How Does COVID-19 Impact the Efficiency of the Chinese Stock Market?: A Sliding Windows Approach

Introduction

The outbreak of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) has already about 20.5 million confirmed cases and killed more than 743 000 people as at August 11, 2020 worldwide, with numbers increasing on a daily basis. To limit the syndrome, social isolation is recommended (Kucharski, Russell, Diamond, Liu, Edmunds, Funk, & Eggo, 2020) and this means that a relevant part of the planet has to stay at home to decrease the likelihood of contamination. This isolation is affecting the entire global production chain and several economic sectors are at a standstill. In this context, President Donald Trump has sent a package of USD 1.6 trillion to the American Congress to help the American economy, the European Central Bank has decided to buy 750 billion euros in public and private bonds, the European Commission approved in July also a total of 750 billion euros to fight the crisis and the Chinese Central Bank has distributed USD 175 billion dollars in aid, making the COVID crisis the biggest since the subprime crisis (Baldwin & di Mauro, 2020). This is affecting the global economy in general and stock markets in particular. The continuing monitoring of stock market behavior is relevant to keep investors informed. In this chapter, we analyse the Chinese stock market because China was the first country to be affected and because at present the disease seems to be controlled, which probably makes the market more stable and suitable for this analysis.

In particular, we will analyse the long-range dependence behavior of the Shanghai Stock Exchange (SSE) Composite Index, in an analysis which could be related to the Efficient Market Hypothesis (EMH), proposed by Fama (1970), stating that a market is efficient when it reflects all the available information at a given time. Identifying three types of efficiency, the weak form of efficiency indicates there is no long-term memory in return rates’ financial series. Based on the hypothesis that asset prices follow a random walk, previously defended and studied by Bachelier (1900), it has been highly scrutinized since the 1970s. For extensive reviews of the study of the EMH in financial markets in general, see the work of Yen and Lee (2008) or Titan (2015).

In this chapter, we will use the Hurst exponent, an indicator often used to analyse the long-term memory of financial time series and particularly stock markets and/or assets, that exponent being identified as a good measure of dependence. We will estimate the exponent using the Detrended Fluctuation Analysis (DFA) and apply a sliding windows approach, allowing us to identify the time-varying behavior of dependence. In particular, we are interested in understanding how the COVID-19 situation affected the dependence of the Chinese stock market over time, so we will use a sliding windows approach with DFA to analyse the SSE, using high frequency data (with one minute-based data). The evolution of the Hurst exponent across time may reveal much information about markets’ features such as liquidity and underdevelopment. In the case of turbulent periods, several studies found a decreasing trend in the Hurst exponent (Kristoufek, 2012 and Morales, Di Matteo, Gramatica &Aste, 2012) with the Hurst exponent less than 0.5 according to Peters (1994), and an increase in short period trading activities, while an increasing trend in the local Hurst exponent, with a value greater than 0.5, may inform about market stability.

This type of information is very relevant since it could highlight financial markets’ behavior in the occurrence of pandemics, which is relevant for both investors and the authorities, in order to be prepared for similar crises in the future.