Artikel - How to Use the Instruments of Macropudential Policy (27 augustus 2014)

27 augustus 2014 09:12

DSF POLICY BRIEFS No. 33/ July 2014

How to Use the Instruments of Macropudential Policy Lex Hoogduin, University of Amsterdam and Duisenberg School of Finance


One of the main lessons from the financial crisis of 2007/2008 was the need for the development of macroprudential policy. That would be a policy focusing on the financial system as a whole, with its own objective(s), instruments, accountability requirements and analytical framework.

This note provides some reflections on the development of the instruments for macroprudential policy. It is, however, impossible to discuss macroprudential instruments in isolation. Therefore, I will discuss the instruments against the background of the initial conditions in which they will be introduced, the objective(s) of macroprudential policy, the theory of crises/instability underlying that policy and the relation with other policies.

Initial conditions

New Policy
It must be acknowledged that macroprudential policy is new. Introduction of a new policy and using new policy instruments or existing policy instruments for other policy objectives than used to be the case, may always have unintended consequences, for the recovery and growth in particular. This calls for a gradual introduction of instruments and for learning by doing and from experience. That means not using a great number of instruments at the same time and initially.

It also calls for initial modesty in objectives of macroprudential policies. Usually two possible objectives are distinguished. The first is to maintain or increase the resilience of the financial system. The other is stabilising the financial cycle. The first objective is more modest, because no data or diagnosis on the state of the cycle is required. Therefore, it would be wise to focus on maintaining and where needed increasing the resilience of the financial system as the objective of macroprudential policy. Instruments should thus be targeted on that objective.

The macroprudential authority needs to be flexible and indeed be able to learn from experience. Therefore, it should be able to introduce new instruments rapidly without having to go through a lengthy process, also involving politics.

To enable flexibility a basis in law could be created for the macroprudential authority to develop and use specific instruments from four categories of instruments at short notice:

  1. Related to balance sheets of financial institutions: liquidity, leverage, etc.

  2. Related to financial transactions: haircuts, margin requirements, etc.

  3. Related to commercial and/or residential real estate: loan to value ratios, debt to income ratios, etc.

4. Structural policies: size of financial institutions, restrictions with respect to certain (types) of activities like proprietary trading, etc.

The perimeter of macroprudential policy should be wider than banks. If not, over time the shadow banking system will become larger and its resilience cannot be ensured. This may undermine the resilience of the financial system at large.

Low buffers
Another main lesson from the financial crisis is that capital requirements had fallen to too low levels and that there was a need to introduce a regime for liquidity requirements. Basel III sets higher capital requirements and in addition to that systemically important financial institutions (SIFI’s)have to hold even more capital. Basel III also introduces the Liquidity Coverage Ratio (LCR) and Net Stable Funding Ratio (NSFR) with respect to liquidity.

Implementing these policies will increase the resilience of the worldwide financial system. The Basel III countercyclical buffer does not only contribute to increasing the resilience of the financial system, but also to stabilising the financial cycle.

For the coming years, in the context of a gradual introduction of macroprudential policies no additional macroprudential instruments than foreseen in Basle III and the policies related to SIFI’s need to be applied to banks. I have sympathy for the argument that after full implementation of these measures capital buffers in particular are still too low. That would argue for the presumption that Basel IV would further increase capital requirements.

What is urgent is the development of policies for the resilience of the shadow banking system. Capital requirements fully consistent with those for banks should be introduced.

Real estate is very often at the heart of financial crises. Maintaining or increasing resilience of the real estate sector should therefore be a key element in macroprudential policies going forward. That would require introducing loan to value ratios well below 100% where they do not yet exist in residential real estate markets and implementing similar measures in commercial real estate markets. This should be done in steps. It may also require structural changes elsewhere, e.g. in pension systems, to avoid that these policies would dampen aggregate demand more than necessary.

Introducing macroprudential instruments related to insurance and pension funds activities has no priority. I would also be cautious to introduce additional structural policies at this stage. The same applies to macroprudential instruments related to financial transactions.

With respect to financial market infrastructure, it seems wise to now first implement the new regulation and not introduce new instruments on top of those for the next four or five years, but rather fine tune the existing regulation2.

Underlying theory of financial cycles and –fragility

Underlying forces
Part of any macroprudential policy framework must be a theory of financial crises and financial fragility. Fragility is defined here as the negative of resilience, i.e. the potential for the financial system to become unstable or fall into crisis.

It goes far beyond the scope of this note to comprehensively discuss theories of financial cycles and – fragility. I just list what in my view can be identified as the key driving forces of financial cycles and – fragility:

1. The inevitable existence of fundamental uncertainty3. An element of the human condition is that we cannot know now the development of our own knowledge over time. Rationality alone cannot guide decisions and actions as a consequence.

2. Human nature has slowly evolved over time adapting to the (changing) environment and to dealing with uncertainty in an evolutionary process. This has led to overconfidence, herd behavior, emotions, etc. to play a role in decision making and in driving (in)action4.

3. Financial systems and economies are complex systems5, i.e. they consist of a great number of interacting subjects and institutions responding also to their environment. All necessary data about the state of the financial system cannot be centrally collected. Data, information and knowledge are highly dispersed throughout the financial system and are created, deleted and reshaped continuously6.

