Aim: Financial crises are dangerous and frightening events with potentially severe consequences for investors, financial systems and even whole economies. Hence, we suppose that market participants show increased proneness to emotionally biased decisions during times of market distress. We test our hypothesis by analyzing two well-known behavioral effects: ambiguity aversion and selective perception. Design / Research methods: The authors should clearly explain the way in which the aim or objective is achieved. The main research methods as well as the approach to the research should be provided that enable effective dealing with the paper's aim.First, we use GARCH volatilities of major stock indices as a measure of market distress and monthly data from the Economic Policy Uncertainty Indicator (EPUI) as a proxy for the level of market uncertainty. By estimating the Granger causality, we test whether uncertainty causally influences market volatility, which could be interpreted as a sign of ambiguity aversion of market participants. Second, we use sub-indices of the EPUI regarding financial regulation, monetary policy, and economic policy as a proxy for market awareness of these topics. By regressing on GARCH volatilities, which serve again as the measure for crises, we analyze if investors' attention differs depending on market distress due to selective perception Conclusions / findings: Overall, we find mixed results. For ambiguity aversion, we find causality for the total sample as well as for the subsamples of the first oil crisis, the Latin America crisis, the Asian crisis, and the subprime crisis. For selective perception, we find significant results for the total sample as well, as for the Dot.Com bubble and the subprime crisis. Originality / value of the article: We add value by examining specific severe financial crises with respect to behavioral aspects of market participants. We want to learn whether the awareness of investors regarding important topics like monetary policy, financial regulation, and economic policy is stable over time and if uncertainty drives the market distress or vice versa. This knowledge is important to investors and policy makers. Implications of the research: Investors and decision-makers need to focus e.g. on current discussions regarding financial regulation not only in times of distress but also in normal times. Otherwise, policy makers will be forced to react in times of pressure and cannot proactively devise regulation. Limitations of the research: First, we did not check for spill-over effects. The question if volatility creates subsequent ripple effects in our framework is left for future research.Second, for the Japanese crisis we did not find causality in our ambiguity aversion analysis. The question whether the link between levels of uncertainty and volatility is stronger once a bubble bursts on domestic soil remains unanswered in our paper.
