Testing the Semi Strong form of Market Efficiency

• Performance of the property pricing model

It is impossible to truly explain all pricing differences in the residential property market since nearly all properties are unique. But with an average adjusted R2 of 0.7, the model when used in this manner can identify a portfolio with a majority of underpriced properties and conversely a portfolio with a majority of overpriced properties.

Intuitively if properties are in fact underpriced with respect to others with the same qualities, then these underpriced properties should yield higher excess returns and alternatively overpriced properties should yield a lower excess return. And by testing this relationship we can have an insight on the accuracy of this pricing model when it is used to implicate these underpriced or overpriced properties. The following regression is used to test the performance of the pricing model.


If the pricing model can in fact identify underpriced properties then the coefficient β for the residuals from pricing, should be negative and significant, since a negative residual implies an underpriced property and thus should on average yield a higher return.


The regression shows a definite relationship in the excess returns and estimated residuals, with the correct sign for the coefficient. And thus confirms that the pricing model performs quite well in identifying underpriced and overpriced properties.

Since there is some estimation errors of the pricing model, the collective portfolio of under priced properties, should contain a majority of properties that are in fact underpriced and a minority of properties that are errors. And accordingly the estimated Sharpe ratio of this underpriced property portfolio should contain some downward bias and subsequently the significance test is less likely to show that there is a difference between the portfolios of total properties. The fact that the results of the z-test in table 7 show that there is still a significant difference between the total and underpriced portfolio irrespective to this estimation biased, we can safely conclude that the underpriced properties truly outperforms the market portfolio.

In summary these results show evidence that there are identifiable properties that are underpriced relative to the market situation in Auckland, and that opportunities of abnormal profits exist. And thus there is evidence to reject the Semi strong form of market efficiency, which states that all available information is reflected in the current market prices. Although these results are very convincing, to fully reject the Semi Strong form hypothesis, consistent abnormal returns has to be “exploitable”, and unfortunately by merely following this simple trading rule, it does not guarantee abnormal profits can be consistently gained.

There are large search costs involved to identify these properties, and these properties collectively provide only a marginal improvement on the market returns. Furthermore to get access to these returns the whole portfolio of underpriced properties has to be purchased in the ten years, which is both impossibly expensive and impractical in a residential property market. Thus we cannot fully reject the semi strong form of market efficiency, despite the fact that properties with higher returns can be identified using this trading rule.