Temporal aggregation and seasonality problems in Malaysian housing index

Abstract

The housing market has is one of the largest asset bases in Malaysia and housing index is an important housing performance indicator. The rapid growth of property investment leads to transformation of physical property investment into financial products such as housing derivatives or home equity insurance based on housing index. There are two issues to be considered in this paper which are temporal aggregation and seasonality problem in housing index. Geltner (1993) mentioned that temporal aggregation will lead to smooth and erroneously eliminate noise in house price movement. As a result, houses become less risky and the volatility of house price is understated which may not be true. For example, the risk and return of quarterly index may be lower compared to monthly index because the information of monthly index is missing. Moreover, Kaplan (2010) defined seasonality in a time series is a pattern that tends to repeat from year to year. In other word, seasonality appeared in time series when the price or index movement is repetitive around the trend line. If the housing index consists of seasonal effect it will mislead measurement on real housing performance based on economic situation. In order to investigate these problems quality adjusted index will be developed by hedonic model based on Laspeyres, Paasche, Fisher and time dummy price index formula. The case study to be investigated is housing index of double storey house in Plentong Johor. Thus f-ratio and volatility of housing index for each period (quarterly, monthly and yearly) used to evaluate temporal aggregation problem. Seasonal Analysis will be adopted in this study to identify the existence of seasonality problem in monthly housing index. It contained two objectives in this study. First is to determine whether housing index should be represented in the most disaggregated level (monthly index). Second is to identify the existence of seasonality problem in housing index.