Abstract
Hamilton's Markov-switching model [5] was extended to the simultaneous equations model. A framework for an instrumental variable interpretation of full information maximum likelihood (FIML) by Hausman [4] can be used to deal with the problem of simultaneous equations based on the Hamilton filter [5]. A comparison of the proposed FIML Markov-switching model with the LIML Markov-switching models [1,2,3] revealed the LIML Markov-switching models to be a special case of the proposed FIML Markov-switching model, where all but the first equation were just identified. Moreover, the proposed Markov-switching model is a general form in simultaneous equations and covers a broad class of models that could not be handled previously. Excess sensitivity of marginal propensity to consume with big shocks, such as housing bubble bursts in 2008, can be determined by applying the proposed model to Campbell and Mankiw's consumption function [6], and allowing for the possibility of structural breaks in the sensitivity of consumption growth to income growth.
본 논문에서 Hamilton의 마코프-스위칭 모형을 연립방정식으로 확장한 FIML 마코프-스위칭 모형을 제시해 보았다. 본 논문의 FIML 마코프-스위칭 모형을 LIML 마코프-스위칭 모형 등과 비교하면 LIML 마코프-스위칭 모형은 FIML 마코프-스위칭 모형의 특별한 경우이며 FIML 마코프-스위칭 모형은 연립방정식으로 확장된 일반화된 모형 형태를 띄게 된다. 본 논문의 FIML 마코프-스위칭 모형을 Campbell and Mankiw 소비함수에 적용해 본 결과, 2008년 부동산 거품 붕괴와 같은 경제충격 시기의 한계소비성향은 매우 민감도가 높아진다는 것을 알 수 있다.
