Application of the Threshold Vector Autoregression Model (TVAR) in Nonlinear Analysis of Exchange Rate pass-through on Inflation in Iran

Considering nonlinearities in the exchange rate pass-through to domestic prices, this study analysis the Exchange Rate pass-through on Inflation in Iran by estimating a Threshold Vector auto-regression (TVAR) model. We estimate a Threshold Vector Auto-regression (TVAR) model on quarterly data over the period 1990Q2-2016Q3. The nonlinearity test for a TVAR model against a linear VAR model suggests the presence of two regimes with one threshold value of inflation. The threshold value of inflation is estimated endogenously. The quarterly rate inflation of 3.9 % select as a threshold level between two regimes. We find that domestic prices in Iran response strongly to a positive exchange rate shock in two regime of inflation. The rate of inflation below the threshold level constitutes a low inflation regime, and above the threshold level it constitutes a high inflation regime. The response of inflation is statistically significant in the two regimes and this result confirms the Taylor’s theory. Also the calculation of the ERPT coefficient confirming this result and show that exchange rate pass-through is incomplete.