Limits to Diversification: Tail Risks in Real Estate Portfolios

Abstract

This study addresses real estate’s riskiness from a distributional standpoint. Several studies have found that real estate returns are best modeled with stable Paretian distributions. This is confirmed using National Council of Real Estate Investment Fiduciaries individual property returns, but the first application of stable distributions to commercial real estate portfolio returns provides evidence that diversification effects ultimately reduce tailedness and, surprisingly, drive the tail parameter toward normality. Further insight is provided by highlighting the importance of a complete view, beyond pure tail parameter considerations. Even when tail parameters reflect normality, the return risk may still be tremendous, and it can only be reduced by diversification effects in property portfolios (and only to a certain, time-dependent extent).

TOPICS: Real estate, tail risks, statistical methods