leverage effect has been found in our data, similarly no support for the time-varying risk
premium as an explanation was found (see further discussion in Talpsepp and Rieger,
2009).
As our findings show that the level of volatility asymmetry tends to be increasing in
time and can differ significantly across countries, it is natural to wonder whether
economic development or structure can play a role in explaining the differences in
asymmetry. We test a number of more or less direct measures of market development
(including GDP/capita, different published market development and efficiency indexes,
etc.) under different regression setups to check the hypothesis. All our test results
indicate that the level of asymmetry is not related to a lack of market efficiency: quite
the contrary, a higher level of economic development and market efficiency is associated
with a higher level of volatility asymmetry!
This is certainly a surprising result that we want to understand in the remainder of this
chapter.
11.2.4 Volatility asymmetry, news, and individual investors
Recent research has argued that the media have the power to influence investor
sentiment and thus prices on the stock market (see, e.g., Tetlock, 2007). Hence,
information obtained from the media might potentially cause volatility asymmetry.
In fact, volatility asymmetry is positively related to analyst coverage in the data
(Talpsepp and Rieger, 2009). This links back to the role of the media in financial
markets, since analyst opinions are generally transmitted by different media channels.
What, then, are the characteristics of the media coverage time-series that could shed
light on this relationship? In a forthcoming working paper Dzielinski, Steude, and
Subasi (forthcoming) look at the daily media sentiment for the constituents of the
Volatility asymmetry, news, and private investors 259Figure 11.2.Time dynamics of volatility asymmetry (gamma) for the MSCI World and MSCI
Emerging Markets indexes. Values over zero indicate asymmetry where volatility is higher when
prices fall, values below zero mean that volatility is higher when the market goes up.