RBA Annual Conference – 2017 Discussion

1. Mark Zelmer

Let me begin by thanking John Simon and Tams Pretty for your hard work in organising this conference. I appreciate having the opportunity to escape an Ottawa winter and join you here in beautiful Sydney. I would also like to thank Luci and Charles for such an interesting paper. It is a testament to the excellent working relationship between the RBA and APRA that the two of you could combine forces to offer us a common view on the evolution of Australian financial stability policy, and the current challenges you are facing from rising household indebtedness and high housing prices.

Reading your paper and listening to your presentation was a good reminder of the many things Australia and Canada have in common when it comes to the structure and vulnerabilities of our financial systems. But it also reminded me of some differences that exist, both in our histories and in how we are approaching similar challenges. In the interest of stimulating discussion, I will focus on the latter.

In doing so, I want to leave you with three questions to ponder for the general discussion.

1. Are the close working relationships that exist between the central bank and prudential regulator as robust and effective as we might like to think?
2. Can one assume that the exchange rate will always be so well behaved in an inflation-targeting regime with a floating exchange rate?
3. Different countries have been dealing with the household debt and housing price issue in different ways. To what extent are structural differences across our housing finance systems responsible for our different policy responses?

Let me take each of these questions in turn.

Relationships among financial sector authorities need to be constantly nourished with a system-wide perspective

The institutional settings for cooperation between the central bank and prudential authority are similar in Australia and Canada, though the histories of how we got here are rather different. As in Australia, we in Canada have committees that foster close working relations among the agencies tasked with overseeing the financial system. And staff at all levels have forged good working relations over their careers. Unlike the RBA, however, the Bank of Canada never had any responsibilities for overseeing individual banks and insurance companies. Instead, Canada has had a long history of delegating these tasks to separate agencies with specific mandates granted to their heads by Parliament.

It is easy to take this strong working relationship for granted. Looking around the world I have often been surprised at the rivalries that exist in many, if not most, countries among central banks, finance ministries and financial regulatory agencies. Not to mention that in central banks with supervisory responsibilities, some supervisory staff may feel that they are not loved as much by their executive parents as their monetary policy siblings.

Charles and Luci offer some good advice on how one can forge good relations across agencies. Much of what they said rang true for me in Canada as we have similar interagency committees, memorandums of understanding and strong personal working relationships at all levels.

But I have to ask whether these arrangements have been truly fully tested? Certainly, the relationships among the agencies worked very well for Australia and Canada during the global financial crisis; but we should remember that neither of us was confronted by the prospective failure of a major domestic institution during that crisis.

I believe that the true test of those arrangements will be if we ever find ourselves having to manage such a domestic crisis. You might be surprised at how what may appear to be good working relations before a crisis can quickly become contentious in the midst of a crisis when the narrow interests of each agency start to diverge. Thus, I encourage you to continue what you are doing to foster strong working relations, anchored to a system-wide focus. No doubt you are doing this already, but I would also encourage you to look for ways to encourage your staff to spend time during their careers working in different agencies to help further cement a system-wide focus. I know I certainly benefited from the insights and perspectives I gained from working in more than one institution over the course of my career.

Before I close off this issue, let me also note that both Australia and Canada are rather unusual in giving so much discretion to prudential authorities to set regulatory requirements and supervise individual institutions with limited political oversight. That discretion is valuable, but fragile.

There are many ways it could be lost if we are not careful. If we are too strict, and are not sufficiently attuned to innovation and competitive pressures, financial institutions and their customers could start advocating for more political oversight to ensure that prudential regulation meets their day-to-day needs. Conversely, if prudential supervisors are too lax, the financial crisis is a good reminder that major failures can be a recipe for political intervention that can displace supervisors.

We should not take our independence for granted. We need to continue looking for ways to demonstrate that we are looking after the prudential interests of depositors and policyholders, while respecting the fact that institutions take risks, are responsible for their own behaviour, and need to be allowed to compete to meet the needs of the public they serve.

