Conclusions

There is ample evidence that the financial system amplifies economic cycles. This crisis has illustrated that, even worse, this mechanism could be exacerbated by financial regulation. One key objective of the ongoing efforts towards reform of the international financial architecture is to reduce such pro-cyclicality, although a healthy starting point for such efforts is to assume that regulation cannot completely eliminate pro-cyclicality.

The adoption of dynamic provisions typically pursues a double objective: (i) to smooth credit growth and (ii) to allow for the creation of reserves in the good times that would serve as buffers in the bad times. The experiences so far indicate that, when the boom has a certain size, the usefulness of provisions for the first objective is very limited; their role from the viewpoint of the second objective is much more promising.

A key question in designing an anti-cyclical device is to decide whether it should be rules-based or discretionary. The problem of a credible commitment by the authorities argues in favor of rules. But this requires a very reliable calibration of the cycle “ex ante”, an assumption that is not realistic, especially taking into account that this crisis has highlighted that the usefulness of models depends inter alia on the length and quality of the data in which they are based. It seems therefore more realistic to assume that any system would require “ad hoc” adjustments and certain degree of discretion, as illustrated by the Spanish experience. This does not imply, however, that total discretion is a superior option as in the Colombian system. The Colombian authorities themselves acknowledged this by announcing recently a reform towards a more rules-based system.

One interesting comparison between the Spanish and Peruvian cases is the use of GDP vs credit as the key variable to determine the volume of provisions. In the case of EMEs, GDP would allow to accommodate financial deepening, whereas credit would need to deal explicitly with this problem. It has also the advantage, at least in some countries, of being a leading indicator of credit. On the other hand, it has the drawback of neither being a banking variable, nor one provisions have a direct impact on.

One implication of the use of GDP versus credit is that the former variable is systemic whereas the latter is institution-specific. A systemic mechanism would be coherent with the idea of having to deal with a systemic problem, but it has implications in terms of competition and equal treatment that need to be considered carefully. One may argue that, if the regulator wants a systemic variable, overall credit could also be used. But it follows that the anti-cyclical provisioning mechanism (based on credit) could be activated for the system as a whole, perhaps because only one or a few institutions are behaving more aggressively in its credit policy. If one accepts that this would be hardly acceptable, it follows that the choice of GDP could not be based exclusively on its virtues as a systemic variable.

To what extent should dynamic provisions be applied differently to industrial versus emerging countries? One key requirement for such a system is the availability of good quality data, ideally corresponding to more than one full economic cycle. This limits the possibilities in EMEs. Another important requirement for EMEs is that the system should allow for financial deepening to occur (in other words, it should be able to differentiate “ex ante” between an excessive credit boom and a legitimate financial deepening process, something that is indeed very complicated). From this point of view GDP is superior to credit.

The debate on whether to use provisions or capital/reserves to inject an anti-cyclical element in banks' behavior has received a lot of attention recently. The arguments in favor of provisions are related to their link with expected losses, whereas the arguments in favour of capital point to the link with profits distributions (since provisions, but not capital, can be used to distribute more dividends in the downturn). The jury is still out, but international consensus seems to favor the use of both mechanisms. This seems sensible given the strong forces towards pro-cyclicality that need to be counteracted, insofar as the system does not become too complex.

Finally, any solution to the problem needs to maintain the equilibrium between making regulation more anti-cyclical while at the same time reinforcing transparency of banks' accounting statements. It is important to keep in mind that this crisis has been the result of (i) pro-cyclical financial system behavior and regulation, but also of (ii) opaqueness of financial institutions, which implies that both aspects need to be addressed in the forthcoming reforms. Reinforcing anti-cyclical mechanisms at the expense of transparency is not a solution.

Download this Paper [ PDF 259.4KB| 32 pages ].

Comment(s)

There are [0] comment(s) for this entry. Post a comment.