Financial Vulnerability Indicator in Malaysia: Construct and Forecasting Ability

This study attempts to investigate the fluctuations of the Malaysian financial market through predicting the early signals of financial market vulnerability using the Financial Vulnerability Indicator (FVI) constructed through the signalling approach and the dynamic approximate factor model (DAFM), using monthly data spanning from 1997 to 2017. The indicators were constructed from macroeconomic, financial and property market data, which comprises of imports, the real GDP, exports, M2, total reserves, the share price, loans and advances, the consumer price index, housing approvals and the real effective exchange rate. The empirical results from the FVI based on the signals approach proposed by Kaminsky and Reinhart (1996) and Kaminsky and Reinhart (1999) suggested that weak exports, a decline in housing approvals, an overvaluation of the exchange rate and an economic slowdown are the common triggers of the occurrence of a financial crisis in Malaysia. For the FVI constructed through the DAFM, following Banbura and Modugno (2010, 2014) and van Roye (2014), the empirical findings indicated that imports, the real GDP and exports contained good predictive ability in gauging financial vulnerabilities. Both indicators highlighted the importance of internal and external macroeconomic environments in defining the financial vulnerability of an economy. In the evaluation of forecasting performance, the FVI constructed through the DAFM contained better predictive power, reflected by a lower noise-to-signal ratio. The feedback and amplification of the effects of financial vulnerabilities on the real sector were explored through the MSBVAR, methodology proposed by Sims et al. (2008). The outcome revealed that financial vulnerability shock brought about significant negative effects on industrial production and inflation while producing significant positive effects on the short-term interest rate, KLIBOR, in a high-vulnerability regime. This highlighted the prominence of the nonlinearities during a high vulnerability period and this is consistent with studies on other economies. To conclude, this study highlighted the usefulness of composite indicators as an early warning mechanism to gauge vulnerabilities in the financial system from a macroeconomic perspective through a non-parametric and model-based indicator construction methodology while recognising the feedback effects of the real sector during periods of high vulnerability. Hence, policymakers and stakeholders in the financial market may improvise financial vulnerability indicators as early signalling mechanisms to foretell the developments in the financial market.