Check leverage phenomenon on ARCH-GARCH models in R I estimated an ARCH(10) and a GARCH(1,1) models on R. I have to verify the presence of leverage phenomenon, graphically or descriptive evidence. Leverage effect: volatility reacts asymmetrically to the changes in prices (i.e. usually in stock returns the volatility increases more after bad news than after good news).
How can I test it?
 A: I do not remember if vanilla ARCH and GARCH models are capable of accounting for leverage. At least directly they are not (there is no parameter tied specifically to leverage), but there are models like GJR-GARCH that are more suitable for that. 
An idea: collect all the instances of bad news and the volatilities of the following periods; do the same with the good news. Compare the distribution of the former volatilities with that of the latter volatilities. The differences in the distributions can be informative of the leverage effect (you would look for difference in location of the distributions). Use a formal test, e.g. something like a $t$-test or rather a version that is robust to nonnormality of volatilities, to test whether the means of the two distributions are equal. One problem will be that volatilites are not observed, just estimated, so the observations are noisy proxies for the real volatilities, and you cannot take the test results at their face value. 
Another idea: use a GJR-GARCH model and see (test) if the leverage parameter is estimated at zero.
