Financial Analytics for Lending: NIM, Cost-to-Originate, and Loan-Level Profitability

Lending financial analytics isn't standard enterprise FP&A — it's loan-level math layered on funding costs, servicing costs, capital charges, and the CECL reserve that moves based on portfolio composition and macroeconomic scenarios. Net interest margin varies by product, vintage, and channel. Cost-to-originate differs by LO, channel, and loan complexity. MSR valuation depends on prepayment speeds that respond to rate movement and borrower behavior. Loan-level profitability requires fully-loaded cost allocation that few lenders do consistently. Each of these depends on methodology that generic financial analytics skips entirely, and the absence of consistent methodology means the lender is making pricing and channel investment decisions on numbers that change depending on who pulled them.

Lender financial analytics done right locks the methodology first and builds the analytics around it. Funding cost allocation to loans based on the actual funding source (warehouse line, retained, securitized). Servicing cost allocation per loan with the productivity assumptions the finance team has validated. Capital charge by product based on the risk weight and the target capital ratio. CECL reserve attribution at loan level. MSR valuation using the prepayment model the MSR team maintains. With these foundations, loan-level profitability is defensible and the pricing, channel, and product decisions rest on consistent numbers. Without them, every board meeting debates the methodology.