Skip to main content

Table 1 Backtesting results based on Z

From: A note on calculating expected shortfall for discrete time stochastic volatility models

Method

1%

2.5%

5%

1%

2.5%

5%

  

SP500

  

Russell

 

SV-\({\mathcal{N}}\)

− 3.662

− 2.092

− 1.140

− 1.895

− 1.252

− 0.687

SV-\(t_2\)

0.725

0.672

0.586

1.000

0.917

0.777

SV-t cal

− 3.190

− 1.583

− 0.793

− 1.842

− 0.828

− 0.446

SV-lev

− 1.514

− 0.854

− 0.560

− 1.408

− 0.512

− 0.432

GARCH

− 2.565

− 1.039

− 0.585

− 1.936

− 0.761

− 0.365

DFGARCH

− 0.298

0.004

0.117

− 0.366

− 0.154

0.087

Hist

− 1.310

− 0.813

− 0.545

− 0.831

− 0.205

− 0.434

  

Dow Jones

  

NASDAQ

 

SV-\({\mathcal{N}}\)

− 3.692

− 2.403

− 1.229

− 3.768

− 2.078

− 1.252

SV-\(t_2\)

0.719

0.707

0.585

1.000

0.701

0.652

SV-t cal

− 3.070

− 1.891

− 1.056

− 2.756

− 1.607

− 0.889

SV-lev

− 2.268

− 1.162

− 0.608

− 1.906

− 0.990

− 0.712

GARCH

− 3.121

− 1.425

− 0.608

− 2.662

− 1.082

− 0.673

DFGARCH

− 0.367

− 0.148

0.020

0.004

− 0.039

0.071

Hist

− 1.731

− 1.009

− 0.597

− 1.188

− 0.683

− 0.568

  1. The method with the smallest absolute value is considered to be better