Industrial Real Estate Cycles: Markov Chain Applications

Abstract

Executive Summary Adding a stochastic element to a well-understood real estate cycle model offers opportunities like those seen in earlier such syntheses of real estate analysis and statistics. The discrete real estate cycle points in the model require a discrete probability model, here a first order Markov chain. Many statistical applications flow from the combined model. Three Markov chain count variables have obvious real estate cycle appeal. Staying time, first recurrence time, and first passage time already exist in the Markov chain literature but only staying time is in the real estate cycle literature. The most fundamental innovation is in probabilistic forecasting. Being able to describe real estate cycle risk, cycle point by cycle point many quarters ahead, could improve evaluation of prospects for property disposal. It is also a simple spreadsheet application to describe real estate cycle risks that influence cash flows from operations across four-quarter spans.