mr Jaime Caruana, General Manager of the BIS, at the SAARCFINANCE Governors' Symposium 2011, Kerala, 11 June 2011.


Ladies and gentlemen, it is a privilege to be able to speak today to this distinguished group of Governors. As for my subject, I will focus on how monetary policy might need to adjust to the implementation of new macroprudential frameworks. To speak on this topic in India is almost presumptuous, given that the Reserve Bank of India is among those central banks, mainly Asian ones, that have successfully combined both monetary policy and macroprudential measures. So please take my remarks more as an invitation to a productive debate - one in which your contribution will be immensely valuable.

Much has already been done to ensure that financial supervision and regulation incorporate a systemic view of risks and to establish effective macroprudential frameworks. Basel III will increase the resilience of the banking system and will introduce a macroprudential overlay. Oversight bodies and macroprudential authorities will actively monitor systemic risk and act to constrain excessive leverage and maturity transformation. These are just two important elements that will have healthy long-lasting effects on the financial system and economy. As we know, however, more needs to be done to strengthen the financial system.

To be sure, a more stable, more resilient and less procyclical banking system will also improve the effectiveness of monetary policy transmission. But to understand the full impact on monetary policy, we need to understand how the new macroprudential frameworks will change the behaviour of the financial system and the real economy. These changes mean that monetary policy will have to adapt. How it adapts will depend on the way in which macroprudential and macro policies interact. This brings to the fore not only technical issues but also policy and governance considerations.

Perhaps it is too early for definite conclusions. Key reforms are still under way. For example, the Basel III liquidity ratios will improve liquidity management in banks, but may also affect capital markets and monetary policy transmission mechanisms. These effects will need to be analysed closely during the scheduled observation period.

Today, however, I would like to offer some tentative thoughts on some propositions and principles. My main focus will be the interplay between macroprudential frameworks and monetary policy.

1. Define macroprudential policy and its aims narrowly

In what follows, I will define macroprudential policy as the use of primarily prudential tools to limit system-wide financial risk, and so prevent disruption to key financial services and the economy.1

Thus, macroprudential policy is defined by its aim (limiting system-wide financial risk), the scope of analysis (the financial system as a whole and its interactions with the real economy), a set of powers and instruments and their governance (prudential tools and those specifically assigned to macroprudential authorities).

This definition highlights a couple of points.

First, the set of macroprudential tools is not as large as sometimes believed. It may be tempting to consider as macroprudential any tool that can influence systemic risk and financial stability. But such a definition is too broad, as almost anything can have an impact on systemic risk. Monetary, fiscal and competition policies are cases in point. Moreover, a too-broad definition could argue for the adoption of measures whose primary aim has nothing to do with addressing systemic risk and whose effectiveness may be doubtful in that context. Capital controls could be one example.

Second, macroprudential policy should not be considered a tool for the management of aggregate demand. To safeguard macroeconomic stability, there is no substitute for sound monetary and fiscal policies. Unless these policies anchor domestic inflation and ensure the country's long-term solvency, the result will be serious macroeconomic instability. To be sure, the economy will be more stable if systemic risk is contained. But to rely on macroprudential policy as a first line of defence against inflation or other macroeconomic imbalances is bound to lead overall policy astray. Macroprudential policy can, at best, play a supportive role.

This leaves open the question of how narrowly the macroprudential objective should be defined. To answer that question, we first need to ask how a macroprudential framework should deal with financial cycles and procyclicality - a key source of financial instability. Recent experience has confirmed that financial crises often result from mutually reinforcing feedback between the financial system and the real economy. Financial forces can drive and feed economic expansions. Unsustainable developments often show up in unusually rapid credit and asset price growth, together with burgeoning risk appetite. As external funding constraints are eased, they promote additional risk-taking and economic exuberance. During the boom, the financial system may miss the chance to build up sufficient capital and liquidity buffers while this could easily and cheaply be done. As a result, it cannot withstand the subsequent bust. When the unsustainable can no longer be sustained, a financial crisis breaks out. This can be very costly, both economically and socially.

