"Econometrics of testing for jumps in financial economics using
bipower variation"
	
	
Ole E. Barndorff-Nielsen
The Centre for Mathematical Physics and Stochastics (MaPhySto),
University of Aarhus, Ny Munkegade, DK-8000 Aarhus C, Denmark

and

Neil Shephard
Nuffield College, Oxford OX1 1NF, UK.
		
Abstract: In this paper we provide an asymptotic distribution theory for some
non-parametric tests of the hypothesis that asset prices have continuous
sample paths. We study the behaviour of the tests using simulated data and
see that certain versions of the tests have good finite sample behaviour. We
also apply the tests to exchange rate data and show that the null of a
continuous sample path is frequently rejected. Most of the jumps the
statistics identify are associated with governmental macroeconomic
announcements.
	
Keywords: Bipower variation; Jump process; Quadratic variation;
Realised variance; Semimartingales; Stochastic volatility.