In an efficient market, better performers gain market share. If a lender denies fewer borrowers for the same profiles, more borrowers should route there over time. MBI measures whether this happens. MBI = 1 − |r|, where r is the correlation between lender performance and subsequent market share change. Near zero = efficient. Near 1 = the market doesn't react to quality.
FHA mortgage market 2023–2025: OLS regression of market share change on prior-year denial rate across 317 lenders yields slope +0.0013, r = 0.017, p = 0.326. The best two lenders carried a 30–40 point denial advantage for 8 years and gained only +3.7pp of combined market share. MBI = 0.883.
Borrowers don't choose lenders — they're assigned. By advertising budgets, broker relationships, servicer transfers, and geographic accident. The information exists in public federal data but has never been translated into a borrower-facing decision tool. Quality exists. Signal exists. The channel to the decision point doesn't.
Pending entries are hypotheses, not claims. Other markets will be added as verified measurements are completed.
A market with MBI near 0.883 doesn't self-correct. This is why Shadow Approvals persist year after year. This is why the 9+2 lender hierarchy doesn't converge. And this is the mechanism FRC Research exists to interrupt — by putting lender decision surfaces directly into the borrower's moment of choice.