We discuss two more universal features of stock markets: the so-called leve
rage effect (a negative correlation between past returns and future volatil
ity), and the increased downside correlations. For individual stocks, the l
everage correlation can be rationalized in terms of a new 'retarded' model
which interpolates between a purely additive and a purely multiplicative st
ochastic process. For stock indices a specific market panic phenomenon seem
s to be necessary to account for the observed amplitude of the effect. As f
or the increase of correlations in highly volatile periods, we investigate
how much of this effect can be explained within a simple non-Gaussian one-f
actor description with time independent correlations, In particular, this o
ne-factor model can explain the level and asymmetry of empirical exceedance
correlations, which reflects the fat-tailed and negatively skewed distribu
tion of market returns. (C) 2001 Elsevier Science B.V. All rights reserved.