CEFI

CEFI (Constrained Efficient Frontier Index) runs constrained mean-variance optimization across 13 ETFs spanning developed equities, bonds, emerging markets, commodities, and inflation-protected assets. Asset-class caps (e.g. max 50% equities, max 50% bonds, max 25% alternatives) enforce diversification.

The optimizer uses a 6-month rolling covariance matrix with shrinkage estimation to reduce noise in the correlation structure. Weights are rounded to 5% increments. Volatility is managed by targeting a specific annual level through the optimization constraints.

Rebalancing occurs on the last trading day of each period. The strategy differs from a static 60/40 allocation by shifting weights based on recent return and risk data, but remains fully systematic with no discretionary inputs.

At each rebalance, the optimizer calculates expected returns (6-month average) and the covariance matrix (6-month, Ledoit-Wolf shrinkage) for the 13-ETF universe. It then solves for the maximum-return portfolio at the target volatility level, subject to the asset-class and position-size constraints.

If no feasible portfolio exists at the target volatility, the target is increased by 1% iteratively, or the minimum-variance portfolio is selected as a fallback. All weights must be multiples of 5%.

The benchmark is a static 60/40 portfolio (SPY/AGG), rebalanced at the same frequency. All dashboard metrics update dynamically based on the selected start date and initial capital.