RBA Annual Conference – 2017 Wrap-up Panel Discussion

The final part of the Conference was a panel discussion focused on the use of monetary and macroprudential policies to address financial stability in a world of low interest rates. The discussion was moderated by John Simon, Head of Economic Research at the Reserve Bank of Australia, and included the following panellists:

• Ricardo Caballero, Professor of Economics at the Massachusetts Institute of Technology
• John McDermott, Assistant Governor and Head of Economics at the Reserve Bank of New Zealand
• Cecilia Skingsley, Deputy Governor of the Sveriges Riksbank.

As the Conference and panel were conducted under the Chatham House Rule, no individuals' comments are attributed.

1. Introduction

The session opened with the moderator reiterating that the questions and issues covered in the conference were live ones that many countries, including Australia, were trying to navigate.

Two panellists then discussed their experiences in implementing macroprudential policies to address financial stability risks. They outlined the economic backdrops and risks that had prompted action, what policies were implemented, the processes for implementing these policies, and whether the policies had effectively addressed the relevant stability risks. The panellists also discussed a number of lessons they had learnt from these experiences.

2. Importance of Institutions

The two case studies highlighted that institutional arrangements play an important role in determining the nature and effectiveness of macroprudential policies. This was a major theme of discussion throughout the session. In one of the two countries, the body with the mandate for macroprudential policies had fairly narrow powers and needed legislation to be passed in order to implement its desired policies. This delayed implementation significantly. In contrast, in the other country the relevant body had much broader powers. This allowed it to implement polices more quickly and easily, without the need for formal legislative approval. Nevertheless, the body still consulted with the government to ensure they had adequate support from other public institutions.

In this context, one panellist noted that Rochelle Edge and Nellie Liang's work on the institutional arrangements around macroprudential policy, presented earlier in the conference, is important. Institutional arrangements tend to form organically over time based on country-specific factors and history. Thinking properly about what potential structures could be implemented, and what the relative strengths and weaknesses of these structures are, is likely to allow countries to better direct the development process.

Nevertheless, while the panellist agreed that institutional arrangements are important, they argued that other factors, like institutional culture, are also relevant. Regulators, and the financial system more broadly, need to have a culture of trying to serve the greater good, rather than one of self-interest. Referring to an earlier session, the panellist also cited the importance of coordination between public sector institutions. This can take many forms, including the provision of public ‘moral’ support. Another panellist noted that such public support is particularly crucial in the context of macroprudential policies as these policies involve saying ‘no’ to people, which will always be unpopular.

3. Future Research

The third panellist focused their comments on three areas for future research. The first area is networks – the linkages between different agents and institutions in the economy. They noted that understanding networks is crucial for understanding how shocks to individual households or institutions can interact and proliferate, and affect the macroeconomy.

Better understanding networks would also facilitate further research into the use of reverse stress testing. Traditional stress testing examines whether the system can withstand a given set of shocks and so requires the user to calibrate specific scenarios. This could be difficult for countries that have experienced relatively few crises and so have relatively few empirical observations upon which to base scenarios. In contrast, reverse stress testing involves identifying the types and magnitudes of shocks that would be sufficient to cause a crisis. This would allow countries to identify weaknesses in the system that have not previously been exposed or identified in past crises.

The third area for future research is how best to build risk into macroeconomic models. The panellist noted that risk and financial systems tend to be appendices that are added onto existing models, but that in reality they are central to the economy – as exemplified by the global financial crisis. To this end the panellist outlined a stochastic model of the economy with risk on the production side of the economy. They showed that such a model can help to explain the simultaneous occurrence of low levels of economic growth and high asset price growth that has been observed in a number of countries in recent years. Moreover, the model can help to elucidate the potential benefits of macroprudential policy.

A number of participants agreed that more research should be carried out into embedding risk and financial sectors into macroeconomic models, and improving these models more generally. One contended that the most important graph presented at the conference showed real-time estimates of output gaps leading up to the crisis, and compared them with current estimates of what the output gap was at the time. The current estimates show that the US economy was overheating in the lead-up to the financial crisis, whereas the real-time measure did not. They argued that this showed that, with the benefit of hindsight, monetary policy should have been set very differently. Another participant suggested that more research is needed to improve the measurement of risk and our understanding of when and why the economy ‘switches’ to periods of high risk and volatility.

Panellists and participants also noted a number of areas for research related to the effectiveness of macroprudential tools. These included: how and whether macroprudential policies make balance sheets more resilient; whether macroprudential policies are more or less effective when interest rates are low; and how macroprudential tools interact with monetary policy. One panellist also suggested that further research is needed on the interaction between new liquidity regulations and monetary policy.

4. Difficulties Associated with Macroprudential Policies

When asked about their key takeaways from the conference, two panellists cited the difficulties in implementing and assessing the effectiveness of macroprudential policies. One suggested that some of these difficulties reflected the nature of macroprudential policies. Unlike monetary policy, where the central bank plays a coordinating role in ensuring mutually beneficial price stability, macroprudential policies are aimed at constraining people's behaviour. Still, the panellist argued that policies could potentially become more effective as policymakers became more ‘credible’, citing parallels to the anchoring of inflation expectations in an inflation-targeting regime.

