time-varying parameter vector autoregression model (tvp-var)

The impact of financial and economic shocks and uncertainty is not always limited to the target market and may spread to other markets as well. Empirical research results, such as those by Jurado et al. (2015) and Gabor and Gabota (2020), indicate that the contagion of cross-sectoral uncertainty and the significance of these uncertainties are not constant over time and may change. Traditional time series regression models assume that a relationship with fixed coefficients can be applied across different time periods. The misleading results of this unrealistic assumption have led to the development of dynamic models that better reflect the realities of the external world. The state-space approach is a modeling method for dynamic systems that predicts and analyzes system behavior under these modeling conditions. One of the applications of this approach is to account for structural instability in parameters and to allow coefficients to vary over time. Models of this type are known as time-varying parameter (TVP) models. This research aims to study the reaction of the financial, housing, and macroeconomic sectors in Iran to each other's shocks, with a focus on the effects of uncertainty contagion.
 
The present study is applied in terms of purpose and correlational analysis in terms of nature and method. It is post-event and utilizes past information. In this study, a library method was used to collect theoretical sources, while an archival method was employed to gather the data needed for hypothesis testing. To examine changes in cross-sectoral uncertainty contagion, the time-varying parameter vector autoregression model (TVP-VAR) is used with monthly data from January 2008 to December 2020. In this context, uncertainty indicators are calculated using GARCH models and then tested using the TVP-VAR approach, along with an analysis of variance of the generalized prediction error of total dynamic connectedness, as well as the directional dynamic connectedness of the indicator pairs.
 
The research results indicate that the primary source of uncertainty is the macroeconomic sector, which acts as the main source and transmitter of uncertainty to the other financial and housing sectors. Additionally, the housing sector is a net recipient of uncertainty from the other two sectors. The findings suggest that the contagion of uncertainty between the financial and housing sectors is bidirectional and conditionally dynamic, while the contagion of uncertainty from the macroeconomic sector to the financial and housing sectors is unidirectional.
 
According to the results, the contagion of cross-sectoral uncertainty and the significance of these uncertainties are not constant and change over time. Therefore, identifying the different channels of contagion between markets and pinpointing the source of contagion can help in selecting policies that reduce vulnerability and enhance the performance of asset portfolio risk management.