Sophisticated sensitivity : can developers guess smarter?

Abstract

The commercial real estate industry is currently in a state of turmoil, as access to capital markets is as constrained as consumer demand. Today many real estate development firms find themselves in difficult positions, with plummeting net operating income and upwardly mobile capitalization rates. Tail events - market events that were believed, based on statistics, to be rare occurrences - seem to be occurring with more and more regularity. With increasing uncertainty and market volatility, the question must be asked: how well can real estate developers predict returns? The purpose of this thesis is threefold: First, to determine whether real estate developers are accurately projecting real estate development returns; second, to determine where input assumption estimation errors are made in the ex ante proforma; and third, we undertake an analysis and application of Monte Carlo Simulation to ascertain whether, by providing practitioners another layer of transaction information, simulation is additive to the development return forecasting process. Through the careful analysis of both ex ante and ex post proformas of real estate development projects, this thesis is one of the first to show how well developers predict the outcomes of their projects. Our findings are rather surprising. We determine that ex ante and ex post real estate returns vary dramatically. On average expected development returns are shown to be 23.2%, while realized returns are only 9.4%. To understand this discrepancy we analyze each project proforma to identify where, during the valuation and development processes, developers made mistakes.

(cont.) Our findings suggest that developers are overly optimistic, especially when estimating hard costs, soft cost, and cashflow timing. The thesis results are consistent with the findings of a study by Dr. James Shilling, who analyzed the discrepancy between ex ante and ex post proforma returns for stabilized institutional properties. Shilling deduced that institutional investors are also misjudging returns, overestimating by an average of nearly 650 basis points. We also seek to augment and improve the valuation process employed by developers by applying Monte Carlo Simulation to discounted cashflow analysis. Applying Monte Carlo Simulation to the ex ante proforma of a real development transaction, we assess whether discounted cashflow analysis coupled with simulation provides an ex ante return that more closely approximates the realized ex post return. Again, our results are surprising. Among our findings, we learn that the simulation preparation process better informs a developer of sensitivities in input assumption variables for the transaction. However, industry data is not comprehensive, transparent, or available for a sufficiently long period of time to apply Monte Carlo Simulation. Despite the additional information provided by simulation, there remains the risk that a simulation proforma using incomplete data will yield inaccurate results. Due to the limited sample size used in our study we acknowledge that our results must be interpreted with some caution. However, we are hopeful that this initial effort to better understand and forecast development returns will encourage further study in this important area.