Skip to Content
MathematicsMathematics
  • Article
  • Open Access

19 November 2026

20 Pages

Polynomial Approximation of Conditional Expectations by Meixner–Pollaczek Martingales

Department of Mathematics, College of Science, King Saud University, Riyadh 11451, Saudi Arabia
This article belongs to the Section D1: Probability and Statistics

Abstract

We study the approximation of conditional expectations associated with the symmetric Meixner Lévy process by Meixner–Pollaczek polynomial martingales. Starting from the classical orthogonal polynomial martingales, we compute explicit Fourier multipliers for the spatial coefficients in the conditional expectation expansion and identify its finite sums as fixed-time orthogonal projections. The time independence of the coefficients follows from the standard martingale principle. We establish an exact formula for the squared L2 error in terms of the tail of a negative binomial distribution. We prove convergence for every function in L2(R) and obtain an estimate for the expected squared maximal error using Doob’s inequality. For bandlimited functions, the convergence is geometric. For functions in a bounded subset of the Sobolev space Hs(R), with s>0, we establish the sharp uniform rate (logN)−2s for the squared approximation error, both at fixed time and for the expected maximal error, up to multiplicative constants. We also determine the Sobolev regularity of the indicator function of an interval and show that it belongs to the Sobolev space of order s precisely when s<1/2. For a particular symmetric choice of parameters, we obtain the asymptotic formula 1/(πlogN) for its squared error at the terminal time.

Article Metrics

Citations

Article Access Statistics

Article metric data becomes available approximately 24 hours after publication online.