Jens Weidmann: All for one and one for all? The roles of microprudential, macroprudential, and monetary policy in safeguarding financial stability

Speech by Mr Jens Weidmann, President of the Deutsche Bundesbank, at the Deutsche Bundesbank symposium on financial stability and the role of central banks in Frankfurt, 27 February 2014.

The views expressed in this speech are those of the speaker and not the view of the BIS.

1. Introduction

Ladies and gentlemen

About six years ago, on 17 February 2008, the British government decided to nationalise Northern Rock. The bank run that preceded this decision was the first run on a British bank in over a century.

As we know now, the worst was yet to come. The fall of Lehman Brothers seven months later sent shockwaves through the global economy, marking the onset of what is now known as the Great Recession. Still today, many economies are struggling with the aftermath of the crisis, not only economically, but also socially and politically.

The tectonic shifts in financial conditions that erupted so violently took place beneath a seemingly calm macroeconomic surface. Inflation was low and stable, and estimated output gaps were small. Thus, when taking stock of the lessons learned from the crisis, one lesson seems rather straightforward. Price stability is a necessary, but not a sufficient condition to guarantee economic stability. Hence, financial stability has to be in the mix as well.

Another lesson the crisis taught us is that, in order to safeguard financial stability, it is not enough if financial regulation and supervision focus on the soundness of individual institutions. The interlinkages between different financial actors and between financial institutions and sovereigns have been a key factor in explaining the crisis.

How, then, is financial stability to be maintained? What went wrong in the run-up to the crisis? Have we drawn the right conclusions from what has been the worst financial crisis since the Second World War? And is the implementation of the lessons learned, i.e. the regulatory agenda, proving sufficient to make a similar crisis unlikely?

These are the overarching questions of this Bundesbank Symposium. It's a great honour for the Deutsche Bundesbank and a great pleasure for me to welcome you to our symposium, and to see that so many distinguished representatives from academia, central banks, the financial sector, politics, and unions, have accepted our invitation. I am fully aware that some of you - like myself - have come more or less directly from the G20 and BIS meetings held last weekend in Sydney, but I am sure that the lively debate ahead of us on that very relevant topic is the best way to overcome the jetlag.

Karl Popper once said that "without a free exchange of ideas there can be not freedom of thought. To find out whether our ideas are sound, we need other people to try them out on." This is what this event is all about. So, to kick off our exchange of ideas, let me try out some thoughts on you as to how financial stability might best be achieved.

2. The roles of monetary and macroprudential policy in macroeconomic management

If anybody or anything has actually emerged from the crisis a winner, it is probably the word "macroprudential". Microprudential policies such as banking regulation and supervision are still necessary in safeguarding the soundness of individual institutions. But as it turns out, they are not enough when it comes to safeguarding the financial system as a whole. For the management of systemic financial stability risks we need an additional toolkit - macroprudential policies. And it seems to me that this assessment is by now widely shared.

The role of monetary policy in this context is, however, still being debated. While macroprudential policy frameworks are still under development, we need to take a stand on whether and how monetary policy should play a supporting role without undermining its pursuit of price stability.

First, we have to acknowledge that in the world we live in, macroprudential policy can never be perfectly effective - for instance because safeguarding financial stability is complicated by having to achieve multiple targets all at the same time.

While maintaining price stability translates into one single target variable, that is, the average inflation rate related to the price of a consumption basket, maintaining financial stability is more complex. Risks to financial stability can emanate from different sources, such as price bubbles in specific asset markets, but also reckless risk-taking by a single but systemic player, and so on.

While the level and volatility in individual prices of goods and services is usually not of much concern to central bankers as long as the impact on the price index is small, the level of and volatility in particular asset prices could lead to instability of the financial sector. This is why preserving financial stability is a multi-dimensional problem.

Additionally, financial and business cycles do interact and the monetary policy stance does affect risk-taking by the financial sector. Indeed, some current research supports the view that monetary policy seems to have contributed to the high risk appetite in the financial system and thus to the build-up of financial imbalances.

