Degree Name

Doctor of Philosophy


School of Accounting and Finance


This thesis fills a number of gaps in both the Australian and overseas literature on the value premium, particularly with regard to the dearth of Australian studies on behavioural explanations and the role default-risk plays in such explanations. First, whilst a number of studies have investigated the value premium in Australia in the context of rational asset pricing, there are no identified studies in the academic literature that specifically investigate the existence of mispriced Australian value and growth stocks. This gap is addressed by exploring the overvaluation of financially distressed growth stocks and the undervaluation of low default-risk value stocks. Second, there are no Australian studies that directly test the errors-in-expectations hypothesis using biases in analysts’ earnings forecasts, as there are for some overseas markets. This analysis is carried out for the Australian market, but unlike previous studies introduces a highly-relevant control variable: financial distress. Finally, previous studies have shown conflicting patterns of returns and analyst forecast errors across value and growth classifications (high book-to-market stocks simultaneously have over-optimistic forecasts and high returns). There have been no attempts to reconcile these conflicting patterns, a gap addressed here by studying the relative importance of book-to-market, default-risk and analyst agreement to the market reaction to earnings surprises.

The findings of this thesis are consistent with overvaluation of distressed growth stocks and undervaluation of low default-risk value stocks, and support the validity of mispricing as a potential explanation of the value premium in Australia; a conclusion to which previous Australian studies have been averse. The findings also demonstrate a marked bias in analysts’ earnings forecasts which is consistent with analyst underreaction to distress, which dominates the relationship between analyst optimism and valuation ratios, and which reveals a previously-undocumented problem with the use of such forecasts to test the errors-in-expectations hypothesis. The findings also shed light on the conflicting patterns of returns and forecast errors across value and growth stocks. This anomaly is largely explainable by the previously-documented asymmetric reactions of value and growth stocks to earnings torpedoes; the novel finding here is that it is not explainable by differences in analyst agreement (which affects earnings response coefficients) or in default-risk.

In general terms, the thesis makes contributions to the literature on market and analyst efficiency, behavioural finance, the momentum life cycle, the pricing of default-risk, and earnings response coefficients.