The reported P/B of 20.68 is not a premium in any comparable sense — it is an arithmetic artifact of roughly $6.9B of goodwill and intangibles against a $3.3B equity base, so any attempt to benchmark it against SPGI's 3.80x or ICE's 3.21x is a category error. The forward P/E of 28.82 against a trailing 35.37 embeds an assumption of continued issuance recovery, which appears reasonable given the 2026Q2 EPS peak of $5.02 and the refinancing wall management has flagged, but the PEG of 4.53 suggests the market is already paying handsomely for that trajectory. A blended SOTP framing — SPGI-like multiples for MIS and MSCI-like multiples for the Analytics recurring book — remains the more defensible lens, and on that basis the 0.8% dividend yield signals the market prices buyback-driven EPS compounding, not income, as the return mechanism.

A DuPont decomposition shows the classic non-bank structure: ROA of 5.0-6.0% is the genuine operating engine, driven by 102.7% fee income contribution and a 68%+ gross margin on incremental ratings, while the 4.3x equity/assets leverage multiplier does the rest. That distinction matters because ROA itself is unstable — it ranged from 2.5% in 2024Q4 to 6.0% in 2026Q2 — tracking issuance volumes rather than execution quality. The quality of ROE therefore rests on whether MA's subscription base can lift the structural floor of that ROA during issuance troughs; surveillance fees and analytics retention are the metrics that would validate the current 22-28% ROE band as sustainable rather than cyclical.

The negative NIM is entirely consistent with the balance sheet findings — Moody's carries borrowings whose cost exceeds the yield on an investment portfolio that has been drawn down 91% to just $108M, so the spread structure will likely stay adverse as the earnings-asset base shrinks further under continued repurchases. On the cost side, the efficiency ratio's ~660bp expansion over seven quarters is the more actionable signal; if expense growth is now outpacing revenue growth (operating income slipped to $906M sequentially), it suggests either investment build in Analytics or integration costs that warrant confirmation against management's cost-savings commentary. Analysts should watch whether the 25-31% efficiency band reverts toward the historical mid-20s or represents a structurally higher cost base.

Traditional CET1 and Tier 1 adequacy frameworks do not apply here — Moody's is not a deposit-taking institution subject to Basel capital floors — so the operative constraint is management's own comfort band on equity/assets, and the trajectory implies that band is being deliberately compressed. The concern is not solvency but optionality: with cash and bank balances steady at $1.5B but the liquid securities buffer effectively exhausted, the marginal buyback dollar increasingly depends on operating cash flow, which itself swung between 0.89x and 1.71x conversion of net income. A further meaningful drawdown would leave less cushion for an issuance downturn or for funding the RMS integration, so investors should monitor whether the equity ratio stabilizes near 0.22 or continues to grind lower.