5.2 SVAR 

  1. Model 1 
The first model estimated the SVAR model with five variables, including global uncertainty(GU) and domestic uncertainty(KU), KOSPI index, GDP(KGDP) and consumer price index. The V2 model last placed the estimated sequence of domestic uncertainty variables (KU). Assuming that global uncertainty is the most exogenous of these, while domestic uncertainty is unclear, so it is divided into two models (V1,V2) to compare impact response. 
\[v_1:y_t=[gu_t,ku_t,kospi_t,kgdp_t,kcpi_t]\]
\[v_2 : y_t =[gu_t,kospi_t,kgdp_t,kcpi_t,ku_t]\]
As seen in [Fig2], GDP and KOSPI declined significantly due to shock reactions to global uncertainties, and have since rebounded and recovered.  This is consistent with the findings of \cite{bloom2009impact,Jeon2017}. KCPI, on the other hand, reacted rising up as uncertainty shocks were given. It would indicate that foreign investors will leave the country if uncertainties arise and the exchange rate will tend to rise, apparently due to a mechanism in which prices of imports will rise and consumer prices will rise in the short term. This will be confirmed by adding exchange rate variables in Model 2. Meanwhile, GU shock does not significantly affect KU, meaning that our GU estimates and KU estimates are orthogonally  well separated. Responses to domestic uncertainties(KU) also show a similar path to global uncertainties(GU). However, it shows a decreasing response (GDP, KOSPI) over time, and recovery is tend to be delayed. The consumer price index showed a more sensitive response than the impact of global uncertainties, showing a stabilizing trend after the fifth period.