The causal link between money and output has been widely debated in the literature over the past decades; however, the results are inconclusive. The review of the literature shows that results of causality tests appear to be sensitive to the sample period employed. To overcome this problem, we use a VAR model with time varying parameters, and these variations are governed by a Markov chain. In this paper, we use quarterly data from 1376 to 1391 and a Markov-Switching VAR model with two regimes to examine the causality between the output and money in Iran. The results show that money Granger-causes output only in the first regime, which covers the period from 1384:3 to 1390:3. Therefore, in the Iranian economy money is not always neutral. In addition, the results show that the output is a Granger-cause of the money in both regimes; however, the link is negative in the regime 1. In other words, it indicates that money and output were changing in opposite directions in the regime 1.
)
