On the Impact of Long Memory on Market Risk: Pre- and Post-Crisis Evidences – Long Memory and Stock Market Risk

On the Impact of Long Memory on Market Risk: Pre- and Post-Crisis Evidences – Long Memory and Stock Market Risk

Salim Ben Sassi (High Institute of Management of Tunis, Tunisia) and Azza Bejaoui (High Institute of Management of Tunis, Tunisia)
DOI: 10.4018/978-1-5225-3767-0.ch003
OnDemand PDF Download:
No Current Special Offers


This chapter investigates the influence of the long memory behavior in returns and volatility on the market risk for four emerging stock markets during the pre- and post-crisis periods. In this respect, the authors consider four major political events (Tunisian revolution, Egyptian revolution, assassination of Prime Minister Rafik El Hariri, and a series of suicide bombings in Morocco). Using the modified R/S test and GPH test, they show the long memory property in returns and volatility over the two sub-periods. To explore the dual long memory property, the authors apply the joint ARFIMA–FIGARCH specification on the returns and volatility of the four emerging stock markets. The dual long memory property is prevalent in the returns and volatility of the emerging stock markets over the pre-crisis period. During the post-crisis period, the dual long memory process is only detected in the Moroccan market. The authors also display the dynamic behavior of VaR during the two sub-periods. In addition, based on the backtesting test, VaR performed better during the two sub-periods for all countries.
Chapter Preview


Over the past decade, many researchers have correspondingly investigated the behavior of stock returns in different financial markets. Thereby, volatility clustering, asymmetric leverage effects, and long memory, among others, are considered as stylized facts in finance. In particular, whether stock returns long memory, they show significant autocorrelation among observations which are largely separated in time (Barkoulas et al., 2000). In this respect, many researchers focus on investigating the long memory property in emerging markets. Even though the long memory property has become of paramount importance in empirical finance, the debate on existence of this issue in emerging stock returns still remain inconclusive. The recurring market microstructure change coupled with the alternate periods of financial turmoil can be impetus for analyzing the issue of long memory in such markets. Even though the rapid growth of emerging markets and their increasing importance attract the investors, these markets suffer from biases due to market severe thinness and nonsynchronous trading. As well, investors in emerging markets tend to react gradually and slowly to new information compared to developed markets (Barkoulas et al., 2000). Apart from the specific features of these markets, many emerging markets have experienced dramatic country-specific events such as the 1994 Mexican crisis, the 1997 East Asian crisis, the 2001 Argentine crisis and the 2011 Arab Spring. The revival of emerging financial markets, after the country-specific circumstances, represents a new challenge for many researchers. Notably, research on emerging markets focuses on the statistical nature of the market fluctuations and risk management, especially amid financial turmoil.

As aforementioned, emerging markets experienced great upheaval, involving that such markets collapsed and their volatility raised. A great deal of econometric research has been devoted to examine the influence of crises on the behavior of stock prices. In this respect, there is no precise and rigorous definition of the concepts of “stock market crisis”. These commonly used expressions are used to describe a brutal and widespread collapse of stock prices. The crash resulted in an almost continuous fall in stock market prices, following massive sales of securities held by investors. From empirical standpoint, the stock market crisis should not be confused with price volatility. The latter represents the amplitude of changes in stock prices over a given period (measured on the basis of stock market indices) and reflects the chronic instability of the stock market. The stock market crisis is generally accompanied by an increase in volatility. Thereby, the emerging markets exhibit an increased level of risk during the recurring turbulence in those markets. As a result, the use of standard measures can misevaluate the market risk. Under uncertainty in financial markets, measures of risk have a crucial role. Standard quantitative tools, such as Value-at-Risk (VaR) are used for measuring, managing and controlling risk. The Value-at-Risk (VaR) expresses the maximum loss of an asset or a portfolio in a given time line and confidence level. Several researches have focused on the evaluation of the VaR using different types of models. Despite the multitude of methods, there are no standard ones. The researches in this area have focused on the search for the most adequate method to the structure of financial returns especially in developed markets.

In this chapter, we focus our attention on the following questions: Does the long memory property in return and volatility change over time and among stock markets? If so, has the long memory property affected the market risk in the emerging markets? and how should the VaR be adjusted to these specific criteria?. For this end, we use stock markets indices for Tunisia, Egypt, Lebanon and Morocco for the different periods. Choosing such emerging markets is motivated by a comparable degree of financial markets and the country-specific circumstances. The sampled market data is divided into sub-periods based on the onset of political unrests.

The chapter is organized as follows. Section 1 presents the review literature on long memory and its definitions and measures. Section 2 reports the features of emerging markets. Section 3 presents the methodology and Section 4 reports data and descriptive statistics. Estimation results are shown in Section 5 and Section 6 concludes.

Complete Chapter List

Search this Book: