In this thesis, I analyse the pricing ability of major asset pricing models using an out of sample data set from the New Zealand stock market in 1899-1929. Previous literature argues that the observed predictive ability of models is a result of the research design and database used to conduct the research. This unique dataset provides a way to test these models on data that has not been a component in previous models and testing. I find evidence of positive returns in the value and size portfolios, with high returns in momentum. There is strong evidence that Carhart’s (1997) four factor model captures variation in returns, with observed size and weaker book-to-market effects. Cross-sectional tests identify existence of a strong positive relation to momentum and profitability.