Skip to main content

Table 8 Summary of the backtesting results for each probability level for the stock indices. Number of test rejections for the four indexes

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

Model

τ = 0.01

τ = 0.05

UC

IND

CC

DQ

Mean (M-Loss)

UC

IND

CC

DQ

Mean (M-Loss)

Hist

1

1

1

4

0.4504

1

0

1

1

0.8882

Normal

2

1

1

4

0.6490

1

0

1

2

0.9201

CAViaR-SAV

0

0

0

4

0.3144

0

0

0

1

0.7912

CAViaR-AS

0

0

0

3

0.2044

0

0

0

0

0.6677

CAViaR-IGARCH

0

0

0

3

0.2678

0

0

0

1

0.7088

CAViaR-Adaptive

0

0

0

4

0.2904

0

0

0

1

0.9012

QRNN

1

0

1

2

0.1990

2

0

2

3

0.7970

LASSO-QR

3

0

3

4

0.6661

1

0

0

2

0.8867

QR-RF

2

0

0

2

0.1949

0

0

0

0

0.4057

QR-GBDT

3

0

2

4

0.2164

0

0

0

1

0.4592

QR-SVM

0

1

0

4

0.3907

1

0

1

1

0.8367

QR-CNN

0

0

0

3

0.3032

0

0

0

2

0.6485

QR-LSTM

0

0

0

4

0.2891

2

0

2

2

0.7939

QR-GRU

0

0

0

4

0.2655

1

0

0

1

0.6533

Proposed Model

0

0

0

1

0.1379

0

0

0

0

0.3370

  1. Smaller values are preferred