Financial Shenanigans

The Forensic Verdict

MetLife screens Watch (38/100) — not because any single line on the financial statements is broken, but because three structural features deserve underwriting before sizing a position: (i) a long history of regulatory and internal-control settlements (1995-2019) that the company itself acknowledges in its FY2025 risk factors, (ii) layered non-GAAP framing where management's headline metric ("adjusted earnings excluding total notable items") was redefined twice in two years to remove items that would have hurt the print, and (iii) a Bermuda Class E sidecar (Chariot Re) launched July 2025 in which MetLife retains a 15% economic interest, exclusive asset-management fees through MIM, and is reinsuring liabilities originated in its own pension risk transfer engine. None of these is a smoking gun. The income statement reconciles, GAAP cash from operations is genuinely strong (5-yr CFO/NI = 3.5x, sector-normal for a life insurer), the auditor (Deloitte & Touche) has clean opinions and only 20% of fees are non-audit, goodwill was not impaired, and there is no open SEC investigation. The single data point that would most change the grade is Talcott or Chariot Re structure disclosure: confirmation that funds-withheld reinsurance with the Bermuda affiliate uses arm's-length pricing and is not used to mask deteriorating reserve adequacy in the run-off variable-annuity / long-term-care book.

13-Shenanigan Scorecard

Breeding Ground

The structural risk profile is mixed. The board is unusually clean by US insurer standards — 10 of 11 directors are independent, the chair is separate from the CEO, the audit committee includes two former insurance partners (KPMG/EY), and there is no founder, family, or controlling-shareholder dominance. Auditor concentration risk is low: Deloitte audit fees are $53.3M for 2025 and non-audit fees are 20% of total ($13.0M of $66.3M), well within independence guidelines. Compensation is heavily equity-weighted (CEO Khalaf's 2025 total of $22.4M is 66% stock awards) and tied to adjusted earnings, FCF ratio, and direct expense ratio — metrics whose definitions management controls.

What unbalances the picture is MetLife's enforcement history. In the past decade the company or affiliates have paid: $10M to the SEC (2019, internal-control failures, "MetLife overstated its reserves and understated income related to its variable" annuities, per the SEC's New York office), $19.75M plus $189M in restitution to NYDFS (2018, escheatment / Death Master File misuse), $25M to FINRA (record fine for misleading variable-annuity customers), $123.5M to DOJ (MetLife Home Loans mortgage fraud, 2008-2012 underwriting), $13.5M settlement with FBI (improper foreign payments), $32.5M (employment discrimination class), $40M (life insurance beneficiary identification), and $84M (variable-annuity class settled 2020). The FY2025 risk factors include unusually frank language: "the Company's internal controls have in the past proved, and there is a risk that they may in the future prove, to be deficient or ineffective" — a direct acknowledgement.

The breeding ground does not strongly tilt this case to "elevated" — every flag has a sector-normal explanation, and the board/auditor structure is genuinely robust. But the cumulative regulatory record and the company's own admission that controls have failed in the past is reason to weight earnings-quality findings more skeptically than at a peer with a clean record.

Earnings Quality

Reported earnings reconcile cleanly to the cash-flow statement and balance sheet over a multi-year window, but management's "adjusted earnings" diverges meaningfully from GAAP and the divergence is widening. GAAP net income has been on a downward trajectory since FY2021 ($6.65B → $5.10B → $1.38B → $4.23B → $3.17B), while management's preferred number — "adjusted earnings excluding total notable items" — is presented as smoothly growing ($5.5B → $5.8B → $5.9B over 3 years). The mechanism is the carve-out of net investment gains/losses, net derivative gains/losses, MRB remeasurement, and "notable items" the company classifies as non-recurring. Most are technically defensible under the LDTI regime, but the gap is large and growing.

The wedge between GAAP and adjusted earnings has averaged $1.9B annually but widened to $2.8B in 2025. The dispersion is driven primarily by remeasurement of market-risk benefits (MRBs) and derivative hedging losses. Reading the FY2025 reconciliation: net investment losses of $1,145M, net derivative losses of $1,939M, MRB remeasurement gains of $508M, and policyholder liability remeasurement gains of $150M are all stripped from GAAP to reach the higher non-GAAP number. The hedging losses recur every year, which makes the recurring-as-non-recurring criticism legitimate.

Other earnings-quality tests come back clean. Goodwill of $9.6B was tested in Q3 2025 and "estimated fair values of all such reporting units were substantially in excess of their carrying values." DAC deferral and amortization look proportionate. The ratio of EPS (boosted by buybacks) to comprehensive income is sector-normal. The one item that warrants close watch is commercial real estate: the mortgage allowance for credit losses rose 75% in one year, from $461M at year-end 2024 to $807M at year-end 2025, and direct real-estate impairments rose 5x ($36M → $190M). With $42.4B in commercial mortgage loans of which 38.9% is office and 7.5% has DSCR under 1.0x, this is the most plausible accelerator of earnings deterioration in 2026.

