Portfolio Optimizer
Design allocations with explicit objectives, constraints, and walk-forward context before saving them into portfolio research.
Portfolio Optimizer is where an allocation idea becomes a candidate set of weights under a chosen objective and constraint set.
This page covers the optimization methods available, what constraints you can set, and how to validate the resulting weights.
Start with one question
Constraint-aware candidate weights
Which allocation fits the return, risk, and holding limits in this setup?
Output previewCandidate weights | objective value | constraint and covariance inputs
Open Optimizer →
Backtest the candidate
Does the optimized allocation survive a historical walk-forward check?
Output previewWalk-forward metrics | realized path | comparison to benchmark
Run Portfolio Backtest →
The biggest risk is treating optimized weights as truth instead of as outputs of specific assumptions. Expected returns, covariance windows, and constraints all materially change the result.
Use the optimizer to generate a candidate allocation, then test that allocation in Portfolio Backtest under the same assumptions that matter to implementation.
Should I optimize directly on a short sample window?
Usually not without caution. Short windows make optimizer outputs more brittle, which is why walk-forward validation and explicit constraints matter.