Revisiting monetary policy frameworks in the light of macroprudential policy
Panel remarks by Mr Jaime Caruana, General Manager of the BIS, at the IMF seminar on "Revisiting monetary policy frameworks", Lima, 10 October 2015.
Abstract
The often proposed separation principle - whereby monetary policy is assigned to price stability and macroprudential policy is assigned to financial stability - has an intuitive appeal, but is unconvincing as a general proposition. These panel remarks offer three reasons why the separation principle is not compelling, and address three often cited concerns about using monetary policy to deal with financial imbalances. The conclusion is that, given the damage that financial instability can inflict on the real economy, policy priorities need to be rebalanced towards mitigating financial booms and busts. Relying exclusively on macroprudential policy to tame the financial cycle is insufficient. There is a case for including monetary policy in this effort.
It is a real pleasure to be part of this panel. In my remarks, I would like to take a look at the nexus of monetary and macroprudential policies.
I will point out that the often proposed separation principle - whereby macroprudential policy should deal with financial booms and busts (the financial cycle) while monetary policy should deal with inflation and shorter-term output fluctuations (the business cycle) - is intuitive, but unconvincing as a general proposition. I will first offer three reasons why the separation principle is not compelling. I will then address three typical concerns about using monetary policy to deal with financial imbalances.
The conclusion I draw is that, given the great economic costs of financial instability, we need to rebalance policy priorities towards reducing the likelihood of such instability. While there is some acceptance of the idea that monetary policy has a role to play in leaning against the financial cycle in some circumstances,1 I will argue that this case applies much more generally.
Given how powerful monetary policy is in affecting the price of leverage, credit growth, asset prices and financial risk-taking, simply arguing that it forms the last line of defence is inadequate and somewhat risky. It assigns too modest a role to such an influential policy. The proposed rebalancing of policy priorities will no doubt require additional analysis. But relying exclusively on macroprudential policies to tame the financial cycle would simply be insufficient.
In fact, I would go further and argue that even using both macroprudential and monetary policies may still be insufficient in some situations because the endogenous build-up of financial imbalances can be very powerful. In such cases, policymakers will also need other policies - not only prudential/macroprudential and monetary, but also fiscal policy or even structural reforms - to address the imbalances.
Part of the problem in discussing the costs and benefits of alternative policies is that current models and traditional analytical approaches take little or no account of the endogenous cumulative effects of interest rates being too low for too long. They tend to assume that monetary policy has limited influence on the financial cycle - and hence on the costs of financial booms and busts. The international dimension of monetary policy, the spillovers and spillbacks, also tend to be underestimated.
Why the separation principle is not convincing
The separation principle has the merit of yielding clear and neat policy assignments: monetary policy for price stability and macroprudential policy for financial stability. However, the real world is more complicated than that. This principle is therefore not very convincing as a general rule, for three reasons.
Same channels of influence
First, both macroprudential policy and monetary policy fundamentally influence the same channels - funding costs, leverage incentives and risk-taking - which, in turn, affect credit growth, asset prices and the macroeconomy. As such, these two policies are not really neatly separable - they often interact and can create tensions. For example, in a situation where policy rates are lowered for business cycle reasons and macroprudential tools are tightened to address credit booms and rising asset prices, economic agents face incentives to borrow more and to borrow less at the same time.
One consequence of this observation, also supported by empirical evidence, is that the two sets of tools are most effective when used as complements, pulling in the same direction.2
Different effectiveness
Second, although macroprudential policy tools have the advantage of being more targeted to specific sectors or practices, experience suggests that these tools are not as effective as monetary policy rates in preventing excessive risk-taking that is widespread across the financial system.
