Abstract:
This paper provides limit distribution results for power variation, that is, sums of powers of absolute increments under nonequidistant subdivisions of time and for certain types of time-changed Brownian motion and $\alpha$-stable processes. Special cases of these processes are stochastic volatility models used extensively in financial econometrics.
Keywords:power variation, $r$-variation, realized variance, semimartingales, stochastic volatility, time change.