Central banks have long engaged in risk management, adjusting policy preemptively when tail risks rise. This paper sets out a formal framework for doing so. Using options data, I show empirically that tail-risk shocks act as negative aggregate demand shocks, with output and inflation falling together. I then develop a New Keynesian macro-finance framework in which all higher-order moments of the shock distribution enter asset prices, aggregate demand, and the neutral real rate. Tail risks are quantitatively important for the neutral rate: a one percentage-point increase in disaster probability lowers it by approximately 16 basis points. For policy implications, I find large welfare gains from central banks actively adjusting policy in response to the full shape of the shock distribution, not just its mean and variance. I then construct a counterfactual for the US during the Global Financial Crisis, finding that ignoring tail risk would have more than doubled the output gap decline in 2007Q3–Q4. Finally, I provide a new measure of the neutral rate of interest adjusted for risk. This risk-adjusted neutral rate documents that policy has been overly restrictive during periods of elevated tail risk, with sharp crisis-driven contractions absent from standard estimates.