Applying System Dynamics to Simulate and Forecast Rental Real Estate Market

Abstract

This research explores the utilization of system dynamics modeling methodology to simulate and forecast a sub-market within the real estate industry. By doing so, this research examines the feasibility and potential of a system dynamics-based tool that could reliably forecast future trends and inform decision-making for businesses in a sub-market. It is based on the original system dynamics model for real estate markets as developed by John Sterman (Sterman, Case Study: Boom and Bust in Real Estate Markets 2000), and other subsequent examples of this methodology’s application in a real estate context since. It expands on this existing literature by recognizing and incorporating concepts central to the real estate industry, such as rental rates, affordability, absorption, inflation, cap rates, and rental prices, as key for predicting market movements.

As a test bed, the multifamily rental housing in the South Boston region is identified for application. The study thus predicts short-term movement for the multifamily assets in this sub-market in comparison to forecasts from other major sources. It also highlights the limitations of this approach, such as the smoothing effect of generated data and its limitations in capturing seasonality in the market. The study further explores potential avenues for enhancing the functionality and accuracy of forecasts by endogenizing additional factors, thus establishing a foundation for subsequent research.