Optimization
Solving for the portfolio weights that maximize an objective (e.g., return or utility) subject to constraints (e.g., risk or budget).
At Level I the exam rarely asks you to solve an optimizer; it tests what optimization produces and why the output misbehaves. The classic “tell” is a question describing wildly different weights from a tiny change in inputs — the answer is input sensitivity (error maximization), and the fixes the curriculum credits are constraining weights, shrinkage, resampling, or Black–Litterman (which anchors estimates to equilibrium/CAPM reverse-optimized returns). Know the priority: expected returns drive far more instability than variance or covariance estimates, so estimation error there matters most.
The classic trap is conflating optimization with its outputs. The efficient frontier is the result of optimization, not the procedure; diversification (correlations below 1) is why the frontier bows and curves, not the objective function itself. Another trap: assuming “optimal” means well-diversified — unconstrained mean-variance often piles into a few assets, which is exactly why it’s distrusted. A hook: the optimizer maximizes errors, not just returns — it loves whatever asset your estimate happened to overstate.
PlayPrepHQ study notes are written and reviewed against primary exam sources. How we create & review content →