To be clear, macroprudential policy can be helpful in increasing the resilience of the financial system, ie in building buffers that will protect it when a boom turns to bust. Research also shows that some tools, such as requirements on the loan-to-value ratio (LTV) or the debt-to-income ratio, can be effective in influencing credit and property price developments, ie in constraining the build-up of financial imbalances in the first place. That being said, estimates of such effects generally imply that these instruments would need to be tightened by quite a lot in order to be able to contain the typical dynamics during a boom.3
In contrast, the monetary policy rate is the key determinant of the universal price of leverage in a given currency. It is not susceptible to regulatory arbitrage, and it affects all financing in the economy. In particular, if the price of leverage has been too low for a long time, allowing financial risk-taking to take hold and spread across the system, it would then be much more difficult for macroprudential policy tools to address the excessive credit growth and asset price increases.
This point is consistent with the experience of some economies that have made extensive use of macroprudential measures in recent years against the backdrop of very accommodative monetary policy conditions. I can cite, for instance, Hong Kong and Switzerland, among others. Despite the tightening of macroprudential tools such as LTV requirements and countercyclical capital buffers or dynamic provisions, these economies have not been able to fully avert the build-up of financial imbalances.4
This challenge is also reminiscent of Spain's experience in the 2000s, when it found out that dynamic provisions could not be sustained at levels sufficient to contain the credit boom. Moreover, being in a monetary union, Spain could not have independently used monetary policy to deal with the boom. A monetary union is a special case in this regard.
Market-driven booms
Third, there are market-driven booms - and this reinforces my previous points. Financial intermediation has been changing: capital markets are gaining prominence and the search for yield is an active mechanism for transmitting financial conditions across markets. Most of our experience so far with macroprudential tools has come from banking. But now, imbalances are building not so much in the banking sector but in capital markets, which are not within the direct reach of traditional macroprudential tools.
In this situation, monetary policy again has an advantage. By changing the universal price of leverage in a given currency, it affects all financing denominated in that currency and is much better positioned to work in a world in which capital markets are vast and macroprudential tools are narrowly directed at banks.
What I would conclude from all this is that exclusive reliance on macroprudential tools to deal with financial stability risks is insufficient and ill advised. Macroprudential tools can increase resilience. They can address localised issues, such as the overheating of specific markets. And they provide policymakers with additional options to lean against the build-up of financial imbalances. But they cannot "get in all the cracks" in the system, as Jeremy Stein so aptly put it.5 Arbitrage can move the build-up of financial imbalances from one place to another, finding the inevitable cracks that exist in any prudential regulatory regime. For this reason, there is a case for using monetary policy.
Concerns about using monetary policy to deal with financial imbalances
Let me now turn to the typical concerns about using monetary policy to deal with financial imbalances. These concerns should be taken seriously. But I believe they are manageable - and they should be managed.
Lack of good metrics
One often cited argument is the lack of good metrics with which to track the financial cycle. This is a serious concern for policymakers. But one should recognise that the past decade has seen considerable progress in devising and improving such metrics. One practical approach, followed by economies such as Hong Kong, Norway and the United Kingdom when setting countercyclical capital buffers, is to track credit and asset price developments, and to compare current dynamics to historical benchmarks.
At the same time, one should not forget that even the more familiar yardsticks used in monetary policy are themselves not without problems. Take, for example, the output gap, a measure for economic slack, which is not observed directly and thus has to be estimated. It is known that the estimates are subject to considerable uncertainty.6
In fact, some recent research suggests that using information about the financial cycle, such as the behaviour of credit and property prices, can produce better estimates of potential output and underlying slack compared with using traditional methodologies, which often draw on the behaviour of inflation.7 In this sense, metrics informed by the state of the financial cycle may help improve the calibration of monetary policy and fiscal policy.
Policy trade-offs
A second, and more challenging, concern is the potential trade-offs between financial stability on the one hand, and price stability and near-term output stabilisation on the other.
To some extent, this concern can be ameliorated by looking at the relevant policy horizon. Financial vulnerabilities build up over time. And a financial bust can have long-lasting effects on the macroeconomy, including on inflation. Hence, extending the policy horizon beyond the traditional two to three years would help to reconcile the financial stability objective with the traditional price stability (and output growth) objective. After all, financial instability is a concern precisely because of the damage it imposes on the real economy.
