standard deviations of residuals for a Time varying structural VAR
Hetero_TVP_VAR_stdresid(res_TVP,resid,pvalue,plt)
* res_TVP = a 'heteroskedastic tvp var' result tlist produced by a call to Hetero_TVP_VAR
* resid = a vector of integers, collecting the indexes of the residuals whose standard deviations the user wants to graph
* pvalue = a real between 0 and 1, the p-value of the confidence band (1 minus the probability that the standard deviation lies in the confidence band)
* nothing: results are graphed on graphic windows
stacksize('max') load(GROCERDIR+'\data\primiceri.dat') mkdir(GROCERDIR+'\temp') nlag=2 bounds('1953q3','1963q2') prior=tvpvar_prior0(nlag,4,4,0.01,0.01,0.1,['Inflation';'Unemployment';'rate_3m']) bounds('1963q3','2001q3') res=Hetero_TVP_VAR(nlag,['Inflation';'Unemployment';'rate_3m'],10000,2000,prior,path,2000) Hetero_TVP_VAR_stdresid(res,1:3,0.32) // After having estimated Primiceri (2005) tvp-VAR, the standard deviations of the residuals // are calculated and graphed for the residuals numbered 1 to 3 (that is here all) and a 68% // confidence band | ![]() | ![]() |