On Pathwise Uniform Approximation of Processes with Càdlàg Trajectories by Processes with Minimal Total Variation

Abstract: For a real cadlag function $f$ and positive constant $c$ we find another cadlag function, which has the smallest total variation possible among the functions uniformly approximating f with accuracy c=2. The solution is expressed with the truncated variation, upward truncated variation and downward truncated variation introduced in the papers R. \L ochowski, Truncated variation of Brownian motion with drift, Bull. Pol. Acad. Sci. Math. Vol. 56 (2008) and R. \L ochowski, Truncated variation, upward truncated variation and downward truncated variation of Brownian motion with drift - their characteristics and applications Stochastic Processes and their Applications 121 (2011). They are analogs of Hahn-Jordan decomposition of a cadlag function with finite total variation but are always finite even if the total variation is infinite. We apply obtained results to general stochastic processes with c?adl?ag trajectories and in the special case of Brownian motion with drift we apply them to obtain full characterisation of its truncated variation by calculating its Laplace transform. We also calculate covariance of upward and downward truncated variations of Brownian motion with drift. Keywords: total variation, truncated variation, uniform approximation, Brow- nian motion, Laplace transform