Institut numerique

Excess volatility puzzle of stock prices: Explanation by behavioral finance across overconfidence and herd behavior. « Empirical validation on the Tunisian Stock Exchange Market »

Abstract

The emphasis on behavioral finance in recent decades has motivated us to dissert this memoir which aims to
allow readers, not only, to know different behavioral biases, but also, to understand their contributions to
explain the excess volatility of stock prices.
In order to achieve this purpose, we began by challenging the efficient market hypothesis and its inability to
elucidate observed anomaly in financial markets. This was the reason bellow the appearance of a new fieldbehavioral
finance, which assumes to bring a new light on the actual functioning of markets and its deficiencies,
through a variety of behavioral biases. Our interest in this approach is focused particularly on the clarification of
excess volatility across psychological aspects of investors, that is, the bias of overconfidence and the bias of herding
behavior.
Once excess volatility is detected on the Tunisian Stock Exchange Market using Variance Bounds Test of
Shiller (1981), a literature review will be conducted including the presentation of the two biases mentioned above
and their contributions to explain excess volatility puzzle in stock prices. After this theoretical body, two empirical
studies will be developed: the first one intended to detect overconfidence and to identify whether excess volatility
of observed securities on the Tunisian stock market result from investors‘s overconfidence, is inspired by the study
of Chuang and Lee (2006) based on an asymmetric model EGARCH (1, 1). The second one intended to detect herd
behavior and to identify whether excess volatility of observed securities on the Tunisian stock market result from
investors’s herd behavior, is inspired by the study of Tan et al. (2008) based on the calculation of Cross-Sectional
Absolute Deviation (CSAD: measure of detection of herd behavior) according to the methodology proposed by
Chang, Cheng et Khorana (2000) then on the regression of the asymmetric effects of herd behavior through market
return, trading volume and volatility.

Key words: Efficient market hypothesis, Anomaly, Excess volatility, Behavioral finance, Variance Bounds Test,
Overconfidence, Herd behavior, EGARCH.

Retour au menu : L’énigme de volatilité excessive des cours boursiers : explication par la finance comportementale à travers l’excès de confiance et le comportement grégaire. Validation empirique sur la BVMT.