Aflac heads into Q2 with the stock near all-time highs. AFL closed at $125.59 on Aug 5, up roughly +14% year-to-date (from ~$110 in early January) and up about +10% since the Q1 report on April 30. Shares touched ~$129.55 in late July before easing slightly into the print. That run-up matters: expectations are elevated, and this is a company whose "beat" magnitudes are typically small, so the bar for a positive reaction is high.
Consensus for Q2 2026 sits at roughly $1.77 adjusted EPS on ~$4.19B revenue, essentially flat versus the $1.78 adjusted EPS a year ago (a ~0.6% decline). Full-year 2026 consensus is around $7.07, down ~5.6% from 2025, with a rebound expected in 2027. Note Aflac's recent track record: it has beaten the EPS consensus in three of the last four quarters, but Q1 2026 was a miss — $1.75 adjusted vs. ~$1.80 expected (a ~2.8% negative surprise) — so investors will be watching whether the underlying trends firm up.
A reminder on the revenue optics: the large reported YoY revenue "decline" is mostly a function of net investment gain/loss swings (derivatives/FX) that wash out of adjusted earnings. Focus on adjusted EPS, currency-neutral segment trends, benefit ratios, and capital return rather than the GAAP headline.
This is the central Aflac debate. The marketing/sales transformation and new products have driven a genuine sales surge — Q1 2026 Japan sales rose 25.5% (¥17.7B), led by the new medical product Anshin Palette, cancer product Miraito, and first-sector Tsumitasu. A year ago in Q2 2025, sales were up 23.2%.
But underlying earned premium is still shrinking (down ~1.3% in Q1), because lapses and policies reaching paid-up status are running ahead of new premium. On the Q1 call, CFO Max Broden framed the math clearly: Japan needs to reach roughly ¥90B of annual sales to hold in-force premium flat, versus ~¥74B booked in 2025. CEO Dan Amos said he'd "like ¥80 billion" for 2026 — still short of the breakeven level. Management guides to underlying earned premium of roughly -1% to -2% for full-year 2026.
Watch: the Q2 sales figure (against a tough ¥20.7B comp), and any updated commentary on when the earned-premium inflection to flat/positive actually occurs.
Japan has been the earnings engine — Q1 pretax margin hit 35.0% (up 320 bps YoY) with a benefit ratio of 62.9%, helped by favorable cancer/hospitalization experience and reserve remeasurement gains (~70 bps above plan). Management reiterated a full-year Japan benefit ratio range of 60–63%. Q1 ran near the high end excluding favorable items, so watch whether the favorable claims trend persists.
Key structural item: management flagged a long-term assumption unlock in Q3 2026 (net premium ratio reset). Any preview of that on the Q2 call could move the stock.
The U.S. story is a tale of two channels: - Group business is strong — group VB, dental/vision, and group life/absence/disability were up ~12% in Q1, and the "buy-to-bills" (dental/vision + group L&D + consumer/direct) were up ~25%. Dental & vision, a prior problem area, rose 52%. - The core individual/agent business is flat-to-down. Overall U.S. sales rose only 2.9% in Q1 ($318M) because the traditional agent channel isn't growing. Recruiting remains tough; management targets ~10–11K new agents for the year and is investing in conversion/productivity.
Persistency held at a solid 79.3%, and net earned premium grew 3.5%. U.S. benefit ratio was 47.2% in Q1, well below the full-year guide of 48–52%, boosted by favorable cancer and group disability claims (~80 bps above plan) — management explicitly cautioned that group disability is volatile quarter-to-quarter and expects normalization. So a higher U.S. benefit ratio in Q2 would not necessarily be a negative surprise. Also watch for any follow-up on the one-state regulatory rate-cut headline from earlier in the year, which management characterized as immaterial.
Capital return remains the core of the AFL thesis. In Q1 the company returned $1.3B ($1.0B buyback + $315M dividends). The Board raised the quarterly dividend 5.2% to $0.61 and declared the same for Q2 — extending a streak now at 43 consecutive years of increases. Capital ratios are robust: ESR ~227% (243% with USP), estimated combined RBC ~560%, leverage 21.2% (within the 20–25% target), and $3.4B of holdco liquidity ($2.4B above minimum). Watch the Q2 buyback pace (Q2 2025 was $829M) and remaining authorization.
A newer strategic angle: at end-Q1, Aflac Re Bermuda assumed a block of whole-life annuities from Japan Post — the first external Japan reinsurance deal (following internal deals). It's immaterial today (mid-single-digit-million near-term drag to Japan earnings that fades to ~zero over time), but management framed it as a platform that could become "material over time" for deploying excess capital without cutting shareholder returns. Any sizing of the pipeline is a potential positive catalyst.
Aflac is a "quality compounder" story — strong margins, fortress capital, and relentless capital return — trading near highs into the print. The base case is another in-line-ish quarter (~$1.77 EPS) with continued Japan margin strength, robust U.S. group sales, and $1B+ of capital returned.
The debates that will drive the reaction: 1. Does strong Japan sales momentum show any sign of translating into earned-premium stabilization? 2. Do the favorable reserve/claims tailwinds in both segments persist, or do benefit ratios normalize toward guidance (US 48–52%, Japan 60–63%)? 3. Any early read on the Q3 assumption unlock and the scale of the external reinsurance opportunity? 4. Cyber and CRE — any negative surprises?
With the stock priced for continued execution and Q1 having been a small miss, the risk/reward into the print is asymmetric: a clean beat is largely expected, while a benefit-ratio normalization, an FX drag, or cyber-related noise could pressure a stock that has already run ~14% YTD.
(Sources: Aflac Q1 2026 earnings release and conference call transcript, Q2 2025 earnings release, publicly reported analyst consensus, and Aflac SEC risk-factor disclosures. Figures are company-reported; consensus estimates are third-party and approximate.)