NMI Holdings' Strong IIF Growth: Can It Drive Future Earnings?

NMI Holdings' Q2 2026 primary insurance in force rose 5.8% year over year to $227.1 billion, while new insurance written jumped 29% to $16.1 billion, expanding its future premium base. Net premiums earned increased 5.7% to $157.5 million, adjusted EPS grew 14% to $1.38, and the loss ratio improved 70bps to 8.3%. Consensus projects 2026 EPS and revenue growth of 6.7% and 6.5%, respectively, with estimates for both 2026 and 2027 up 0.4% over the past month; NMIH shares have gained 14.9% over three months and trade at 1.23x book versus the industry's 1.45x.

Analysis

NMIH's faster policy growth matters less as a near-term revenue surprise than as evidence of share capture in the purchase-mortgage channel. New business typically earns into premiums with a lag, so the key 1-3 month catalyst is whether management raises its in-force/premium run-rate outlook rather than simply reporting another strong origination quarter. If sustained, a modest rerating toward the private-MI peer price-to-book range is plausible over 6-12 months, provided persistency remains intact and pricing has not been sacrificed to win volume.

The more important competitive question is risk selection. NMIH can grow faster than MTG and RDN because of distribution gains, but rapid expansion late in a housing cycle can increase exposure to higher-LTV, thinner-FICO cohorts that produce losses only after unemployment rises. Current loss experience is backward-looking: mortgage insurers face a convex downside if home-price appreciation reverses and delinquency cures weaken simultaneously. Watch quarterly risk-in-force composition, average FICO/LTV, persistency, and incurred-but-not-reported reserve development; these are more decision-useful than NIW alone.

Consensus may be underestimating the operating leverage from a lower-rate refinance revival: lower rates can initially reduce earned premiums through cancellations, but subsequently expand originations and create a larger addressable low-down-payment borrower base. That ambiguity argues against a broad housing-beta trade. NMIH is the cleaner relative winner if share gains persist, while MTG and RDN offer more mature books and potentially less underwriting-tail sensitivity; a weakening labor market would reverse the relative thesis quickly.

Key Decisions for Investors

• Initiate a 6-12 month long NMIH / short MTG pair, sized beta-neutral, only if the next filing confirms stable or improving average FICO/LTV and no adverse reserve development. Target relative rerating of roughly 10-15%; exit if NMIH's NIW growth falls below peer growth for two quarters or loss-ratio guidance rises materially.
• Do not chase NMIH after the recent momentum; accumulate on a 5-8% pullback or following earnings if management confirms premium growth acceleration. Base case is multiple expansion plus mid-single-digit earnings growth; principal risk is that volume was purchased through lower pricing or weaker credit selection.
• Set a macro risk trigger on unemployment and national home prices: a sustained rise in unemployment toward 5% or a mid-single-digit national home-price decline should prompt reducing private-MI exposure, particularly NMIH. Those conditions can drive loss ratios disproportionately higher over the following 6-18 months.
• Use MTG and RDN as confirmation rather than substitutes: broad NIW acceleration with stable credit characteristics supports a mortgage-origination tailwind, whereas NMIH-only growth points to distribution share gains and raises the need to verify unit economics and book quality.