Empirical investigation on loan loss provisions: a cross country analysis / Hasni Abdullah

The present study examines the determinants of loan loss provisions with the presence of earnings management and capital management by commercial banks in the emerging and developed markets during the period of 2004 to 2012. Besides the main objective, there are external factors as well as the internal factors investigated in determining the loan loss provisions. This study focuses on the banking sectors that specify the commercial banks for the period of 2004 to 2012 in selected emerging and developed markets. In order to achieve the objectives of this study, Malaysia and Thailand are chosen to represent the emerging markets, while the developed markets are represented by Hong Kong and Japan. The empirical findings based on Pooled Ordinary Least Square (Pooled OLS) with Newey-West Standard Error for emerging markets and Fixed Effects (FE) with heteroscedasticity and autocorrelation for developed markets recommend that earnings management, gross domestic product and certain bank internal factors have a statistically significant impact on loan loss provisions. This study found that the proxy of earnings management, earnings before tax and provisions is positively related to loan loss provisions in both economies. Surprisingly, the variable of capital management is found to be insignificant in emerging market economies and positively associated with LLP in developed markets. Based on the empirical findings of GDP implying that during the period of study, the economic expansion does influence the decision of loan loss provision by following the normal pattern of business cycle for both markets and contributes to the procyclicality behaviour of loan loss provisions. With respect to the effect of bank specific factors on the loan loss provisions of emerging and developed markets, it clearly indicates from the overall perspective of the results, that the obvious factors that have really contributed to loan loss provisioning practice of both markets are only the non-performing loan and the return on assets. Meanwhile, the robustness check using Generalized Method of Moments (GMM) has basically generated the same results as Fixed Effects with heteroscedasticity and autocorrelation for developed markets.