We modify the work-horse pay-for-performance sensitivity design (where annual changes in CEO compensation are regressed on annual performance measures) by distinguishing between stock returns on information event days, when stock returns better reflect CEO effort, versus those on “regular” (non-event) days. We find that the sensitivity of annual CEO pay to stock returns is 74% larger on information event days versus “regular” days. We distill information events further based on accounting versus non-accounting information events and find that the sensitivity is 80% larger on accounting days than other event days. These findings are distinct from RPE or pay-for-luck. Consistent with our methodology offering a parsimonious way to assess the efficiency of compensation contracts, pay-information return-sensitivities are stronger in firms focused on shareholder value maximization. Our study uncovers variations in pay-for-performance sensitivity based on managerial effort that are not documented previously and not readily observable from explicit contracts.