The philosophy of "mopping up" after a crisis that prevailed in the years before the Great Recession has come under attack because it led to the expectation of a dramatic loosening of monetary policy in the event of a financial crisis. Such an asymmetric policy, in turn, may itself generate an additional incentive to take on higher risks and thus contribute to financial and macroeconomic instability.^{1, 2}

Hence, some observers argue that price stability and financial stability are so closely intertwined that, in the case of both upturns and downturns, monetary policy has to play an active role in stabilising the financial system. Some even call for a dual mandate for central banks, with financial stability on a par with price stability.³

But in my view, even if monetary policy instruments were an efficient tool to influence financial stability - a premise which I would still doubt - an important argument against the comprehensive involvement of monetary policy in financial stability lies in the risk of losing credibility with regard to the pursuit of price stability.

The credibility of monetary policy depends both on the clarity of its objectives and on transparency with regard to its limitations. Adopting financial stability as an additional monetary policy objective on a par with price stability would risk raising unrealistic expectations about the effectiveness of the monetary policy instruments. It could create trade-offs which would undermine the primacy of the price stability objective, reflected for instance in the form of an inflationary bias.

Tasking the central bank with both financial as well as price stability might give rise to a time-inconsistency problem. In such a setting, a central bank should find it optimal to deliver the social optimal level of inflation ex ante, but it might not do so ex post.⁴ The time-inconsistency problem may arise because macro-prudential policy is not perfectly effective in preventing financial imbalances and it is only a pre-emptive tool, while monetary policy can pre-empt as well as correct financial imbalances. This is because macroprudential policy can affect only flows of credit but not the outstanding stock of debt. By contrast, monetary policy can also reduce the real stock of debt by creating surprise inflation.

Moreover, studies being more optimistic with regard to the efficacy of macroprudential instruments suggest that relying on macroprudential policies for financial stability purposes yields socially more desirable results, because macroprudential instruments allow for a targeted policy response and thus minimise output and inflation volatility.^{5, 6}

To put it succinctly: when it comes to handling financial stability risks, monetary policy is a sledgehammer rather than a scalpel. So, the latter should be the preferred choice of tool, and getting it ready for use has to be a political top priority.

I am therefore convinced that in order to enable monetary policy to fulfil its mandate and to ensure clear accountability, it's important that price stability remains the primary objective of monetary policy. But that is not to say that when it comes to monetary policy and price stability there are no lessons to be learned. Or "re-learned", as Claudio Borio has pointed out with the words, "Understanding in economics does not proceed cumulatively. - So-called 'lessons' are learnt, forgotten, re-learnt and forgotten again."

It is therefore more of a reminder from the crisis than a discovery that monetary policy decisions must attach greater importance to financial sector developments, as risks to financial stability may ultimately culminate in risks to price stability - a topic, I am sure, that will be discussed at greater length during our first session today.

Indeed, recent research⁷ reaffirms how taking into account credit developments can enhance the stabilisation of the overall economy: output volatility can be dampened without increasing inflation variability. The Eurosystem's monetary policy strategy is in principle rooted in such an approach, as real economic analyses are regularly cross-checked against those from monetary analyses. Nonetheless, the Eurosystem has stepped up its efforts to continually enhance its monetary analysis, including early warning indicators for unsound developments in the financial markets.⁸

In this context, I would like to stress that what I call a "more symmetric monetary policy" approach of the Eurosystem is not identical to a prototypical approach of "leaning against the wind". Roughly speaking, a "leaning against the wind" strategy can be seen as an unconditional monetary policy response to asset price changes. In contrast, a "more symmetric monetary policy" advocates a conditional policy response.

By this I mean that in the context of our monetary analysis, we identify longer-term risks to price stability based on money and credit developments.⁹ The critical condition is therefore whether asset price movements and credit developments give rise to monetary and financial imbalances, which are ultimately relevant for price stability.

In this sense, our monetary analysis acts as a filter to extract the monetary policy relevant movements in financial markets. This prevents us from being myopic, helps us to conduct monetary policy more symmetrically across the financial cycle and thereby contributes to overall financial stability - but always on the basis of our price stability objective.

