Comment on “Realized Variance and Market Microstructure Noise” by Peter R. Hansen and Asger Lunde

We find ourselves very much in agreement with the thrust of HL’s message concerning the complexity induced by microstructure noise. In particular, we agree that noise is time dependent and correlated with the efficient price - features that in our view are a necessary consequence of the observed form of market transactions, as we have argued above - and that the properties of noise inevitably evolve over time, again just as the efficient price is itself evolutionary. We further agree that microstructure noise cannot be accommodated by simple specifications. Since microstructure noise at ultra high infill sampling frequencies often off-sets the actual transactions data to the latent efficient price, the complexity of microstructure noise includes local nonstationarity and perfect correlation with the efficient price. These are properties that are not permitted in the models and methods presently used in the literature. However, there are empirical procedures that are capable of addressing these additional complexities as we have indicated in parts of our discussion. We join the authors in saying there is still much to do in this exciting field and we look forward to further developments that build on the work they and others have done recently.