Symmetry classification and invariant solutions of the classical geometric mean reversion process
Jin Zhang, Dapeng Gao
TL;DR
This work applies Lie symmetry analysis to the Feynman-Kac PDE associated with the classical geometric mean reversion process, identifying a rich symmetry structure that depends on model parameters. It derives the full Lie algebra of infinitesimal symmetries, constructs an optimal system of one-dimensional subalgebras, and performs symmetry reductions to produce closed-form invariant solutions expressed through special functions. The resulting invariant solutions provide exact pricing- and value-related formulas for the GMR dynamics and offer insights for applications in discounted project values and commodity pricing. Overall, the paper delivers a systematic symmetry-based framework for solving and interpreting the GMR pricing PDE with nonnegative, mean-reverting dynamics.
Abstract
Based on the Lie symmetry method, we investigate a Feynman-Kac formula for the classical geometric mean reversion process, which effectively describing the dynamics of short-term interest rates. The Lie algebra of infinitesimal symmetries and the corresponding one-parameter symmetry groups of the equation are obtained. An optimal system of invariant solutions are constructed by a derived optimal system of one-dimensional subalgebras. Because of taking into account a supply response to price rises, this equation provides for a more realistic assumption than the geometric Brownian motion in many investment scenarios.