Picking up on the theme of difficulties in constraining people's behaviour, a number of panellists and participants contended that macroprudential policies could cause riskier borrowing to shift to less heavily regulated parts of the financial system, and to the creation of ‘grey markets’. While conceding the point, one panellist argued that a way to avoid this regulatory leakage was to indicate that the policies are temporary. This would limit the incentives for people to invest in creating these markets. Moreover, coordinating behaviour between banks can help ensure a cooperative equilibrium.

Another participant argued that more fundamentally, while macroprudential polices can be used to treat the symptoms, it is crucial to deal with the root causes: distortionary policies that create misaligned incentives. This sentiment was echoed by a number of other participants. In particular, one noted that they were sceptical of using macroprudential policies to control risk-taking behaviour, as individuals are better judges of what risks they can manage. Rather, the issue is a lack of adequate information, and policies should be aimed at addressing information asymmetries through better product disclosure rules. Others were less convinced by this, noting that even with better disclosure it is difficult for individuals to know the future and that they tend to use simple heuristics in making decisions. Moreover, there are numerous agency issues and externalities that would be difficult to address using disclosure rules.

The moderator asked the panellists whether the difficulties and uncertainties associated with macroprudential polices are of sufficient concern that policymakers should eschew macroprudential tools. All three agreed that this was not the case. One suggested that policymakers can't wait to gather more evidence, but need to act. However, they need to do this carefully and test different tools as they do. This was seconded by another panellist, who suggested that the only way to learn about these policies was to use them. If they fail, at least we will have learnt something. The third panellist echoed the others, noting that it was probably safer to overdo the policy response, rather than to underdo the response. Nevertheless, one of the panellists suggested that the difficulties probably mean that monetary policy will remain the first line of defence against financial stability risks.

5. The Role of Monetary Policy in Financial Stability

In contrast, the other two panellists argued that macroprudential policies should be the first line of defence against financial stability risks, though there may be a secondary role for monetary policy.

One panellist argued that there was some scope to use monetary policy as a secondary tool when doing so does not conflict with a central bank's key role of achieving low and stable inflation. To this end, a central bank should only consider ‘leaning against the wind’ when inflation is near target and inflation expectations are well anchored. Carrying on this discussion, a number of participants pointed out that there is not always a trade-off between ensuring financial stability and maintaining price stability. They suggested that one of the biggest risks to financial stability is unemployment, and that lowering interest rates can simultaneously lower unemployment and raise inflation back towards target.

The other panellist noted that the significant degree of interaction between macroprudential and monetary policies meant that they both need to be set with the other in mind. For example, macroprudential policies affect the allocation of credit in the economy and so could affect the pass-through of monetary policy just as the level of interest rates affects risk-taking. This line of discussion was picked up by the participants. One argued that macroprudential policies are essentially trying to constrain the interest-sensitive sectors of the economy, which will necessarily make monetary policy less effective. Instead, if policymakers are worried about financial stability risks it may be better to use fiscal policy to stimulate the economy, as fiscally sensitive sectors are likely to be different to interest-sensitive sectors.

The interactions between the policies were also discussed more broadly. Participants noted that, if financial imbalances are supporting the economy, implementing macroprudential polices aimed at reducing these imbalances may lead to an economic slowdown and the need for more stimulus. At the same time, lowering interest rates could encourage more risk-taking and therefore essentially offset any financial stability gains stemming from macroprudential policy.

6. The Role of Macroprudential Policy in Financial Stability

There was also a discussion of the role of macroprudential tools and, in particular, whether they should be seen as cyclical or structural tools. In general, panellists and participants were sceptical about using macroprudential tools in a countercyclical manner. One panellist stated that, while they found the idea appealing, they believed that policymakers do not have a sufficient understanding of the business and financial cycles to do so. As such, the focus needs to be on making the system more resilient.

Echoing this, one participant argued that they find it hard to imagine that any of the supposedly cyclical tools that have been employed around the world will actually be removed. In this sense, the polices just become ‘good regulations’, which contribute to banking system resilience by helping to ensure that appropriate lending standards are employed throughout the boom and bust. While broadly agreeing, another participant noted that, as yet, we do not have a clear understanding of if and how these tools actually contribute to structural resilience.

The importance of avoiding Ponzi schemes was also discussed in the context of ensuring financial system resilience. A number of participants noted that Ponzi schemes and predatory lending activities are generally associated with particularly severe economic and financial crises. Policies aimed at preventing such activities could therefore play an important role in ensuring financial system resilience. To this end, consumer protection regulations could help to promote financial stability.

There was also a broader discussion of the role and purpose of macroprudential policies. One participant commented that they found it surprising that macroprudential policies generally focused on setting limits and quotas on the quantity of lending, rather than changing the price of lending more directly. Another remarked that most of the discussion had centred on using macroprudential tools to ensure that the banking system is resilient, but that the resiliency of households' and businesses' balance sheets is also important. Building on this, the moderator noted that one of the crucial questions is what risks are we really worried about, and that it is difficult to design policies without knowing what they are intended to address.