Testing the weak form of the Property Market Efficiency

• Excess Returns

The excess returns of the property samples were calculated by the change in the nominal house price index of samples A and B, less the annual Treasury bill yield for the relevant period (risk free rate). No exact rental data is available for the time period of study, and even if available, an application of a proxy for relevant rents will be highly inaccurate since there may be significant quality differences in owner occupied properties (Case & Shiller 1989). So all excess returns are based solely on the appreciation of properties, and since in general there is no capital gains tax in New Zealand, taxes were also not included in the profits. The same test that was used to predict the next quarter returns from before is used to test the predictability of excess returns, the results shown below.




According to results above, there is insignificant proof to conclude that the past excess returns can explain future excess returns. Even though the real appreciation of residential property prices can be well predicted by past movements, following the transformation with the quarterly risk free rate, these resulting quarterly changes of the excess returns has no significant predictive power of future changes. This subsequent unpredictability is entirely due to the unpredictability of the real interest rate (an unpredictable opportunity cost), a resulting relationship that is in contrast to the results obtained by Case and Shiller (1989). Their study based on an alternative geographical location and time period of study, the real interest rates were highly predictable and in fact improved the predictability of the excess returns (Case and Shiller 1989).

In summary these result imply an acceptance of the weak-form market efficiency hypothesis, when no superior profits can be made by simply trading on the past movements in property prices. Due to the unpredictability of the real interest rate, the future excess returns or profits cannot be properly determined by trading on past information portrayed by the index..