Forecasting VaR and ES by using deep quantile regression, GANs-based scenario generation, and heterogeneous market hypothesis

Financial Innovation

Table 11 Comparision of models using the coverage rate and correlation coefficients

	FTSE100		N225		SPX500		DAX
	CR	CC	CR	CC	CR	CC	CR	CC
\(\tau = 0.05\)
GD	0.9762	0.1846	0.9881	0.2427	0.9786	0.2336	0.9762	0.1531
KDE	0.9857	0.1744	0.9929	0.1976	0.9881	0.2408	0.9881	0.1478
MCMC	0.9667	0.2268	0.9691	0.2515	0.9596	0.3002	0.9715	0.1751
LSGAN	1.0000	0.2651	1.0000	0.3147	1.0000	0.3140	1.0000	0.2418
\(\tau = 0.01\)
GD	0.9549	0.3328	0.9644	0.3175	0.8955	0.3190	0.9620	0.3077
KDE	0.9881	0.3110	0.9810	0.3407	0.9596	0.3170	0.9834	0.3072
MCMC	0.9335	0.3845	0.9359	0.4695	0.8670	0.4659	0.9216	0.3937
LSGAN	1.0000	0.4339	1.0000	0.6691	1.0000	0.6314	1.0000	0.6006