HELENA CHULIÁ ET AL. 317(ε^22 ). Moreover, the coefficient for its own asymmetric term (η^22 ) is significant,
indicating that the negative shocks on the IBEX35 returns affect more its
volatility than the positive shocks.
Panel B indicates that, during the pre-September 11 period, the S&P500
volatility is directly affected by its own volatility (h1,1) but not by the IBEX35
volatility. The results show that the S&P500 volatility is neither affected by
its own shocks (ε^21 ) nor by the shocks on the IBEX35 (ε^22 ). Thus, our findings
suggest that S&P500 volatility is only affected by its own volatility.
The equation for the IBEX35 in Panel B, shows that the index volatility is
only affected by its own volatility (h2,2). Therefore, higher levels of condi-
tional volatility in the past are associated with higher conditional volatility
in the current period. Our results indicate that the IBEX35 volatility is not
affected by shocks affecting neither the IBEX35 returns (ε^22 ) nor the S&P500
returns (ε^21 ).
Panel C displays the result for the post-September 11 period. The S&P500
volatility is directly affected by its own volatility (h1,1) but not by the IBEX35
volatility (h2,2). Our findings suggest that the S&P500 volatility is affected by
its own shocks (ε^21 ) and the IBEX35 shocks (ε^22 ). Finally, the coefficient for its
own asymmetric term (η^21 ) is significant, indicating that the negative shocks
on the S&P500 returns affect more its volatility than the positive shocks.
The IBEX35 volatility is affected by its own volatility (h2,2) and by the
S&P500 volatility (h1,1). Again, the IBEX35 volatility is not affected by its
own shocks (ε^22 ) but it is affected by the S&P500 shocks (ε^21 ). Finally, the
coefficient for its own asymmetric term (η^22 ) is significant, indicating that
the negative shocks on the IBEX35 returns affect more its volatility than the
positive shocks.
If we compare the results for the pre-September 11 and post-September
11 periods, it can be observed that, before September 11, both variances
(S&P500 and IBEX35) were only affected by their own past volatilities. How-
ever, after September 11, the S&P500 volatility is also affected by its own
shocks and the IBEX35 shocks. On the other hand, the IBEX35 volatility is
also affected by the S&P500 shocks and the S&P500 variance. These results
show that, after September 11, there has been an increase in the volatility
transmission between the USA and Spain.
16.4.2 Asymmetric volatility impulse response functions (AVIRF)
Figures 16.2 to 16.7 present the AVIRFs computed following Lin (1997) and
Meneu and Torró (2003). When unexpected shocks are positive (Figures 16.2,
16.4 and 16.6), graphical analysis shows that in the post-September 11 and
total period, there exists a significant volatility spillover from the S&P500 to
the IBEX35 (about 5% of the shock, Figures 16.2-C and 16.6-C). Thus, there
is no significant volatility spillover from the S&P500 to the IBEX35 before