Modeling Business Cycles with MarkovSwitching Arma (Ms-Arma) Model: AnApplication on Iranian Business Cycles

In this paper, the Iran Business Cycle characteristics were investigated vianumerous univariate and multivariate Markov-switching specifications. In thiscase Markov switching model MSM-ARMA is proposed for determiningbusiness cycles. We examined the stochastic properties of the cyclical patternof the quarterly Iran real GDP between 1988 (1) – 2008 (2). The empiricalanalysis consists of mainly three parts. First, a large number of alternativespecifications were tried and few were adopted with respect to variousdiagnostic statistics. Then, all selected models were tested against their linearbenchmarks. LR test results imply strong evidence in favor of the nonlinearregime switching behavior. In line with the main objective of research,proposed model for Iran business cycle is estimated by and result of thisestimation showed that economic of Iran despite of having two periods ofrecession 1992(3) - 1992(4) and 1995(1)-1995(2), is out of recession withmoderate growth and also experienced growth with high rate in early period ofstudying. Also the possibility of resistance of recession regimes with moderateand high growth is 0.3, 0.92 and 0.5 respectively. The results show theeconomic tend to stay in moderate growth regime

Keywords:

Markov Switching Models , Business Cycles , MS-ARMA , Iran Economy