Skip to main content

Table 9 The MCS-test rank order of all methods for 1-step ahead predictions on selected datasets

From: A model-free approach to do long-term volatility forecasting and its variants

 

GE-NoVaS

P-GE-NoVaS

GA-NoVaS

P-GA-NoVaS

GARCH-direct

2018-AAPL-1step

2

4

3

1

6

2019-AAPL-1step

4

1

3

2

5

2018-BAC-1step

4

2

3

1

6

2019-BAC-1step

6

1

4

2

3

2018-TSLA-1step

5

4

1

2

3

2019-TSLA-1step

4

1

5

2

3

2018-MCD-1step

4

2

3

1

5

2019-MCD-1step

3

1

4

2

6

2018-MSFT-1step

2

3

1

4

5

2019-MSFT-1step

5

2

3

1

4

2018-Djones-1step

4

3

2

1

5

2019-Djones-1step

4

1

3

2

5

2018-NASDAQ-1step

1

4

3

2

6

2019-NASDAQ-1step

5

2

3

1

4

2018-NYSE-1step

2

4

1

3

5

2019-NYSE-1step

4

1

2

3

5

2018-S &P500-1step

4

2

3

1

5

2019-S &P500-1step

4

1

3

2

5

11.2019–10.2020-IBM

3

4

2

1

6

11.2019–10.2020-KO

3

4

1

2

6

11.2019–10.2020-MCD

2

4

1

3

5

11.2019–10.2020-SBUX

3

4

2

1

6

11.2019–10.2020-CADJPY

1

4

2

3

5

11.2019–10.2020-CNYJPY

2

4

3

1

6

11.2019–10.2020-USDCNY

4

2

3

1

5

11.2019–10.2020-EURJPY

3

2

1

4

5

11.2019–10.2020-Djones

3

4

2

1

6

11.2019–10.2020-S &P500

3

4

2

1

6

ARO

3.357

2.679

2.464

1.821

5.071

CSR

0.964

1.000

1.000

1.000

0.642

BMR

0.071

0.250

0.214

0.464

0.000

  1. The value 6 in Table 9 means the corresponding model is eliminated from the MCS for the specific dataset. We take the bold font to mark the best results according to three criteria