Money demand, bank credit and real exchange rates in a small open developing economy : an econometric analysis for Malaysia
This is essentially a three-part thesis on money demand, bank credit and real exchange rates in Malaysia. Long and short run real money demand functions with money variously defined as MO, M1 and M2 have been estimated using the Johansen cointegration technique and the error correction approach respectively. While liberalisation and innovation in the Malaysian financial system have not ruled out the existence of stable long run money demand relationships as attested to by the presence of cointegrating vectors, they have rendered short run relationships unstable. This called for a reestimation of short run dynamics over more recent periods and all the revised estimates could withstand a battery of diagnostic tests akin to original full sample estimates. The estimated short run functions appear to track the direction of actual changes in the demand for money reasonably well. The second part of the thesis is basically concerned with the possible practice of equilibrium credit rationing (a la Stiglitz & Weiss, 1981 & 1983) amongst commercial banks in Malaysia and the significance of commercial bank credit vis-a-vis other monetary variables in the determination of economic activity in Malaysia. Two of the major implications of equilibrium credit rationing are the irresponsiveness of lending rates to changes in the factors determining loan demand and supply and the presence of a 'ceiling' on the lending rate. Via an application of cointegration and error correction techniques, the lending rate is found to be insensitive to determinants of loan demand while only nominally sensitive to loan supply determinants. This is corroborated by an evidence derived from an application of Sims' VAR technique that shows a lack of responsiveness of the lending rate to changes in the inter bank rate used as a proxy for the cost of financial market funds. With regard to the ceiling on the lending rate arising from equilibrium credit rationing, its effect on the volume of deposits and hence loanable funds mobilised by banks and the interest rate payable on them may depend on the interest elasticity of their flows. Two separate cases can be considered namely the case of zero elasticity and the case of non zero elasticity. In the former case, if it is against the banks' interests to impose a high lending rate owing to possible adverse selection effects, banks may suppress the deposit rate instead. In the latter case however, the higher is the interest elasticity of deposits, the greater will be the amount of loanable funds derived and the interest rate paid on them. In our empirical analysis involving the application of cointegration and error correction techniques, commercial bank deposits in Malaysia are found to have a zero elasticity in the short run. Hence the extent of excess demand arising from an any practice of equilibrium credit rationing may be relatively limited. By applying the Sims' VAR technique, commercial bank credit has been found to exert a greater influence than MI, M2 and the lending rate on the Malaysian GDP. The final part of the thesis pertains to exchange rates. In an adaptation of Dornbusch's (1976) model, it appears that any policy measure aimed at alleviating the asymmetric information problem in the domestic banking system could lead to a depreciation in the long run equilibrium exchange rate and a rise in the long run equilibrium price level. The impact effects are a weaker domestic currency and a higher output level. However the magnitude of the long run and impact effects would vary directly with the interest elasticity of money demand. The cointegration and error correction techniques have also been relied upon for estimating the long run equilibrium real effective and bilateral exchange rates of Malaysia and the short run dynamics of these rates a la Edwards (1988a, 1988b & 1989). The estimates suggest that there has been no sustained overvaluation or undervaluation of the Malaysian real exchange rates. By implication then, the question of a real exchange rate misalignment does not arise and that Malaysia's success in economic development so far has not been due to any deliberate undervaluation policy. Moreover the analysis of causal relationships amongst real exchange rate movements on one hand and exports and GDP on the other has highlighted no significant relationships existing between them. Finally, the results from modelling the short run dynamics of real effective exchange rates indicate that excess domestic credit could induce their depreciation instead of an appreciation, contrary to popular belief.