Malaysian Bank Capital and Risk Profiles: Causality Tests

Abstract

The structural relationships among bank capital and risk taking are
empirically examined by utilising unit root tests and Granger
causality tests using the time series data. The Granger causality test
results are not very robust with respect to different types of banking
institutions, risk variables (NPL and RWA) and time period. With
merchant banks and finance companies aggregate data, there
appears to be an absence of a Granger causality effect in the
Malaysian banking sector. The evidence for Granger causality
running from capital to risk or risk to capital appears to be
statistically significant when the test is performed using the
commercial bank aggregate data. Our results also show that there is
no strong indication that using non-performing loans implies
likelihood of finding a significant relationship. Finally, the evidence
of lead-lag relationship between capital and risk is generally weak
before the1997-98 banking crisis.
Keywords: Bank Risk, Capital Ratios, Risk-based Capital, Risk Index,
Capital and Earnings.
JEL classification: G21, G28