Type a ticker to see what a fund actually holds, judged only by its returns: exposures to ten liquid ETFs, style tilts, how closely a passive clone tracks it, and whether the fee bought anything a clone could not. Forty-five funds precomputed, including every public Goldman Sachs QIS product.
Method. Monthly total returns (Yahoo Finance) since inception or Jul 2008. Exposures: OLS of excess returns on ten ETFs (SPY, IWM, EFA, EEM, IEF, LQD, HYG, DBC, GLD, UUP) with HAC standard errors; time-varying exposures from a random-walk Kalman filter; style tilts from Fama-French 5 + momentum; the ETF clone re-estimates betas on the trailing 24 months and holds them the next month, so its returns are out of sample. Data generated .
Limits. Returns-based analysis infers holdings statistically and cannot see them; R² tells you how much it is missing. Short histories (under five years) give noisy loadings. Nothing here is investment advice or a recommendation, and the author has no affiliation with any fund sponsor. Deeper write-ups: a multi-factor strategy built from scratch and a two-part review of Goldman Sachs QIS products. Code.