About

RiskModels decomposes an equity portfolio into the layers that actually drive its returns and risk — market, sector, subsector, residual — and lets you see what you're betting on with the same precision an institutional risk desk would demand.

Every position in a portfolio carries four kinds of exposure simultaneously. Some is broad-market beta you could replicate with SPY. Some is sector exposure — the part of NVDA that moves with semis, or AAPL that moves with consumer-tech. Some is subsector — narrower still. The rest is idiosyncratic — the stock-specific component a stock-picking mandate is actually compensated for.

Most retail and even most institutional reporting flattens these into one number. RiskModels keeps them separate. The output is an F1 Tearsheet for a portfolio, a DD sheet for a single name, or a Risk DNA cohort comparison — each ground-truthing what's *systematic* and what's *yours*. From there, an analyst can tell you which positions deserve a hedge, which are stock picks, and which are sector tilts in disguise.

The system is built around a 3-level hierarchical equity risk model (ERM3) over the full US equity universe, refreshed daily. The same engine answers questions in the workspace, computes the API responses at riskmodels.app, and produces the F1 / DD / Risk DNA institutional artifacts. No drift between the UI and the data layer.

Where to start