The report illustrates three possible approaches for calculating contributions on the basis of the risk profile of the DGS members.
The investigation of risk-based models for computing contributions of Deposit Guarantee Scheme (DGS) members has become a crucial goal for the European Commission (EC) as the global crisis hits the banking sector all over the world. The new Article 12 of Directive 94/19/EC on DGS1 explicitly calls for possible models for introducing risk-based contributions and requires the Commission to submit to the European Parliament and to the Council a report on the topic by the end of 2009.
The first two models are based on approaches currently applied by some of the DGS in the EU and rely on the use of accounting-based indicators to assess the risk profile of the DGS members. More precisely, 8 indicators are proposed for the first two models, covering 4 key areas (risk classes) commonly used to evaluate the financial soundness of a bank: capital adequacy, asset quality, profitability, and liquidity. All existing DGS that adjust contributions according to the riskiness of banks use accounting-based indicators.
The first model (the ‘Single Indicator Model’) uses a single accounting-based indicator from these risk classes (e.g. a capital adequacy indicator such as the Tier I capital ratio defined in the Basel II framework). Contributions are determined as a fixed percentage of a contribution base and subsequently adjusted by a risk factor specific to each DGS member. The fixed percentage is a political decision depending for instance on the overall amount of contributions the DGS would like to collect. The risk adjustment factor is a percentage used to increase the contributions for risky members and to decrease them for well-behaving banks.
The second model (herein identified as the ‘Multiple Indicators Model’) aggregates information from 4 indicators, one from each of the 4 risk classes mentioned above, to choose the adjustment coefficient. In the present study equal weight is assigned to each indicator; however, this assumption could be relaxed in the case of DGS covering a particular sector of the banking system, e.g. schemes covering cooperative or savings banks. Also, in contrast to the first model, the overall amount of contributions to be collected is set in advance and then apportioned among the DGS members depending on their risk profile.
It must be remarked that the backward-looking nature of accounting-based indicators is a common drawback of the first two models. This could be overcome by using the third model (the ‘Default Risk Model’), which is more sophisticated from a mathematical point of view, and aims to estimate the probability of default of member institutions by including forward-looking information (i.e. market price information). This estimate then serves as the basis for calculating the contribution of a bank. In practical terms, the lack of market price information for many DGS members across EU MS (i.e. for those that are not listed on a stock exchange) makes the third model very difficult to implement. Presenting this approach therefore serves to set the bases for future research rather than offering a real alternative.
Numerical experiments thus focus on the first two models and compare estimated current contributions paid by a sample of banks in certain EU Member States (MS) with estimated risk-based contributions. Although the sample of banks used for the experiments only covers a small subset of the entire banking system of the respective MS, evidence clearly points to an advantage for the second model.
The Single Indicator Model is tested using each of the 8 proposed indicators. Results show that focusing on a single indicator would ignore valuable information on the risk profile of the bank that is captured by other indicators which are not used. In fact, the impact on current contributions would be very different depending on the indicator selected. For this model, the EU average maximum decrease in contributions for a single bank would be 19.6 %, while the EU average maximum increase in contributions would be 27.8 %, which is a very wide range.
The Multiple Indicators Model overcomes these drawbacks. Since it takes into account information from different risk classes, the model better captures a bank’s overall risk. Numerical experiments are performed using two sets of four indicators to test the change in contributions arising from the adoption of this model. Results show that the variability of the impact on current contributions is significantly reduced: the EU average maximum decrease/increase in contributions are -4.1 % and +3.8 % respectively. It is also found that changing the set of indicators does not have much influence. This model thus seems the best option.
A certain number of choices had to be made to conduct the experiments for both models. These mainly cover how risk ratings are assigned to banks, and how differences in risk profiles should be reflected in terms of contributions. These choices can in practice differ for different DGS, depending for instance on the MS banking system, on the specific banking sector covered by a DGS and, more generally, on the behaviour of financial markets as a whole.
© European Commission
Key
Hover over the blue highlighted
text to view the acronym meaning
Hover
over these icons for more information
Comments:
No Comments for this Article