From this perspective, one could single out two possible objectives for a macroprudential framework. A narrow aim would be to increase the resilience of the financial system. A broader, more ambitious one would seek to constrain the upswing of the financial cycle itself. To achieve the narrow aim, all we need to do is to build up buffers during the boom so that they can be used as risks materialise during the bust. For the broader objective, the build-up of the buffers should itself act as an effective speed-limit, restraining the credit and asset price boom. The narrow objective would accept that financial cycles and imbalances could be material despite the best efforts of policymakers. At the same time, it would recognise that, by cushioning the bust, the macroprudential framework would limit the downside of the financial cycle. But it would remain more agnostic about its restraining impact during the boom.

My view is that we should be modest in our expectations. The evidence strongly indicates that macroprudential tools strengthen the banking system's resilience against the bust. At the same time, it suggests that their effectiveness in restraining the boom is more mixed and varies across instruments and financial structures. For example, some countries report that loan-to-value ratios and special provisioning requirements have helped to contain asset price inflation and credit growth in the real estate sector. Several years ago, the Reserve Bank of India raised the Basel weights for household loans, as well as mandating higher loan loss reserves, in the face of rapid household credit growth. Household loans subsequently slowed, even as business loans accelerated; this suggests prima facie that the measures were effective. But it is less clear how far this experience might apply to other financial systems with different capital markets and banking structures. In addition, capital buffers may need to be raised substantially before they can restrain credit expansion: by its nature, capital is ample and cheap in good times.

I draw two conclusions from this. First, a macroprudential strategy is likely to have to rely on a range of reinforcing policies and instruments in order to effectively and sustainably constrain credit growth and asset price booms. Second, given the uncertainties involved at this stage, it would be wise to avoid overly ambitious objectives. We should design the frameworks to provide effective speed limits. But we should not go so far as to judge on this basis whether a macroprudential framework has succeeded or failed. We should be modest in our expectations about this wider objective.

2. Macroprudential policy is not enough to ensure financial stability: other policies have to play their part

Can we rely exclusively on a macroprudential framework to ensure the desired degree of financial stability? I would argue that the answer is "no". At a minimum, both fiscal and monetary policies need to play a more active role than they have in the past.

Let me just say a few words about fiscal policy, as it is not the main subject of my presentation. The plight of the euro area is a telling example of how sovereign solvency is the prerequisite for financial stability. Emerging market countries know this all too well. History also indicates that, during credit and asset price booms, fiscal positions look deceptively rosy. The conclusion is simple. For fiscal policy, we need to apply the same principles that apply to a macroprudential framework: namely, build up buffers in good times so that they can be drawn down in bad times. This means running prudent budget surpluses in good times. And it means not being fooled by the one-off revenues that private sector financial imbalances generate as they build up.

Views about the appropriate role of monetary policy have evolved in the light of the financial crisis. Pre-crisis, the relationship between financial stability and monetary stability was typically regarded as quite simple, at least in most advanced economies. At the cost of oversimplifying, two propositions summarised this view. First, price stability is sufficient for macroeconomic stability. To put it less provocatively, price stability, together with developed and efficient financial markets, would either prevent financial crises or, if they did happen, keep them to manageable proportions. Second, monetary stability should be achieved by mandating an independent central bank with a narrowly specified inflation target. Not surprisingly, standard macro models treated this view as axiomatic and often failed even to mention banks.

The crisis showed that this paradigm is too narrow.

First, it reminded us that financial imbalances can build up even without inflation. Inflation was subdued in the mid-2000s. Yet, at the same time, unsustainable asset price booms developed in many countries, setting the stage for disaster. Evidently, aiming to maintain price stability over a typical two-year policy horizon is not a sufficient safeguard against financial and macroeconomic instability.

Second, the crisis hammered home the message that the correction of financial imbalances can put a huge strain on monetary policy. During the crisis, deflationary pressures and plummeting output induced many policymakers to lower rates until they effectively hit the zero lower bound. Central banks also engaged in aggressive balance sheet policies. As a by-product, these policies increased central banks' financial risks and put their budgetary independence into question. Thus, the crisis showed that a strategy that limits itself to post-bust cleaning up carries huge costs and can cripple monetary policy effectiveness.

