Panel EGARCH — Exponential GARCH for Panel Data
Panel Exponential Generalized Autoregressive Conditional Heteroscedasticity Model · Also known as: Panel EGARCH model, panel exponential GARCH, EGARCH for panel data, cross-sectional EGARCH
Panel EGARCH extends Nelson's (1991) Exponential GARCH model to a panel setting, allowing conditional variance to evolve asymmetrically over time for each cross-sectional unit. The log specification ensures non-negative variance without parameter constraints, and the leverage term distinguishes whether negative shocks amplify volatility more than positive ones of equal magnitude.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use Panel EGARCH when you have panel data with time-varying volatility and suspect an asymmetric response to positive versus negative shocks — for example, stock return panels, sovereign credit-spread data, or panel energy-price series. It is appropriate when individual time series are moderately long (T >= 50 per unit) yet not long enough to estimate unit-specific EGARCH reliably, so pooling across units is beneficial. Do not use it when your panel is very short in the time dimension (T < 30), as volatility recursions require sufficient history to stabilise; also avoid it when there is no a priori reason for conditional heteroscedasticity (e.g., pure cross-section data, or series that pass an ARCH-LM test).
Strengths & limitations
- Captures the leverage effect: negative shocks can increase volatility by more than positive shocks, a pattern missed by symmetric GARCH.
- Log specification guarantees non-negative conditional variance without imposing inequality constraints on parameters.
- Panel pooling improves precision when individual T is moderate, enabling asymmetry tests that would be underpowered in single-unit EGARCH.
- Flexible: can allow unit-specific intercepts while pooling slope parameters, balancing parsimony and heterogeneity.
- Compatible with fat-tailed innovation distributions (Student-t, GED) for financial data.
- Requires a reasonably long time dimension (T >= 50) per unit; very short panels produce unreliable variance recursions.
- Log-variance recursion complicates multi-step forecasting because the forecast of log-variance does not equal the log of the variance forecast.
- Fully heterogeneous specifications consume many parameters and can overfit when N is large and T is small.
- Maximum likelihood estimation is sensitive to starting values; convergence is not guaranteed and local optima are possible.
Frequently asked
How does Panel EGARCH differ from Panel GARCH?
Panel GARCH models the conditional variance directly, requiring non-negativity constraints on parameters. Panel EGARCH models the log of conditional variance, eliminating those constraints, and adds a signed shock term (gamma) to capture the leverage effect — the tendency for negative shocks to raise volatility more than positive ones.
What does a negative and significant gamma coefficient mean?
A significant negative gamma indicates a leverage effect: negative innovations increase log-conditional variance by more than positive innovations of the same magnitude. This is typical in equity markets where bad news amplifies uncertainty more than good news.
Can I pool all parameters across units?
You can impose full homogeneity for parsimony, but it is risky if units operate in different economic environments. A common compromise is to pool the persistence (beta) and leverage (gamma) parameters but allow unit-specific intercepts (omega_i), testing homogeneity restrictions with a Wald or likelihood-ratio test.
How do I check whether EGARCH is correctly specified?
Compute standardised residuals z_it = epsilon_it / sqrt(h_it). They should be close to i.i.d. Apply Ljung-Box tests to both z_it and z_it-squared; significant autocorrelation in the squares indicates remaining ARCH effects. Also check whether the chosen innovation distribution (normal vs. Student-t) fits the tail behaviour.
Is Panel EGARCH appropriate when the panel has cross-sectional dependence?
The conditional variance recursion itself is unit-specific, but cross-sectional dependence in the innovations can inflate standard errors. Use cluster-robust standard errors or a DCC (Dynamic Conditional Correlation) model when contemporaneous correlation across units is a primary concern.
Sources
- Nelson, D. B. (1991). Conditional heteroskedasticity in asset returns: A new approach. Econometrica, 59(2), 347–370. DOI: 10.2307/2938260 ↗
- Tsay, R. S. (2010). Analysis of Financial Time Series (3rd ed.). Wiley. ISBN: 978-0470414354
How to cite this page
ScholarGate. (2026, June 3). Panel Exponential Generalized Autoregressive Conditional Heteroscedasticity Model. ScholarGate. https://scholargate.app/en/econometrics/panel-egarch
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- EGARCH modelEconometrics↔ compare
- Panel DCC-GARCHEconometrics↔ compare
- Panel GARCH modelEconometrics↔ compare
- Panel TGARCHEconometrics↔ compare