Forecasting Models For Predicting Commercial Rental Values In Dar Es Salaam City
Abstract
This paper uses historical rental values collected from a sample of prime commercial properties in Dar es Salaam city to propose a quantitative model to be used in forecasting rental values for properties falling under this category. The model proposed is one of the variants of regression analysis formulae which are widely used by econometricians in rental forecasting. The historical rental data used in this study were adjusted for inflation. From the regression analysis, four models (linear, compound, power and multivariate) were tested and compound model produced the highest coefficient of determination as well as high level of factor-by-factor interactions. When the data were fitted to the models, the trend of rents per square metre decreased gradually with time from its peak of around USD 20 in 1995 to around USD 18 in 2000 and around USD 13 in 2010. The estimate shows that both time period and inflation that change over time have significant negative effects on the rental values. The performance of the selected model is tested in terms of the goodness of fit between observed and estimated rental values from 1995 to 2010.
Keywords: Rental values, prime commercial properties, correlation, regression, ANOVA
Downloads
Published
How to Cite
Issue
Section
License
Authors must agree to abide by the terms of the Copyright Statement as per the licensing information of this journal.
Licensing Information for this journal:
Attribution-Non Commercial (CC BY-NC). This journal License allows others to download works, remix, adapt and build up works non commercially, and new works must also acknowledge other authors and be non commercial.