Appraisal‐Based Real Estate Returns under Alternative Market Regimes
用蒙特卡洛模拟研究房地产回报的统计特性，比较贝叶斯和非贝叶斯评估规则，发现指数平滑和卡尔曼滤波在个体和组合层面表现良好，且当评估目标是回报而非价值时，应完全依赖当前市场数据。
In this article we use Monte Carlo simulation to study the statistical properties of real estate returns. We set up a model where transactions prices are noisy signals of true prices. We then consider a number of appraisal rules, derived from Bayesian and non‐Bayesian theory, to estimate the current true price and rate of return. The class of exponential smoothing and Kalman filter rules perform well at both the disaggregate (returns on an individual property) and aggregate (returns on a real property portfolio) levels. A special case of exponential smoothing (α= 1.0) places all weight on current market data. Since this case eliminates smoothing, our results suggest that appraisers should place all weight on current data (no weight on past data) provided that they want to estimate returns rather than values. However, these results should be used with caution if sales prices are very noisy.