How Do Investors React Under Uncertainty?

It has long been accepted in finance that risk plays an important role in determining valuation where risk reflects that investors are unsure as to the exact value of future returns but are able to express their prior expectations by way of a probability distribution of these returns. Knights (1921) introduced the concept of uncertainty where we possess incomplete knowledge about this distribution and so are unable to formulate priors over all possible outcomes. A number of writers (Gilboa and Schmeidler, 1989; Epstein and Schneider, 2003) have developed models that suggest that ambiguity, like risk, has a negative impact on valuation. The most common approach taken in these models is to assume that investors take a conservative approach when faced with uncertainty and base their decisions on the worst case scenario (maxmin expected utility). The area on which we concentrate in this paper is how the market faced with uncertainty reacts to the receipt of new information. The proposition being that under maxmin expected utility, the interpretation that the market will place on any information received will become more pessimistic as uncertainty increases, upgrading any bad news and downgrading any good news. Williams (2009) uses changes in the VIX (i.e. implied market volatility) as a measure of market uncertainty in his US study where he evaluates the markets response to the release of earnings news. There is a plethora of evidence dating back to Ball and Brown (1968) that confirms that the market responds positively (negatively) to good (bad) news earnings announcements. Williams finds that this response is conditioned by market uncertainty with there being the predicted asymmetric reaction to good and bad earnings news – the negative reaction to bad news increasing with uncertainty and the positive reaction to good news decreasing. In this study we use Australian data to also examine the impact of uncertainty on the market response to earnings announcements. One important difference in our findings to those of Williams is that it is not only changes in VIX but also the level of VIX that influence how the market responds to earnings information. Although generally confirming a pessimistic response by investors to earnings released at a time of high market uncertainly, we find evidence of a slight optimistic bias in the reaction of investors to earnings released at a time of low market uncertainty. We also find that the level of pessimism engendered when uncertainly is high may be significantly diluted if it occurs contemporaneously with strong market sentiment.