More generally, the maintenance of financial stability is too big a burden to rest exclusively on prudential policies, macroprudential included. First, as already noted, it is difficult to constrain the build-up of financial imbalances even with a combination of policies. And the results are uncertain. But the correction of financial imbalances can have serious macroeconomic costs even if it does not result in a full-blown financial crisis. For example, after the end of Germany's re-unification boom, there were no outright bank failures. Even so, the financial system experienced severe strains, which sapped the economy's strength. Second, the effectiveness of monetary policy in constraining credit and asset price booms is hardly in doubt. It is hard to imagine how monetary policy could influence economic activity without affecting credit conditions and asset prices: these are key elements of the transmission mechanism. Finally, monetary policy can help to address regulatory arbitrage, as it sets the universal price of leverage in a given currency.

Will macroprudential policy tend to lead to an amplification or a dampening of policy interest rate cycles? The answer is not straightforward. On the one hand, the troughs might become less extreme, as macroprudential policy should reduce the likelihood of financial crises and their disinflationary consequences. Likewise, interest rate peaks might also come down, to the extent that macroprudential policy succeeds in restraining credit and asset price booms. On the other hand, the need for monetary policy to contribute to financial stability by leaning against the build-up of financial imbalances points to a greater range of interest rate increases during expansions that are marked by such imbalances.

All told, interest rates could move more symmetrically over the financial cycle. They would rise by more during upswings and fall by less during downswings. By implication, there would also be a reduced risk of hitting the zero lower bound and of having to resort to balance sheet policies.

3. Conflicts between macroprudential and monetary policy are likely to be rare

A concern sometimes raised is that macroprudential frameworks could lead to conflicts between monetary and macroprudential actions. My sense is that such concerns are overdone. It seems likely that, in most circumstances, macroprudential policy and monetary policy will be complementary, tending to move in the same rather than opposite directions.

There are two reasons why these policies should complement each other:

First, the financial cycles that matter for prudential policy have a much lower frequency than business cycles. Most business cycles do not involve financial imbalances or crises. In other words, financial crises happen much less frequently than recessions. Since the worldwide liberalisation of financial markets in the 1980s, financial crises have occurred only about once every 20-25 years in any given country. The literature also indicates that financial cycles associated with serious financial distress tend to be considerably longer than typical business cycles.

This suggests that, most of the time, monetary policymakers can treat macroprudential policy developments as a relatively slow-moving background. It also means, of course, that the pursuit of price stability over horizons of just two years or so is no longer fully appropriate. Rather, monetary policymakers will also need to keep an eye on longer-term trends, if they are to take into account the gradual build-up and unwinding of financial imbalances and their economic and inflationary effects.

This longer horizon dissipates some of the possible tensions between monetary policy and macroprudential decisions. Imagine a situation in which a leveraged asset price boom occurs when inflationary pressures are falling. The apparent tension between a desire to cut interest rates and to tighten macroprudential standards disappears once a longer-run perspective on price stability is taken. Since financial crises can generate huge disinflationary pressure, a tightening of monetary policy will promote longer-run price stability.

As an aside, this point suggests that, if monetary policy is mobilised at times to prevent financial instability, no change in formal objectives or mandates will necessarily be required. More important is the analytical lens through which policymakers see the workings of the economy. Indeed, there may be circumstances in which the adoption of an explicit financial stability mandate could be counterproductive. This would be the case, for instance, if it resulted in stronger political economy pressures to keep interest rates low in order to avoid financial stress at times of rising inflation.

Second, we may need to think in terms of a policy hierarchy. A good example is the potential set of responses to strong capital inflows. Capital inflows into emerging market economies can put strong upward pressure on domestic inflation, as well as on credit and asset price growth. In this situation, the top priority is to apply macroeconomic policies - including monetary, fiscal and exchange rate measures - to safeguard domestic financial stability. The appropriate role of macroprudential policy is to curb excessive risk-taking by the domestic financial system. Such restraint might well help to cool aggregate demand and, as such, should to be taken into account by monetary policy. But the use of macroprudential policy should not be used as an excuse to postpone or reduce the inevitable tightening of monetary policy. As for capital controls, these are measures of last resort and are better viewed as a safety valve for extraordinary circumstances. The longer such controls are left in place, the greater the chance of adverse economic side-effects. In this light, India's higher limits on non-resident investment in rupee bonds represent a welcome development. Such investment can help to deepen capital markets.