Exchange rate movements can complicate life for central banks

Time is marching on, but I can be more succinct in my next point. At a macroeconomic level Australia's history with its exchange rate certainly resembles that of Canada's over the past 20 years. So the lessons drawn in the paper for monetary policy certainly apply most of the time for us in Canada too. But not all the time.

Looking further back in time, I recall that in the first half of the 1990s the Canadian dollar and Canadian debt markets were often buffeted by foreign capital outflows. These flows reflected investor preoccupation with political uncertainty in Canada and the state of public finances, despite our currency having been allowed to float for many decades. As a result, the Bank of Canada often had to pay more attention to exchange rate developments in conducting monetary policy than was the case in Australia.^[1] While this may not be an issue for Australia now, or for the foreseeable future, it is a caveat that other countries should not ignore – especially if public finances and political uncertainty become bigger issues in the future.

Dealing with rising household indebtedness and high housing prices

Let me now conclude with some remarks on rising household indebtedness and high housing prices – the key domestic vulnerabilities confronting both of our financial systems. The Canadian response to these vulnerabilities has differed from that of Australia's in some important ways.

First, the role of the state in housing finance in Canada is more pervasive than in Australia, which has given us more avenues for tackling this vulnerability. All residential mortgages in Canada with loan-to-valuation ratios in excess of 80 per cent at time of origination must be insured against default by law, and the mortgage insurance market is dominated by a federal-crown corporation that operates with a full federal government guarantee. Even the private mortgage insurers operating in Canada benefit from a 90 per cent federal government guarantee, in the event that they cannot meet their claims. On top of that, investors in Canada have only had an appetite for mortgage-backed securities (MBS) consisting of insured mortgages. And, that MBS program is also operated by the federal-crown mortgage insurer, which offers a timely payment guarantee to investors.

While one can debate the merits of such a pervasive state presence in housing finance, it has given the Canadian Government more levers to influence lending behaviour. It allows them to control the terms and conditions governing access to mortgage insurance and mortgage securitisation, over and beyond the measures that can be taken by a prudential regulator or a central bank. The government has not been shy in this regard, as evidenced by the numerous steps taken over the past 10 years to gradually tighten up lending terms and conditions for insured mortgages and the mortgage securitisation market.^[2]

Second, in contrast to APRA, the Office of the Superintendent of Financial Institutions (OSFI), Canada's prudential supervisor, has striven to avoid setting broad-based floors on capital requirements. This reflects a desire on OSFI's part to avoid dis-incentivising low-risk lending. Moreover, in Canada, the dynamics of housing-related vulnerabilities vary considerably across the country, given differences in housing market conditions across Canada. Thus, a more targeted approach has been required in setting bank regulatory capital requirements for uninsured residential mortgages that makes allowances for regional differences (Zelmer 2015).

OSFI also prefers to have capital requirements that operate more mechanistically in the housing finance space given the high public attention devoted to housing finance issues.

The point is that the institutional setting governing household lending differs considerably across countries, even between two countries like Australia and Canada that have similar banking systems. This can make it challenging to draw lessons from other countries.

Finally, I have a quibble with one of the arguments in the paper. I personally do not think that one should downplay increases in household indebtedness that arise as nominal interest rates decline in response to lower inflation. It is true that lower interest rates make it easier for households to carry larger debt burdens, if one is speaking of a sustained decline in real interest rates. But the same is not true if nominal rates fall in response to lower inflation. Why? Because one also needs to take account of the associated decline in future nominal income growth, which will mean that the debt inflates away more slowly. So, while the household may be able to service a larger loan due to the lower nominal interest rate, it will also have a higher level of debt for longer.

That may not pose much of a macroeconomic or financial stability issue in the near term; but it cannot help but reduce the resiliency of the household sector over the medium term. If nothing else, it may affect the responsiveness of households to future macroeconomic policy interventions, be they fiscal or monetary. Clearly this has not been an issue in the past for Australia, but I would not ignore it in the future, given household indebtedness levels are much higher today than they were in the past.

