A ReMeDI for Microstructure Noise

We introduce the Realized moMents of Disjoint Increments (ReMeDI) paradigm to measure microstructure noise (the deviation observed asset prices from fundamental values caused by market imperfections). propose consistent estimators arbitrary moments process based on high‐frequency data, where could be serially dependent, endogenous, and nonstationary. characterize limit distributions proposed construct confidence intervals under infill asymptotics. Our simulation empirical studies show that ReMeDI approach is very effective scale serial dependence noise. Moreover, are quite robust model specifications, sample sizes, data frequencies.