In the context of the topic of this note this has the following consequences. Risk assessments both by economic agents and supervisors/central banks are fallible methods to deal with the uncertain future. There is no solid ground in logic for quantitative risk management. Probability theory and risk management are useful heuristics. No more, no less. Risk assessments are always subjective. The same holds for prices of financial assets. Market to market measurement does not provide an objective method for measuring financial soundness. Such a method does not exist.

The pattern of financial cycles and –fragility, is that prices of financial assets increase and perceived risks and required compensation thereof decrease in the upswing. And then there is a sudden turning point and that triggers a sharp revision of valuations, sometimes with a crisis as a result. Cyclicality is an inherent characteristic of the economic/financial process.

2 That does not rule out that for example the scope for mandatory clearing is extended if need be.
3 See Hoogduin (1991), Keynes (1921, 1937) and Knight (1921).
4 See Keynes (1936), Chapter XII for a vivid description, preparing the ground for what is nowadays called behavioural economics.
5 See Hayek (1967) and Mitchell (2009)
6 See Hayek (1945)

Systemic risk
This implies that the concept of systemic risk is troublesome. A quantifiable measure of the probability of crisis cannot be grounded in logic. Market indicators of risk are pro-cyclical and are lagging indicators. The upswing exists because many subjects are optimistic and/or are prepared to receive a low compensation for taking risk. Therefore, I would not have this concept play any role in deciding on how to employ macroprudential instruments.

Dampening of the financial cycle
Ambitious objectives for macroprudential policy are problematic. The data needed for making a diagnosis of the state of the financial system can never be collected. The assessment of how likely it is that a development will become unsustainable, is very difficult to make and cannot be simply be derived from data.

The nature of financial cycles also makes clear that macroprudential measures in the upswing will always be unpopular, also because they can probably be only effective if taken at a moment when there are no clear signs of an unsustainable development building up. This requires the macroprudential authority to be independent. But this can only work in a democracy with a clear and limited mandate.

This reinforces the earlier conclusion that the focus of macroprudential policy and –instruments should be on increasing/maintaining the resilience of the financial system. If nonetheless instruments will be chosen for dampening the cycle, there is much in favour of making them as much as possible rules based.

Risk weighted capital measures have a shaky foundation. The risks cannot be measured reliably. Ultimately capital should not be seen as insurance against risk, but as a buffer against the unforeseeable, against fundamental uncertainty. But uncertainty cannot be measured.

This would argue for making unweighted capital measures the basis for macroprudential policy and – instruments. Leverage should take centre stage.

Liquidity is endogenous. More thought should be given to what that means for the definition of a liquidity measure in a macroprudential instrument. A narrow or broad range of assets? And how to determine which assets qualify? It also touches upon the importance of diversity for creating resilience. More diversity in the financial system is likely to lead to preserving more liquidity in stressed times than when there is little diversity.

The downswing
Financial cycles are a fact of life. This means that there will always be upswings, turning points and downswings. Macroprudential instruments should not be designed with primarily the upswing in mind. Promoting the resilience of the financial system does not only require to determine how high minimal buffers should be. There should also be a policy with respect to the use of buffers. Buffers that cannot be used, are useless in absorbing losses and providing resilience.

Developing a policy with respect to using capital and liquidity buffers should have high priority. Such a policy should be made consistent with the policies with respect to recovery and resolution and with the lender of last resort role of the central bank.

Relation with other policies and their instruments
A crucial area that I can only mention here is that the deployment of macroprudential policy instruments should be well coordinated with the use of other policy instruments. Economic policy instruments have almost always an impact on more than one policy objective.

Therefore, in defining the objective of economic policies it should always be mentioned that they should support other policy objectives to the extent that it does not hinder achievement of their own primary objectives. In this context one could question if announcing that interest rates will remain low for a long period of time or that monetary policy tightening would be gradual, is sufficiently supportive of maintaining a resilient financial system.

It should be avoided that macroprudential policy instruments will be used to compensate for suboptimal other policies, like fiscal and monetary policies. This would for example be the case if capital controls or lower loan to value ratios would be introduced as macroprudential measures in an environment where monetary policy would be too expansionary.

Hayek, F.A. (1945), “The Use of Knowledge in Society”, reprinted in Individualism and Economic Order, 
London, 1949, pp. 77-91.

Hayek, F.A. (1967), “The Theory of Complex Phenomena”, reprinted in Studies, in Philosophy, Politics and Economics, 
London, pp. 22-42.

Hoogduin, Lex (1991), Some Aspects of Uncertainty and the Theory of a Monetary Economy, 

Keynes, J.M. (1921), A Treatise on Probability, The Collected Writings of John Maynard Keynes, Vol. VIII, 

Keynes, J.M. (1936), The General Theory of Employment, Interest and Money, The Collected Writings of John Maynard Keynes, Vol. VII, 

Keynes , J.M. (1937), “The General Theory of Employment”, reprinted in The Collected Writings of John Maynard Keynes, Vol. XIV, 
London, pp. 109-124.

Knight, F.H. (1921), Risk, Uncertainty and Profit, Boston. Mitchell, Melanie (2009), Complexity. A Guided Tour,