We use cross-sectional data about market participation level to capture the share of
individual investors in the market. We find a significant positive impact of market
participation level on volatility asymmetry. Since market participation data are not
available for the whole sample we run some additional tests to confirm the finding.
We use available time-series data for a single country (Switzerland) which support the
finding also at the time-series level.^2 We also use panel data about market capitalization
divided by GDP as a proxy for share of private investors in the market. Here, we find a
positive correlation confirming the finding that the more individual investors (who are
likely to be less experienced and/or informed) in the market the higher the volatility
asymmetry.
Based on these results we hypothesize that in the case of bad news a higher absolute
number of investors will be selling and pushing prices down more quickly, thus increas-
ing volatility during periods when prices fall. This could be the explanation for the
different behavior of investors after prices fall or rise, which would be consistent with
the ideas of Hens and Steude (2009) and Shefrin (2005). It would also be consistent with
the assumption that more analysts and media attention in the case of bad news can cause
asymmetry.
Due to the lack of cross-country data, we are not able to test whether the activities of
hedge funds (who can also act as arbitragers) can impact volatility asymmetry. But we
are able to check whether introducing electronic trading platforms (facilitating stop loss
orders and better and faster access to the markets) affects asymmetry, which could also
be one of the differences between developed and emerging markets. We do not find any
significant effect of electronic trading on the cross-sectional or panel data level.
11.3 Who makes markets volatile?
11.3.1 Google and volatility
So far we have seen that the degree of volatility asymmetry is linked to two
characteristics of the financial market in question: the share of private investors and
the number of stock analysts. The aim of this section is to illustrate in a more detailed
Volatility asymmetry, news, and private investors 261Table 11.1.Log–log regressions on volatility asymmetry (adjusted gamma)
Coefficient t-statistic Coefficient t-statisticAnalyst coverage 0.47 3.14 0.45 4.21
GDP per capita 1.36 3.61 0.69 2.42
Market participation 0.29 2.48
Media 5.04 4.1 1.95 1.68
Stock market capitalization/GDP 0.23 2.19*
Constant 6.97 3.13*** 0.52 0.2
N 24 40
R^2 0.74 0.71
, , : Significant at the 10%, 5% and 1% level.
(^2) Please see Talpsepp and Rieger (2009) for more details and for a discussion of the variables used.