However, fundamental challenges remain with this approach. As Claudio Borio¹⁰ has pointed out, the financial cycle has a much lower frequency than the traditional business cycle. "Since financial liberalisation, its typical length is of the order of 16 to 20 years; by contrast, as generally conceived in academic and policy work, business-cycle frequencies are up to eight years." So, while extending the monetary policy decision-making horizon is warranted in principle, we are restricted in practice by our limited ability to reliably forecast over longer horizons. For instance, early warning indicators might perform well ex post, but do not necessarily exhibit the same performance ex ante.

We certainly need to look into all of these issues in more detail - I am looking forward to our discussion on these topics today. When it comes to financial stability, monetary policy has a role to play - insofar as price stability is at risk as well. But macroprudential policies, with their more versatile and precise instruments, should provide the first line of defence against financial instability. In many cases, the interest rate will prove too crude an instrument with which, for instance, to prick asset bubbles.

3. Micro- and macroprudential regulation

But it is not only the macroeconomic toolbox that needs an overhaul. The crisis has laid bare structural shortcomings of the financial system. These structural flaws demand a structural - that is, a regulatory - answer.

Market failures and regulatory blind spots have tilted the playing field, hampered competition, and over-stimulated the risk appetite of investors. As you all know, much effort has already been taken to address the fault lines revealed by the global financial crisis and to move towards a more resilient global financial system.

At its meeting in Sydney last weekend the G20 agreed, to focus its current work programme on 1) building resilient financial institutions, 2) ending "too big to fail", 3) addressing shadow banking risks, and 4) making derivatives markets safer. To be more concrete, I would like to revert to one key characteristic of the financial market that also regulation failed adequately to take into account in the past: the interconnectedness of its players.

One of the main characteristics of financial institutions is that the failure of one institution might pull others down with it and cause a systemic crisis if the failing institution has, because of its size, a significant influence on financial markets or its business partners or does business with a lot of partners. The term "too interconnected to fail" might therefore be even more appropriate for this well-known "too big to fail" problem.

The insolvency of Lehman Brothers that I mentioned earlier is a case in point. What was lacking then was an insolvency framework that allowed the failure of this institution without destabilising the entire financial system. More than five years after the Lehman insolvency, the "too big to fail" problem still poses a real threat.

Work has been under way to transpose the Financial Stability Board's "Key attributes of effective resolution regimes for financial institutions" into national law, but we are not there yet, and we must not slacken in our effort - for instance to ensure that G-SIFIs have sufficient resources available to absorb losses in resolution, or to develop resolution rules for financial institutions of other sectors as well, such as insurers.

In Europe, one major bit of progress is undoubtedly the agreement on the Bank Recovery and Resolution Directive. The directive sets forth resolution instruments and the sequence of liability for failing banks. In future, shareholders and creditors will be first in line to bear the costs of bank failures - rightfully so, since they assumed the risks and benefited from the respective returns in the first place.

Containing the challenge of interconnectedness implies the necessity, first, to increase the loss absorbing capacity of institutions, thus making them less likely to fail to begin with and, second, to limit counterparty exposure. In the end, both ways also lead to a better balance of liability and control, which is crucial for the correct functioning of financial markets and, hence, for financial stability.

The new Basel III international capital rules increase the loss absorbing capacity of banks because they now have to hold more and better capital than before. But to some, the risk-weighted approach of the Basel rules is fundamentally flawed.

Critics such as Martin Hellwig, who will elaborate on his views during tomorrow's panel discussion on banking regulation, maintain that the Basel approach causes the rules to miss substantial risks. And he argues that banks are using the risk weightings to leverage their positions up to a hundred times the equity invested.

Still, the reasoning behind risk weighting sounds plausible. If all assets had to be backed by the same level of capital, banks would have an incentive to opt for riskier investments, as this would raise their expected return on equity. This, of course, would do a disservice to financial stability.

Mark Carney therefore has argued for a pragmatic approach: sometimes you need both a belt and suspenders if you are at risk of losing your pants. Hence, the risk-weighted Basel rules should be supplemented by an unweighted leverage ratio.

