Forecasting the Thailand housing price indexA case study on condo HPI during the COVID-19

Abstract

HPI (house price index) measures the price development of houses sold to households. And it measures the price as a percentage change from some specific start date, which is treated as the benchmark with the index at 100. In another word, HPI reflects the housing market fluctuation in some respects. Moreover, real estate is a good type of investment for people for avoiding various risks. Hence, under the global rampant of COVID-19, forecasting the possible short-term floating of HPI would offer some foresight to the residential market both on individual and policy sides. In this paper, five kinds of models were constructed with the data before Thailand reopened (1st Jul 2020) and forecasting for the followed the third quarter HPI. Lastly, the capacity of every model will be compared by an error matrix. And the experiments demonstrated multiple regression model, based on accuracy, performs relatively better than the other four models.