Modelling banking-hall yield for property investment

Abstract

Purpose – This paper aims to build a predictive model for the investment yield of British banking-halls. Design/methodology/approach – Empirical data of similar lots sold at previous auctions are subjected to statistical analyses utilizing a cross-sectional research design. The independent variables analysed are taken from a previous study using the same cases. Models are built using logistic regression and ANCOVA. Findings – Logistic regression generally generates better models than ANCOVA. A division of Britain on a north/south divide produces the best results. Rent is as good as lot size and price in modelling, but has greater utility, because it is known prior to auctions. Research limitations/implications – Cases analysed were restricted to lots let entirely as banking-halls. Other lots comprising premises only partially used as banking-halls might produce different results. Freehold was the only tenure tested. Practical implications – The study provides a form of predictive modelling for investors and their advisors using rent which is known in advance of any sale. Originality/value – The study makes an original contribution to the field, because it builds a predictive model for investment yields for this class of property. Further research may indicate if similar predictive models can be built for other classes of investment property.