Time-Frequency Analysis of CAPM: Application to the CAC 40

Mestre, Roman; Terraza, Michel

doi:10.26493/1854-6935.16.141-157

Cited by 2 publications

(2 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Previously, we test the stationarity of risk premia (See Mestre and Terraza 2018) and we find that all variables have no unit‐root. Results are in Appendix A.…”

Section: Market Line Ols Estimationsmentioning

confidence: 82%

Adjusted beta based on an empirical comparison of OLS ‐CAPM and the CAPM with EGARCH errors

Terraza

Mestre

2020

Int. J Fin Econ

Self Cite

View full text Add to dashboard Cite

This paper empirically investigates the differences between the systematic risk estimated by OLS and it simultaneous estimation with a GARCH errors. The systematic risk of an asset is estimated by the beta coefficient of the market line. According to the OLS hypothesis, the estimators are robust and residuals are white noise process. However various papers show the existence of statistical anomalies (stylized facts) in residuals (heteroskedasticity, autocorrelation and non‐normality) rejecting the BLUE properties of estimators. In order to considerate these anomalies to modelize the hazard in residuals regression, we use ARCH processes class that has proved it efficiency in finance. We estimate simultaneously the parameters of the market line and those of the GARCH process for the 30 perennial equities listed in CAC40 for the daily period 2005 to 2015 and we compare them each other. We select the E‐GARCH model providing the best residuals characteristics and we note significant differences with the OLS Betas particularly for equities with betas greater than 1. On this base, we estimate a linear relationship between the OLS Betas and the E‐GARCH Betas considering this break to modelize the differences between these two kinds of Betas. By this way, Investors can quickly adjust the Beta with this tool without completely reestimating them with GARCH.

show abstract

“…Previously, we test the stationarity of risk premia (See Mestre and Terraza 2018) and we find that all variables have no unit‐root. Results are in Appendix A.…”

Section: Market Line Ols Estimationsmentioning

confidence: 82%

Adjusted beta based on an empirical comparison of OLS ‐CAPM and the CAPM with EGARCH errors

Terraza

Mestre

2020

Int. J Fin Econ

Self Cite

View full text Add to dashboard Cite

show abstract

“…Their results imply that the risk-profile (of a portfolio or a stock) is frequency dependent because the beta is frequency-varying. Mestre and Terraza ( 2018 ), McNevin and Nix ( 2018 ), Shah et al ( 2018 ), and Sakemoto ( 2020 ) all confirmed that systematic risk is conditional on an investment horizon (frequencies). These studies also discovered that frequency dynamics differ across industries.…”

Section: Introductionmentioning

confidence: 83%

Stock profiling using time–frequency-varying systematic risk measure

Mestre

2023

Financ Innov

View full text Add to dashboard Cite

This study proposes a wavelets approach to estimating time–frequency-varying betas in the capital asset pricing model (CAPM) framework. The dynamic of systematic risk across time and frequency is analyzed to investigate stock risk-profile robustness. Furthermore, we emphasize the effect of an investor’s investment horizon on the robustness of portfolio characteristics. We use a daily panel of French stocks from 2012 to 2022. Results show that varying systematic risk varies in time and frequency, and that its short and long-run evolutions differ. We observe differences in short and long dynamics, indicating that a stock’s betas differently fluctuate to early announcements or signs of events. However, short-run and long-run betas exhibit similar dynamics during persistent shocks. Betas are more volatile during times of crisis, resulting in greater or lesser robustness of risk profiles. Significant differences exist in short-run and long-run risk profiles, implying a different asset allocation. We conclude that the standard CAPM assumes short-run investment. Then, investors should consider time–frequency CAPM to perform systematic risk analysis and portfolio allocation.

show abstract

Time-Frequency Analysis of CAPM: Application to the CAC 40

Cited by 2 publications

References 1 publication

Adjusted beta based on an empirical comparison of OLS ‐CAPM and the CAPM with EGARCH errors

Adjusted beta based on an empirical comparison of OLS ‐CAPM and the CAPM with EGARCH errors

Stock profiling using time–frequency-varying systematic risk measure

Contact Info

Product

Resources

About