We investigate whether consistent non-GAAP reporting is associated with investment efficiency. Prior research finds a positive association between non-GAAP reporting and investment levels, concluding that it represents overinvestment. We corroborate this positive association, but additional tests are not consistent with the conclusion of inefficient overinvestment. Specifically, we explore the relation between investment and future cash flows as a proxy for the realization of investments in positive net present value projects. We find that the investments of firms that consistently report non-GAAP metrics are associated with similar or higher future cash flows than the investments of firms reporting only GAAP earnings, which is consistent with efficient investment. We observe similar associations in multiple specifications, performance horizons, and outcome variables, including future returns and earnings. Given the prevalence of non-GAAP reporting and the SEC's ongoing concern with the consequences of non-GAAP disclosure, our analyses offer timely evidence relevant to this important discussion.

Boards of directors encourage risk-averse managers to take risky actions by providing stock options and severance pay. We demonstrate that the ability of these incentives to encourage risk-taking hinges on the level of uncertainty facing the manager. We confirm prior findings that stock option convexity encourages risk-taking but find that this relation only holds when market-wide uncertainty is low. We also confirm prior findings that severance pay encourages risk-taking but find that this relation only holds during high market-wide uncertainty and negative market-wide performance. Finally, we find that compensation committees respond to variation in uncertainty by adjusting the level of option grants. Our results suggest that the effectiveness of incentives to take risk varies with the market-wide uncertainty, and that boards consider this in annual compensation design. Data Availability: Data are available from the public sources cited in the text.