Distributional characteristics of interday stock returns and their asymmetric conditional volatility: Firm-level evidence

• Published in: Physica A Vol. 508; pp. 280 - 288
• Main Authors: Balaban, Ercan, Ozgen, Tolga, Girgin, Mehmet Sencer
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 15.10.2018
• Subjects: Asymmetric time varying volatility; Day of the week effects; Interday effects; Risk–return relationship; Statistical distributions of stock returns and volatility; Systematic risk; Asymmetric time varying volatility; Day of the week effects; Risk–return relationship; Systematic risk; Interday effects; Statistical distributions of stock returns and volatility
• ISSN: 0378-4371; 1873-2119
• DOI: 10.1016/j.physa.2018.05.019

This paper is a pioneering effort to jointly analyze the characteristics of interday distributions of stock returns and their asymmetric time varying volatility using firm-level data in local currency from an emerging stock market, namely, the Bourse Istanbul, for the period January 1996 to December 2015. Using a modified Threshold Generalized Autoregressive Conditional Heteroscedasticity-in-Mean [TGARCH(1,1)-M] model; these distributional characteristics statistically assess in a unique framework (i) the weak-form informational efficiency based on the stylized facts of day of the week effects on stock returns and their conditional volatility; (ii) volatility persistence and asymmetry in conditional volatility; and (iii) the conditional total risk–return relationship, and the impact of systematic risk as an asset pricing factor. It is found that at firm level there are statistically significant positive or negative day of the week effects on either stock returns or their conditional volatility, or both. However, for a sample of 120 firms, a full and cross-sectional analysis of the interday distributions does not lead to a systematic pattern of return differences across days of the week. The average volatility is found to be highest on Mondays and the lowest on Wednesdays. It is reported that – as a proxy of total risk in a mean–variance framework – the estimated conditional standard deviation does not have a significant impact on stock returns for the great majority of the sample firms. With reference to the total risk–return relationship and the asymmetry in volatility, there are no significant differences between the industrial and financial sector companies. It is reported that the systematic risk is always priced; and the results are highly significant with a high explanatory power. Volatility is decidedly persistent for all firms investigated; while, a significant asymmetry in the conditional volatility cannot be reported for most of the firms. Contributing to the existing literature as a first time analysis of firm-level distributional characteristics of interday stock returns and their asymmetric conditional volatility with an additional proper risk-impact investigation, the empirical results are of importance primarily for asset pricing and risk management research and practice.