Investors demand reward for taking risk. Concepts of Knightian risk and uncertainty; risk preference (risk-neutral Q vs. real-world P probability measures); coherent risk; and commonly used metrics for risk are explored. The integration of risk and reward in classical efficient portfolio construction is described, along with the drawbacks of this approach in practice and methods for addressing these drawbacks. The leptokurtic (fat-tailed) nature of financial data and approaches to modeling financial surprises are covered, leading to inherently leptokurtic techniques for estimating volatility and correlation. Scenario analysis, and regime-switching methods are shown to provide ways of dealing with risk in extreme environments. The special nature of modeling long/short portfolios (hedge funds) is explored. The text for the class is a Jupyter Notebook with Python code segments.