Table 4 Volatility forecasts evaluation.

From: Tailoring tail risk models for clean energy investments: a dual approach to long and short position forecasting

 

Full

Sample I

Sample II

Sample III

Panel A: PBW

   

GARCH-N

2.373

3.320

3.450

2.976

GARCH-ST

2.372

3.320

3.428

2.975

gjrGARCH-N

2.373

3.297

3.535

2.976

gjrGARCH-ST

2.368

3.288

3.515

2.980

iGARCH-N

2.374

3.301

3.409

2.981

iGARCH-ST

2.371

3.303

3.396

2.978

Panel B: PBD

   

GARCH-N

1.692

na

2.697

2.209

GARCH-ST

1.687

na

2.662

2.203

gjrGARCH-N

1.684

na

2.802

2.200

gjrGARCH-ST

1.680

na

2.770

2.190

iGARCH-N

1.697

na

2.661

2.210

iGARCH-ST

1.690

na

2.651

2.202

  1. The table presents the mean values of the QLIKE loss function across different sample forecast durations. The forecasted volatility of each model is evaluated against the daily squared returns, acting as a benchmark for realized volatility. The MCS test is applied to assess the precision of the forecasts at a 90% confidence level. Models highlighted in bold denote those included in the superior set due to their statistically equivalent volatility forecasts.
Back to article page