Forecasting housing demand in Terengganu using multinomial logit

• Main Author: Ibrahim, Izzah Sofea
• Format: Thesis
• Language: English; English; English
• Published: 2018
• Subjects: HD7795-8027 Labor policy. Labor and the state; TA329-348 Engineering mathematics. Engineering analysis
• Online Access: http://eprints.uthm.edu.my/155/1/24p%20IZZAH%20SOFEA%20BINTI%20IBRAHIM.pdf; http://eprints.uthm.edu.my/155/2/IZZAH%20SOFEA%20BINTI%20IBRAHIM%20COPYRIGHT%20DECLARATION.pdf; http://eprints.uthm.edu.my/155/3/IZZAH%20SOFEA%20BINTI%20IBRAHIM%20WATERMARK.pdf; http://eprints.uthm.edu.my/155/

Housing demand is defined as the number of homes needed as residence per household. There are 3 types of housing developed in Terengganu which are low medium cost housing, medium cost housing and high cost housing. According to the NAPIC, during the 2nd quarter of 2016, Terengganu recorded the highest number of overhang units in the East Coast. This problem shows that there is a mismatch between supply and demand. Therefore, to overcome this housing supply and demand mismatch, a reliable technique to forecast the demand is carried out as a basis for the house provider to supply adequate units of housing. The aim of this study is to determine the demand for housing in Terengganu state based on the 3 types of housing. Relevant variables from Census data of 2010 was filtered and analysed to determine the headship rate. After that, the population was forecasted using Double Exponential Smoothing (DES). Model of forecasting housing demand was developed and the model was validated. The results showed that the elderly age group recorded the highest headship rate (11%) while the youngest age group recorded the lowest headship rate (0%) for the year 2010. In contrast, the population among the residents in Terengganu was the highest (16%) for the youngest age group and the lowest (0.95%) for the elderly age group. The total housing demand in Terengganu increases in parallel with the population. Significant indicators of housing demand were income, age and education attainment. The forecasting model showed that in 5 years duration, low medium cost housing had the highest demand followed by medium cost housing and high cost housing. The validation analysis showed that this model is reasonable and can be used to forecast housing demand in Terengganu with a Mean Absolute Percentage Error (MAPE) value of 22.45%. These forecasted housing demands will be useful for key players in housing industry such as government authorities and developers in providing numbers of housing unit that match the demand to avoid under supply or over supply.