In my opinion, that sounds sensible. To economists, parsimony is normally key when it comes both to economic modelling and to regulatory proposals. But in this special case, putting multiple safety measures in place might be justified, even necessary. What applies to power stations holds for the financial system as well. When the potential fallout is vast, redundancy, not parsimony, might be the better approach - a point, I am sure, we will have time to delve into further during tomorrow's panel discussion on regulatory reforms.

It is obvious, however, that stricter regulation for some institutions might shift business to less regulated players. Regulators need to take this aspect of interconnectedness into account, too. The so-called shadow banking system can be a source of systemic risk because of unregulated liquidity and maturity transformation, the build-up of leverage, and procyclicality.

The FSB published policy recommendations to address specific risks emanating from certain activities or institutions. In order to identify new risks in due time, the FSB has established a regular monitoring exercise. These are steps in the right direction.

For the international recommendations to be effective, it is crucial to transpose them consistently into national law - and to do so quickly, given that this issue has been on the agenda for quite a while. In Europe, for instance, we are already in the process of implementing international recommendations to regulate money market funds and to enhance the transparency of securities financing transactions.

Finally, there is yet another major aspect of interconnectedness that regulators have not taken into account up to now: the link between banks' balance sheets and public finances. The financial crisis has shown that if a systemically relevant bank or many banks simultaneously run into financial difficulties, the stability of the financial system is put at risk as a result. Given the deficiencies of previous resolution regimes, governments often had no option but to step in and bail out banks to prevent the financial system from collapsing.

In Europe, however, the Greek case in particular showed that the other way round can also be of relevance. Difficulties in government finances might cause substantial problems for banks, in particular because many banks hold large portfolios of government bonds.

From a financial stability viewpoint this vicious circle has to be broken. As I discussed earlier, resolution regimes for banks are important in this regard. Having credible resolution frameworks in place with clear bail-in rules of shareholders and creditors would help to shield the government from having to rescue banks with taxpayers' money.

Another crucial measure to cut the sovereign-banking nexus refers to the regulatory capital requirement regime for banks. One of the key lessons of the crisis was that sovereign bonds cannot per se be considered risk-free. We should therefore put this regulatory fiction to rest.

But sovereign bonds pose a threat to financial stability not only because of preferential risk-weighting. The most important rule in risk management is diversification. Yet when it comes to sovereign bonds, banks all too often seem to neglect this principle. In many cases, European banks hold bonds from one sovereign only - their home country. Large and undiversified sovereign exposure is what makes sovereign default a potentially systemic event. Hence, the large exposure regime which caps the investment in one single debtor has to be applied to sovereigns as well.

Without addressing the regulatory treatment of sovereign exposures, I see no reliable way of breaking the sovereign-banking nexus. In my view, such regulations are an indispensable prerequisite for a healthy financial system in the longer run.

4. Conclusion

Ladies and gentlemen, let me now come to a close.

The challenges I have just described make one thing very clear: safeguarding financial stability is truly a herculean task. It would therefore be an understandable reflex to summon all policies at our disposal towards that end. As D'Artagnan, Athos, Porthos and Aramis would have said: "All for one, and one for all!"

But while indeed all the policies I have mentioned - microprudential, macroprudential, and monetary policy - have a role to play when it comes to securing financial stability, some restraint seems to be in order, especially with regard to monetary policy.

Monetary policy can influence financial conditions. Notably, it can, to a certain extent, influence the risk-taking behaviour of the financial sector. A more symmetric monetary policy that looks through the financial cycle therefore seems appropriate. Monetary policy has to counter financial imbalances inasmuch as they endanger price stability.

But adopting financial stability as an additional monetary policy objective on a par with price stability overstates the accuracy and the effectiveness with which monetary policy can contribute to financial stability. And ultimately it runs the risk of harming credibility with regard to the pursuit of price stability, as this depends on both monetary policy clarity regarding its objectives and transparency regarding its limitations.

This implies, however, that we must continue in our efforts to fully develop macroprudential policy as soon as possible; and that the international recommendations on financial regulation have to be transposed quickly and consistently into national law.

I now look forward to an exciting symposium!

---