Prediction of variability in mortgage rates: interval computing solutions

Purpose – The purpose of this paper is to forecast variability in mortgage rates by using interval measured data and interval computing method. Design/methodology/approach – Variability (interval) forecasts generated by the interval computing are compared with lower‐ and upper‐bound forecasts based on the ordinary least squares (OLS) rolling regressions. Findings – On average, 56 per cent of annual changes in mortgage rates may be predicted by OLS lower‐ and upper‐bound forecasts while the interval method improves forecasting accuracy to 72 per cent. Research limitations/implications – This paper uses the interval computing method to forecast variability in mortgage rates. Future studies may expand variability forecasting into more risk‐managing areas. Practical implications – Results of this study may be interesting to executive officers of banks, mortgage companies, and insurance companies, builders, investors, and other financial decision makers with an interest in mortgage rates. Originality/value – Although it is well‐known that c hanges in mortgage rates can significantly affect the housing market and economy, there is not much serious research that attempts to forecast variability in mortgage rates in the literature. This study is the first endeavor in variability forecasting for mortgage rates.