Explore the
portfolio trade-off.
Adjust a small mean–variance model and inspect how its allocation changes. The datasets are illustrative assumptions; computation runs in your browser, independently of the PRISM engine.
Investment universe
Portfolio constraints
Higher values put more weight on reducing modeled variance.
The minimum cap depends on the number of assets.
What this example computes
Maximize wᵀμ − (λ/2)wᵀΣw subject to a net weight of 100% and the selected position bounds. A projected-gradient routine runs for up to 4,000 iterations. The short-enabled mode permits negative weights within the same absolute cap.
Returns, volatilities, and correlations are fixed illustrative assumptions. Taxes, transaction costs, market execution, and financing costs are excluded. This is an educational model, not investment advice or a PRISM benchmark.
Computed allocation
Compare the assumptions.
Three allocations under one illustrative scenario. Each 24-month path uses the portfolio’s modeled return and volatility with the same seeded shocks.
All three allocations use this page’s browser model. The return-leaning baseline uses λ = 1.4, long-only positions, and an 85% cap. Equal-weight assigns 1/N to each asset. A 4% assumed risk-free rate is used for Sharpe ratios. These simulated paths are not historical backtests or measured PRISM performance.
Evaluate your actual portfolio problem.
Production evaluation starts with your objective, constraints, reference, and quality criteria. Explore the published measurements or scope a representative portfolio workload.
See the benchmarks