Transmission of volatility from the USA Stock Market to SA Stock Market during the subprime crisis

1
Transmission of volatility from the USA Stock
Market to SA Stock Market during the sub-prime
crisis

• Thuletho Zwane and Ziv Chinzara
• Rhodes University
• Paper for presentation at the 18th Annual
  Conference of
• the Southern African Finance Association

2
LAYOUT OF THE PRESENTATION

• INTRODUCTION AND BACKGROUND
• DATA AND DATA ISSUES
• METHODOLOGY
• RESULTS
• CONCLUSIONS
• FURTHER RESEARCH

3
Introduction

• Understanding volatility transmission between
  stock markets is important for the following
• Financial stability
• Monetary policy
• Portfolio diversification
• Regulatory policy

4
Introduction Continued

• Volatility transmission between stock markets has
  been widely studied especially in developed
  countries and emerging markets(cf Becket et al,
  1990, Morana and Belratti, 2007, Koutmous, 1996,
  Segot and Lucey, 2005).
• Some of the issues that have been raised in
  this area are
• Interdependence versus Contagion.
• Symmetry versus Asymmetry
• Contagion is a situation whereby comovement of
  stock markets increase during the financial
  crises.
• Asymmetry occurs when bad news increases
  volatility by a bigger magnitude than good news.

5
Introduction Continued

• The aims of this study are to
• Analyse the transmission of volatility from the
  US stock market to the SA stock market and to
  investigate if there was a significant change in
  the size of risk transmitted during the sub-prime
  crisis.
• Analyse the nature of volatility in each of the
  stock markets (i.e. symmetry or asymmetry)

6
Data and Data Issues

• The study utilizes the daily closing stock market
  indices. We use the SP500 for the US stock market
  and the JSE All share for SA stock market for
  period 30/06/1995 30/08/2008.
• The data was sourced from Thompson DataStream.
• Consistent with some previous studies,
  non-trading days were deleted across all markets.
• Daily data was more preferable to lower frequency
  data since stock market promptly react to new
  information.

7
Methodology

• The study uses the AR GARCH and AR EGRACH
  model.
• The autoregressive component was included to
  minimise autocorrelation.
• The variance equations for the GARCH and EGARCH
  models are specified as follows, respectively
• ht ? ae2t-1 ßht-1 , a ß lt 1

8
Methodology Continued

• In the above specifications, a and ß are the
  coefficients of the squared error term and the
  lagged variance terms respectively.
• For the stationarity of the models a ß lt 1.
• ? is the coefficient of asymmetry. If ? lt 0 ,
  then there is evidence of asymmetry and leverage
  effects.

9
RESULTSDescriptive Statistics
10
Results

• A Comparison of the GARCH Models

11
Findings Testing for structural break using a
dummy variable
12
Conclusion

• Results show that volatility in the SA stock
  market is more explained by own-past volatility
  than volatility in the US stock market.
• Volatility in both stock markets is inherently
  asymmetric and there is evidence of leverage
  effect.
• Volatility transmission from the US to the SA
  stock market did not significantly increase
  during the subprime crisis. This is inconsistent
  to the contagion hypothesis.
• This could point out to more effective regulatory
  measures in the SA banking and financial
  industry than the US counterpart.

13
Further Research

• More data should be used.
• Methodology Multivariate GARCH and other
  econometric techniques could be explored.