Intellectual challenges to financial stability analysis in the era of macroprudential oversight

Speech by Jean-Claude Trichet, President of the ECB, for the panel discussion on: “Global imbalances and financial stability” at the Eurofi G20 High Level Seminar, Paris, 18 February 2011

Comments on the article published in the Banque de France Financial Stability Review Nr. 15, February 2011.

1. To focus the Banque de France Financial Stability Review on “ Global imbalances and financial stability” is extremely timely. After some partial reduction induced by the crisis global imbalances have started to widen again. This poses challenges for international monetary cooperation and exchange rate relationships. Moreover, a further surge in private capital flows to emerging economies could lead to an acceleration of reserve accumulation and to the re-emergence of the pattern of imbalances that contributed to the factors causing the crisis. The potential disorderly unravelling of these imbalances remains one of the main risks to the European and global financial systems over the medium term.

2. Therefore we need to continue thinking about how we can make the international monetary system more resilient to such imbalances and policy structures more flexible in addressing them more effectively than the case in the past. In this regard I welcome the focus of the French G20 Presidency on the process of mutual assessment of its members’ macroeconomic and structural policies. 

3. I decided to focus my remarks today on a related but more specific topic. My starting point is that we have recently entered a new era of financial stability policies, the era of macroprudential oversight. Both the European Union and the United States of America have established new bodies whose task it is to identify early emerging systemic financial risks and to consider policy measures that could mitigate them. These are the European Systemic Risk Board and the Financial Stability Oversight Council, respectively.

4. One pre-condition for matching the expectations to this new era is that the competent authorities possess a comprehensive and reliable analytical apparatus for identifying systemic risks and for assessing the effectiveness and efficiency of policies that could contain them. In my contribution today I would like to highlight four challenges that I see in advancing such an analytical apparatus and point in some directions one could go in meeting these challenges.

5. The first challenge relates to the basic functioning of financial systems. Whereas important parts of – let me say – the “DNA” of financial systems is known, the crisis has shown that there remain significant difficulties (1) in grasping the benefits and risks of some major “mutations”, namely important financial innovations and new business models, and (2) in predicting how the overall “body” reacts to specific stresses.

6. More specifically, I would argue that our analytical apparatus would still have to make progress in characterising the role of non-bank financial intermediaries. For example, the “explosion” of the industry of highly leveraged financial institutions from around 100 billion US dollars of capital under management in 1990 to 3 trillion dollars in 2007 implies that this industry is likely to generally have important systemic implications, even though the present crisis was not caused by failures of large hedge funds. Many hedge funds are highly sophisticated, but this does not imply that in the aggregate they always act in the stabilising contrarian way perceived by some observers. Recent research about the “dot.com” bubble episode by Brunnermeier and Nagel, for example, suggests that many hedge funds were heavily invested in technology stocks rather than acting as a force correcting prices towards fundamentals.

7. The second intellectual challenge for such an analytical apparatus relates to when and how a financial system migrates from tranquility to a severe crisis. It appears to me that in economics we have some way to go in understanding the triggers, speed and abruptness of such migrations. See, for example, the sudden eruption of systemic instability in August 2007 and its dramatic worsening in September 2008, as reflected in a new indicator of systemic stress developed by ECB staff displayed in chart 1 of my article. In physics, however, “phase transitions” are the subject of ample research since a long time. For example, Jean-Philippe Bouchaud from the Ecole Polytechnique here in Paris has shown how fundamental research in physics on “crackling noise” and “self-organised criticality” can be applied to transitions to financial crises. This helps, for example, to understand the initial persistence and subsequent abrupt breakdown of financial bubbles. While this type of approach closely matches a number of important empirical patterns in asset prices, for example, it does not assume very high levels of rationality of economic agents, a strong tendency towards equilibrium situations or universal efficiency of financial markets. This observation might be insightful, if we compare it with the standard models that are nowadays used in economics.

8. The third challenge concerns how financial instability interacts with the macroeconomy. In my article, in chart 2, I have illustrated how drastically major forecasting institutions missed the start of the “Great Recession”, the “free fall” of economic activity in late 2008 and in 2009. A tremendous challenge is therefore to improve macroeconomic forecasting models, so that they give us also faster and more accurate characterisations of such situations over the relevant policy making time horizon. Two directions in which we probably have to go in order to make progress are to generally integrate more realistic characterisations of financial systems in macroeconomic models and to capture the relevant nonlinearities that are so typical for the unfolding of financial crises.

9. Fourth and last, we need to make further progress in the regulation of systemic risk, as opposed to only regulating individual risks. To be fair, Basel III goes some way in this direction. It generally increases the quantity and quality of capital and liquidity buffers. And the conservation of capital against excessive distributions will be complemented with a counter-cyclical element. Moreover, major international initiatives are under way to further reduce the probability and impact of failures of systemically important financial institutions (the so-called SIFIs), including the development of effective resolution regimes and recovery planning. One area where our analytical apparatus supporting macroprudential policies could make further progress is in terms of the assessment of early regulatory responses to credit bubbles. Some moderately sized Asian emerging countries, for example, have made some interesting experiences with the tightening of loan-to-value ratios or debt-to-income limits when property markets started to boom. We would benefit from more work about the effectiveness of these instruments in large industrial countries and as fully counter-cyclical tools, which would mean to also relax them in downturns.

10. Europe has pushed ahead with establishing the European Systemic Risk Board, which started to work last month. The ESRB will monitor systemic financial risks to the EU, including the ones emerging from global imbalances. When necessary, it will issue risk warnings and/or policy recommendations. One of its strengths is the ESRB’s broad membership, bringing around the table the right institutions and expertise. In this way, all relevant issues can be tackled in a coordinated way and its recommendations carry due weight.

11. Let me now conclude. The agenda we have to meet in the era of macroprudential oversight is tall. We need to advance our analytical and policy apparatus along the four types of challenges I just discussed. This will require, inter alia, enriching the way of thinking in economics and finance. New approaches should be considered that do not necessarily rely on the notions of equilibrium and universal rationality and efficiency. Going beyond the latter concepts may benefit from inspiration of approaches used in other fields, such as notably the natural sciences. This would not only help us advancing the foundations of national and regional supervisory policies but also the foundations of global macroeconomic policy coordination.