This paper uses the event study methodology to analyze the impact of unexpected political event on stocks abnormal returns. The objective is twofold. The first is to reach robust estimates of stocks abnormal returns. The second is to reach robust estimates of the effects of unexpected political events on stocks abnormal returns. Design/methodology/approach: This paper experiments with three different methods to estimate stocks abnormal returns, namely: market model, mean-adjusted model and market-adjusted model. The sample includes the firms listed in the leading index in Egypt stock exchange (EGX30). The statistical tests, Anderson-Darling test for normality, Wilcoxon rank-sum test for comparing the significance of the estimates, and Breusch-Pagan, Cook-Weisberg test for heterogeneity of abnormal returns. Findings: The results indicate that (a) statistical differences between the three estimates exist, which indicates that the three methods of abnormal return estimation are not substitutes, or alternatives, to each other, (b) that is, the political event is considered an anomaly which has idiosyncratic effects. This is contrary to the common belief that political events have systematic effects. Originality/value: The contribution of this paper is twofold. First, the estimation of abnormal returns must be examined for robustness in order to ensure reliability. Second, the results offer robust evidence that political risk premium is an anomaly, which is a call for stock market participants not to panic. Eventually, it saves investors' wealth.