Conclusion

Let me close by thanking the RBA again for inviting me here to discuss this interesting paper. I learned a lot about Australia from it, and it made me think more about how things are done back in my home country. I look forward to hearing what the rest of you have to say in the general discussion.

Footnotes

Further information on monetary policy tactics in Canada in the first half of the 1990s can be found in Clinton and Zelmer (1997). [1]

These measures are summarised in the annex of a recent Department of Finance Canada consultation document (Department of Finance Canada 2016). [2]

References

Clinton K and M Zelmer (1997), ‘Constraints on the Conduct of Canadian Monetary Policy in the 1990s: Dealing with Uncertainty in Financial Markets’, Bank of Canada Technical Report No 80.

Department of Finance Canada (2016), ‘Balancing the Distribution of Risk in Canada's Housing Finance System: A Consultation Document on Lender Risk Sharing for Government-Backed Insured Mortgages’, Consultation document, October.

Zelmer M (2015), ‘Updating Capital Requirements for Residential Mortgages’, Letter to Industry published by the Office of the Superintendent of Financial Institutions, 11 December.

2. General Discussion

The discussion began with one participant noting that tensions can arise between prudential regulation and monetary policy when conditions differ between sectors of the economy. In particular, when there are negative shocks to non-interest-sensitive parts of the economy, using monetary policy to boost economic activity and inflation could exacerbate vulnerabilities in other more interest-sensitive sectors of the economy, such as housing. The participant gave the example of the dot-com bust, when US monetary policy was used to stimulate the economy following a shock to the (non-interest-sensitive) technology sector, and mused that similar dynamics may be evident in Australia following the end of the mining boom.

This led to some discussion of the appropriateness of different macroprudential tools. One participant noted that in the situation where conditions differ between sectors of the economy, it is important to choose tools that will directly affect the interest-sensitive sectors, without impinging on the less sensitive sectors. Tools such as countercyclical capital buffers are likely to be too blunt, especially given that activities such as business lending tend to be capital intensive, while housing lending tends to be less capital intensive. In an environment in which monetary policy and prudential regulation are moving in the same direction, these tensions are less of an issue. In either case, communication and coordination between monetary and prudential authorities is likely to be helpful in picking the most appropriate tools.

One participant raised the issue of culture within financial institutions, suggesting that institutional culture was a relevant factor for financial stability. For example, regulatory actions in protecting banks and the financial system more generally could lead to moral hazard and excess risk-taking. It was acknowledged that a ‘toxic’ culture within financial institutions could be detrimental, but that regulators are limited in their ability to assess culture or to intervene. In this sense, regulators need to try, but try humbly, and should use expertise from other public sector institutions.

Discussion turned to the feasibility and appropriateness of using macroprudential tools in a countercyclical manner. In terms of feasibility, one participant asked whether it would be possible under current governance structures to ease prudential policy in a downturn. It was noted that, in Australia's case, prudential regulation is more stringent than the Basel III requirements, meaning there was scope to ease policy in a downturn. More generally, one participant opined that if tighter macroprudential policy is imposed during a boom, it may be easier to motivate an easing in policy during the downturn, since policy would be returning to normal. However, it was noted that it could be difficult for banks to adjust their balance sheets in response to a policy easing during a downturn if they might be penalised by the market, for example, through higher funding costs.

Political constraints on using macroprudential policies in a countercyclical manner were also discussed. It was noted that prudential changes can be unpopular, especially at times when they are most important. In a boom, the need for regulation can be questioned; in a downturn, regulators may be perceived as being too lax. For this reason, central bank support for prudential changes is important, particularly in the form of public moral support.

On appropriateness, participants acknowledged that keeping bank lending markets open during a downturn is crucial to any recovery. In this context, prudential policies need to ensure that banks are strong enough to step in to fill any void left by institutions that fail during a crisis. The example of major Australian banks filling the void after European banks left the domestic market following the financial crisis was cited in this context. Participants suggested that countercyclical macroprudential policies could help to keep lending markets open.