Back to Article

Nigeria Exchange Rate Volatility: A Comparative Study of Recurrent Neural Network LSTM and Exponential Generalized Autoregressive Conditional Heteroskedasticity Models

Journal of Artificial Intelligence and Big Data | Vol 4, Issue 2

Table 3. EGARCH (1,1) models Fitted for Euro, Pound Sterling and US Dollars.

ParticularsEuroPound SterlingUS Dollars
VariableCoef. (p-value)Coef. (p-value)Coef. (p-value)
Constant4.634 (0.000)11.341 (0.000)7.523 (0.000)
1 st Lag Order-0.765 (0.000)-2.112 (0.000)3.244 (0.000)
2nd Lag Order-0.513 (0.000)-1.763 (0.000)-2.332 (0.000)
3rd Lag Order-0.582 (0.000)-0.352 (0.000)3.774 (0.000)
4th Lag Order-0.342 (0.000)-0.996 (0.000)3.798 (0.000)
Performance Criteria
 R-squared 0.8121910.820555 0.887292
Loglikelihood-1387.898-1313.194-1230.568
Durbin-Watson2.0242202.0336072.018676
AIC10.6658810.093449.460294
SIC10.7205110.148079.514922

Note: AIC is the Akaike info criterion and SC is the Schwarz criterion