Capital adequacy assessment in Indonesia: an empirical study

Many Indonesian banks suffered problems and some even failed in the early 1990s. This provided evidence that risk-based capital adequacy regulation in Indonesia had failed to prevent banks from taking excessive risks. Such observations provide the motivation for this thesis which seeks to identify the nature of bank risks in Indonesia and also analyses the operation of risk-based capital adequacy regulation in Indonesia. To obtain a general view of risk in Indonesian banks, this thesis includes an empirical study to identify the determinants of problem banks in Indonesia using a logit fixedeffect model. The model also can be used as an "early warning" device in banking supervision. This study finds that credit risk and operational risk contributed significantly to banking problems. State banks, non-foreign exchange banks and regional development banks are shown to be also sensitive to interest rate risk. Foreign exchange rate risk is less significant for banks (by group) in Indonesia. If we examine cases individually, however, there were some bank failures which were due to excessive foreign exchange rate risk. This thesis also finds that the adoption of risk-based capital adequacy regulation in Indonesia contains some deficiencies, such as focusing only on credit risk (ignoring market risk). This study suggests that market risk should be included in capital adequacy assessment and a number of alternative models of risk assessment [exponential weighted moving average (EWMA) and generalised autoregressive heteroscedasticity (GARCH)] are analysed. The results of the empirical study show that the inclusion of foreign exchange rate risk in capital adequacy assessment results in a higher capital requirement than that resulting from the application of the BIS's standardised methodology. This study also finds that the decay factor of 0.94 suggested by J. P. Morgan (J. P. Morgan, 1995, 1996) is irrelevant for [DR (Indonesian Rupiah) exchange rate returns. Additionally, assessment of foreign exchange rate risk using GARCH suggests a lower capital charge than that applicable under the BIS's standardised methodology and EWMA. The policy implications of these findings are also considered.