چرخه های کسب و کار و عدم تقارن سیاست های پولی: بررسی با استفاده از مدل مارکوف سوئیچینگ

This study assesses empirically the effects of monetary policy on four ASEAN economies in different states. The idea of asymmetry is being examined by using the relatively popular technique of non-linear modeling—Hamilton's Markov regime-switching model. The findings confirmed the existence of two-regimes in all economies under study. Additionally, the null hypothesis of symmetry had been rejected in the case of the four economies and to a great extent, monetary policy was confirmed to have had larger effects during recessions. These findings, thus, may imply the important role that credit market imperfections have on a firm's investment behavior, which in turn suggests that the financial accelerator is a relevant mechanism underscoring the observed asymmetry.

It has long been proposed by theorists such as Mitchell [1] and Keynes [2], that the movement of certain macroeconomic variables across business cycles may be asymmetric. Nevertheless, most empirical and theoretical modeling in macroeconomics has incorporated explicitly linear adjustment mechanisms, which imply that responses to shocks are both continuous and symmetric. Various sorts of non-linear models have been receiving increased attention in literature focusing on the econometric time series. Linear models, with their implied symmetry, have been shown to be deficient in dealing with certain traits of some economic series. They are unable, for instance, to fully capture the high volatility of certain financial variables [3] or the asymmetrical behavior of business cycles [4], [5], [6], [7] and [8]. As mentioned in Hamilton [5], there is “accumulated evidence that departures from linearity are an important feature of many key macro series”. Hamilton introduced a two-regime Markov-switching model with endogenous structural breaks to analyze the behavior of gross domestic product (GDP) data. His results matched the turning points (between recessions and booms) as shown in the data by the National Bureau of Economic Research. Subsequently, the use of regime-switching model has proliferated in numerous areas of applied economics. Engel and Hamilton [9] used this model to explain the long swings in exchange rates. Pagan and Schwert [10] used it to model conditional stock volatility in stock returns. Diebold and Rudebusch [11] as well as Kim and Nelson [12] applied this framework in order to study dynamic factor models in business cycles. Garcia and Perron [13] and Bekdache [14] employed Markov-switching models to investigate the behavior of real interest rate. Ravn and Sola [15] studied the variables that can affect the transition probabilities in these models. Sims [16] used this structure to analyze monetary regimes. In this study, the Markov regime-switching model developed by Hamilton [5], which shares the principle that the generating process of a series is a function of some states of nature, has been utilized to analyze the asymmetric effects of monetary policy in four ASEAN economies—Indonesia, Malaysia, the Philippines and Thailand. Having validated the existence of two-regimes in these economies, the model was extended as suggested by Garcia and Schaller [17] to include a measure of monetary policy shocks in order to examine if monetary policy has the same effect during a recessionary period as it does during an expansionary period. By way of summary, the overall findings provided substantial evidence of the existence of two-regimes in all the economies over the sample period that ended in 2003:4. Most importantly, the results generated from this analysis lent credence to the asymmetric effects of monetary policy over business cycles for all the four economies. In particular, by looking at the monetary policy coefficients, the real impact on a recession period appeared greater than on an expansion period for all model specifications. Section 2 of this study is meant to offer a concise review of the Markov regime-switching approach, followed by Section 3 which discusses the methodology. Section 4 is centered on the empirical results and the relevant insights, while the closing Section 5 presents key conclusions and suggestions.

Using a non-linear framework, business cycle is interpreted as being reflected by two altering regimes which represent the expansionary and recessionary phase. In general, based on the two-state Markov regime-switching model, the null hypothesis of one regime can be rejected at significant level based on the likelihood test. Furthermore, the transition probabilities indicate that an expansion is supposed to continue for a longer period. After confirming of the existence of two-regimes, the model was extended to incorporate monetary policy variable shocks into further testing of various hypotheses. The overall results generated from this analysis provide strong evidence for asymmetry of monetary policy over business cycles of all the economies under investigation. In these economies, the null of symmetry has been rejected in all cases. Judging from the monetary policy coefficients, the real impact in a recession has a greater size for all model specifications. On the other hand, the coefficients are negative and statistically significant during expansions in some cases. As highlighted earlier, this theoretically implausible number shows that monetary shocks in expansionary periods cause an increase in inflation expectations that could create a negative impact that outweighs the positive (real balance) effect of a monetary shock. Therefore, the evidence of the effects of monetary shocks during expansion is less pronounced than during recession. Also, from Fig. 1 and Fig. 2, one may notice that the regimes that were distinguished exhibit plausible business cycle characteristics for the univariate as well as the extended models. In summary, by using the Markov regime-switching approach, the finding that the monetary policy has a larger effect during recession period than expansion period is consistent with the argument of Bernanke and Gertler [21] and [22], and Bernanke et al. [23] and [24]. For these economies, the findings may imply the financial accelerator as a relevant mechanism underscoring the observed asymmetry. Therefore, macroeconomic stability is in peril if financial systems of these economies are not managed prudently. Policies should be designed to reduce the financial sector's vulnerability to a crisis by encouraging prudent risk management practices. Prudent financial intermediary practices can be encouraged through reform of financial sector regulation and supervision of the financial sector is crucial to the health of emerging market economies and the prevention of financial crisis. Financial liberalization must also be managed carefully to guard against exposing the economy to high levels of risk that the economy is ill-prepared to absorb.