However, I should note that extending the policy horizon should not be interpreted as extending point forecasts. Rather, it is intended as a means to examine more systematically the risks to the macroeconomic outlook posed by financial factors, given their longer fuse.
Deviation from mandate
A third concern is deviation from mandate. Given the potential trade-offs I just described, there would be times when the price stability objective (eg inflation target) could not be achieved as quickly as one would like because of financial stability considerations. If one is going to tolerate such deviations of inflation from target, how long should this be allowed to last? And how much importance should one attach to such deviations?
The real concern here, I believe, is the worry that if the deviation from the stated target persists for a long time, it might lead to a loss of central bank credibility. If this is indeed the issue, then our view is that the monetary policy framework should explicitly provide for tolerance of such deviations if and when they are deemed appropriate for achieving its objective over the longer term. Of course, the allowance should be based on proper analysis of the reasons underlying the deviations.
Much less clear, however, is whether allowing such tolerance would necessarily constitute a deviation from the mandate (eg price stability). Central bank mandates are typically worded generally enough to accommodate different ways to interpret and implement them in practice. In particular, given the large negative impact financial crises can have on the real economy, sustainable price stability or macroeconomic stability can indeed be thought of as encapsulating financial stability.
This suggests that the first priority should be to: (i) make use of all the existing room for manoeuvre; (ii) develop a better explanation for why a near-term deviation from target may sometimes be justified for the longer-term good; and (iii) build a constituency for a more systematic incorporation of financial stability concerns into central bank decisions.
In addition, transparency with respect to financial stability policies could be helpful in this regard. Disclosing financial stability decisions and actions, and the reasons behind them, could help to manage expectations about how a central bank would deal with financial stability risks and the potential impact of policy actions. That said, I do recognise that interpreting mandates flexibly in difficult and uncertain times is not at all an easy task.
Conclusion
On balance, arguments against incorporating financial stability considerations systematically into monetary policy, while not without merit, are nonetheless not fully convincing. They tend to overestimate how much is known about the business cycle but underestimate how much has been learned about the financial cycle. They also tend to put too much faith in the ability of macroprudential policy to deal with financial stability risks but underappreciate monetary policy's role in determining the price of leverage and in influencing borrowing and risk-taking behaviour across the board.
Although there may be near-term trade-offs, financial stability and price stability are really two sides of the same coin over the longer horizon. If the ultimate goal of monetary policy is to promote sustainable economic growth, then there is good reason to call for a rebalancing of policy priorities towards mitigating financial booms and busts, which can inflict long-lasting damage on the real economy. Such a rebalancing would be challenging and would confront policymakers with tough questions. But relying exclusively on macroprudential tools to address financial stability risks is simply insufficient. There is a case for including monetary policy in this effort.
1 See IMF, "Monetary policy and financial stability", IMF Policy Papers, August 2015.
2 See V Bruno, I Shim and H S Shin, "Comparative assessment of macroprudential policies", BIS Working Papers, no 502, 2015.
3 See, for instance, K Kuttner and I Shim, "Can non-interest rate policies stabilise housing markets? Evidence from a panel of 57 economies", BIS Working Papers, no 433, 2013; and E Cerutti, S Claessens and L Laeven, "The use and effectiveness of macroprudential policies: New evidence", IMF Working Paper, no WP/15/61, 2015.
4 Early warning indicators measuring financial vulnerabilities for more than 20 countries and regions are presented in Table 1 in BIS, "Highlights of global financing flows", BIS Quarterly Review, September 2015.
5 See J Stein, "Overheating in credit markets: origins, measurement, and policy responses", speech at research symposium sponsored by the Federal Reserve Bank of St Louis, February 2013.
6 See J Weidner and J C Williams, "Update of 'How big is the output gap?'", Federal Reserve Bank of San Francisco, 2015.
7 See C Borio, P Disyatat and M Juselius, "A parsimonious approach to incorporating economic information in measures of potential output", BIS Working Papers, no 442, 2014.
