Is There a Measurable Placebo Effect in Policy Announcements?

This paper tests whether policy announcements generate a distinct 'placebo' effect: an incremental same-day reaction beyond what is explained by the news sign (positive or negative) and the time remaining to implementation. A multi-country event panel with at least 20 positive and 20 negative announcements tracks close-to-close equity index returns (basis points) and 10-year sovereign yield changes (basis points). The method converts lead time to a behavioral salience weight with an exponential half-life (baseline h=7), estimates a baseline relation Delta M = alpha + beta*(ShockExp*s), and compares residuals for anticipated (long lead) versus unanticipated (short lead) events. Equity responses exhibit a steep baseline slope, yield responses are modest, and unanticipated events show more negative residuals than anticipated events. Results are robust to alternative half-lives, anticipation thresholds, and outlier checks.