In this section I first report the warning model estimation. I next present the test results in the short- and long-term windows when self-selection is not controlled for. I then report my primary test results—the tests when self-selection is controlled for. I briefly note the robustness tests concerning sub samples and model specifications. Finally, I examine post-event returns and evaluate trading strategies to provide supplementary evidence for the primary results. Figure 4 summarizes the key results