4. Complementary policy areas still call for policy coordination

Although conflicts between macroprudential and monetary policy are likely to be rare in practice, there will still be a need for mutual consistency and coordination. The close relationship between macroprudential and monetary policy makes that inevitable. More generally, financial stability is a shared responsibility that requires clear cooperation arrangements.

Consistency and coordination could be achieved in a number of ways. At one end of the spectrum, a single institution could take responsibility for all coordination. It would, in fact, determine both policies at all times, with the aim of promoting both macroeconomic and financial stability. Concretely, a single committee or institution could be charged with deciding on the mix of instruments. A central bank would be an obvious candidate for this role. Short of this solution, various other possibilities can be envisaged. For example, policymakers in one area could have veto rights for the other policy; or macroprudential and monetary policy committees could have overlapping memberships. Alternatively, there could be requirements to consult; requirements to notify the other authority before taking decisions; requirements to provide information and advice to the other party; and "best efforts" coordination governed by memoranda of understanding or similar instruments.

The key trade-offs are well known. On the one hand, they involve maximising the credibility and accountability benefits of a narrow policy focus. And, on the other hand, it is a matter of exploiting the technical efficiency benefits yielded by coordination.

This could be done in various ways. For example, multiple objectives could be explicitly ranked. The timing of macroprudential policy reviews could differ from those of monetary policy, which themselves differ from the calendar of fiscal policy actions. Interestingly, some recently established financial stability committees plan to meet quite frequently - perhaps on a quarterly basis. On the one hand, regular review meetings help to keep financial stability in the public eye and could guard against biases towards inaction. On the other hand, this frequency seems quite high given that financial cycles build up so slowly; it might even risk creating the impression that macroprudential interventions will in practice be quite frequent.

5. We need proper governance arrangements: independence, clarity and accountability.

Regardless of the specific type of cooperation mechanisms put in place, financial stability requires governance arrangements that incorporate the principles of independence, clarity and accountability.

Independence from political cycles is needed for macroprudential policy no less than for monetary policy. A common problem for both policies is the need to intervene during the upswing, when things are going well and the public might be sceptical that problems loom down the road. Operational independence will be needed to shield unpopular policy decisions. Strong accountability and clarity of communication will bolster public support for the independence of macroprudential policy and hence its credibility and effectiveness.

Clarity about mandates, responsibilities and powers is important for the effectiveness and timeliness of actions and for managing the difficult trade-offs. Sufficient powers imply control over relevant instruments and appropriate safeguards. For example, access to micro supervisory data is important. At the same time, our limited technical knowledge means that macroprudential frameworks need room to adapt and grow with experience. Very specific and inflexible mandates raise the risk that the specified targets are, or quickly become, poorly matched to the economy's and financial system's needs. As a result, the policymaker's ability to respond to unexpected circumstances could be severely constrained.

Accountability is critical. That said, since financial stability objectives are difficult to quantify or define precisely, accountability is harder to achieve than, say, for price stability objectives in monetary policy. A clear and transparently communicated strategy that sets out the central bank's intentions can serve as the basis for accountability.

Regardless of the specific governance and cooperation arrangements, the emerging reality is that central banks have a key role to play. This role requires mandates and governance structures that are consistent with their primary monetary policy function. In some cases, central banks' duties and powers to promote financial stability are being enhanced. More active financial stability roles will raise issues of reputational risk that central banks will need to manage carefully, especially if their views on specific decisions are not shared by other agencies involved in the process.

Central banks will also face additional challenges. They will face an added burden to be very clear about what policy actions are being taken and for what reason. They will need to be careful not to undermine price stability mandates and hard-won credibility. And they will need to preserve their operational autonomy, including financial independence. In turn, this requires control over their balance sheet and ex-ante clear mechanisms to transfer losses to the Treasury. A forthcoming Central Bank Governance Forum report describes the current range of practice across central banks and analyses the issues posed by various choices.


To conclude, these are early days in our experience with new macroprudential frameworks. The consequences of active macroprudential policy for the conduct of monetary policy will be material, but still need to be understood in light of experience. The Asian experience, your experience, will be extremely helpful in refining macroprudential frameworks and managing expectations as to what they can deliver.

1 FSB-BIS-IMF, "Macroprudential policy tools and frameworks: Update to G20 Finance Ministers and Central Bank Governors", February 2011.