Generalizing the OLS and Grid Estimators

Abstract

The vast majority of market valuations employ either some formal estimator such as ordinary least squares (OLS) or rely upon an informal set of rules defining the grid adjustment estimator. The success of the grid adjustment estimator suggests the data do not obey the ideal assumptions underlying OLS. However, the grid adjustment estimator's lack of a formal statistical foundation makes it difficult to use for inference and other purposes. This article demonstrates how to generalize the grid estimator and OLS to potentially obtain the best features of both. Interestingly, the generalization defines a spatial autoregression. On an empirical example the spatial autoregression outperforms the grid estimator which in turn outperforms OLS.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 74 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.