- Behavioural Heterogeneity in Singapore’s Housing Market
Awards
Author(s):
- Eugene Or
Email:
- eor001@e.ntu.edu.sg
Category:
- Economics
Institution:
- Nanyang Technological University
Region:
- Asia
Winner Category:
- Regional Winner
Year:
- 2020
Abstract:
- In recent years, the Heterogeneous Agent Model (HAM) has been proven to be one of the more successful frameworks in forecasting future asset prices. In particular, HAM acknowledges that the expectation of full rationality, a characteristic of the benchmark Rational Expectations Model (BRM), is often unrealistic. Thus, the model takes into account the behavioural heterogeneity and bounded rationality of players in the asset market. Additionally, we also recognise the possibility of regime-dependent behaviours. Therefore, in this paper, we apply HAM within the Markov switching framework in Singapore’s housing market and compare it with the BRM. We seek to find out if between-group and within-group behavioural heterogeneity exist in Singapore’s housing market, and if the Markov Regime-Switching HAM or the BRM is a better tool in estimating and predicting housing prices. We conduct our research using data of private housing in Singapore from 2004Q1 to 2018Q4. In general, our results suggest a coexistence of four groups of investors (between-group behavioural heterogeneity) as well as difference in behaviours across different market conditions (within-group behavioural heterogeneity). In line with the presence of between-group and within-group behavioural heterogeneity, our in-sample estimation efficiency tests indicate that the Markov Regime-Switching HAM, in comparison to the BRM, has a better fit and is better specified. Similarly, results from out-of-sample predictive power show that the former provides a better forecast than the latter. All in all, the Markov Regime-Switching HAM is a more reliable model to utilise for analysis on housing price dynamics in Singapore.
Attached Documents: