Estimating quadratic variation when quoted prices jump by a constant increment
Jeremy Large
Nuffield College,
Oxford OX1 1NF, UK
Abstract:
Financial assets' quoted prices normally change through frequent
revisions, or jumps. For markets where quotes are almost always
revised by the minimum price tick, this paper proposes a new
estimator of Quadratic Variation which is robust to microstructure
effects. It compares the number of alternations, where quotes are
revised back to their previous price, to the number of other
jumps. Many markets exhibit a lack of autocorrelation in their
quotes' alternation pattern. Under quite general 'no leverage'
assumptions, whenever this is so the proposed statistic is
consistent as the intensity of jumps increases without bound.
After an empirical implementation, some useful corollaries of this
are given.