Cash Flow Quality

Operating cash flow is the cleanest part of MetLife's financial statements — but only if you understand what it represents in life-insurance accounting. GAAP CFO of $17.1B in 2025 (up from $13.0B in 2022, a five-year CAGR of 8.0%) is the sum of premium and policyholder-deposit inflows minus benefits paid, plus net investment income and policy-fee revenue. It is not directly comparable to industrial-company CFO. The forensic test that matters here is whether the upward trend reflects genuine business acceleration or a one-time shift in funding-agreement issuance.

CFO/NI ran 1.86x to 2.56x from 2017 to 2022 — typical insurance-industry levels. From 2023 onward the ratio jumped (9.9x in 2023, 3.5x in 2024, 5.4x in 2025) because GAAP NI compressed under LDTI volatility while the underlying premium engine kept running. The forensic interpretation is that this is a denominator effect, not a CFO inflation. CFO did grow $4.4B in 2025 — but disclosure shows the increase came primarily from a $3.7B (43%) jump in adjusted premiums, fees, and other revenues in the Retirement & Income Solutions segment, driven by pension risk transfer and U.K. longevity reinsurance flows. PAB net flows accelerated from $2.96B to $6.85B (+131%), so a portion of the CFO uplift is institutional-product deposits — economically equivalent to a financing inflow that GAAP classifies as operating because of how insurance liabilities are categorized.

The capital-return story is also worth flagging. Over the past five years MetLife distributed $24.6B to shareholders ($16.8B buybacks + $7.8B dividends) against $20.5B of cumulative GAAP net income — a payout ratio above 100% sustained by debt-funded financing and balance-sheet capacity. Free cash flow ratio at the holding-company level was 81% in 2025, above management's stated 65-75% target — flagged in the proxy as a "core" metric beat, but indicating capital is being returned faster than holding-company free cash flow can durably support.

Metric Hygiene

This is where MetLife earns the bulk of its forensic risk grade. Management's headline numbers are presented in a pyramid of non-GAAP layers: GAAP net income → adjusted earnings → adjusted earnings excluding notable items → core adjusted earnings (excluding both notable items and PRT-related expenses) → constant-currency core adjusted earnings. Each layer adds a definition that is reasonable in isolation but cumulatively gives management substantial discretion over the headline number. The forensic test is whether definitions are stable. They are not.

The other structure to watch is affiliated and quasi-affiliated reinsurance. Three vehicles concentrate risk transfer: (i) MetLife's own captive reinsurers in Vermont, South Carolina, Bermuda, and the Cayman Islands (disclosed in Item 7 of the 10-K and in Risk Factors); (ii) the November 2023 universal-life reinsurance transaction, which the FY2024 MD&A confirms reduced MetLife Holdings adjusted earnings by $170M but does not name the counterparty; (iii) the new Chariot Re sidecar launched July 2025. The Chariot Re structure is unusual and worth explicit forensic underwriting: MetLife and General Atlantic each retain ~15% equity, Chubb and other investors hold the remainder, MetLife Investment Management has an exclusive asset-management mandate on ~$10B of liabilities (structured settlements and PRT-linked group annuities), and the Bermuda Class E framework allows MetLife to release statutory capital while retaining economic upside through the equity stake and IMA fees. This is sector-normal in 2025 but it is a related-party-adjacent transaction that warrants ongoing fee, ceding-commission, and risk-retention disclosure.

What to Underwrite Next

The highest-value forensic items to track in the next four quarters are concentrated, specific, and tied to disclosure that already exists in the 10-K and press releases. None are speculative — each has a defined data point that will resolve the underwriting question.

Bottom line for the investor. This is not a thesis-breaker. It is a position-sizing and disclosure-monitoring case. The accounting risk premium that should attach to MetLife is small — call it 100-200 bps on cost of equity — and is almost entirely concentrated in three things: the credibility of the adjusted-earnings definition going forward, the arm's-length pricing of the Chariot Re relationship, and the speed of CRE reserve build. If all three resolve favorably over the next four quarters, the forensic grade drops back to Clean. If a third Q4 metric redefinition appears, or if the November 2023 reinsurance counterparty turns out to be an undisclosed affiliate of an investment partner, or if the CRE allowance accelerates further while management points to "model recalibration," the grade moves to Elevated and the appropriate response is reduced position size and explicit reserve-adequacy diligence on MetLife Holdings.