Business

Know the Business — AgeSA Hayat ve Emeklilik

Bottom line. AgeSA is a Turkish life-and-pensions compounder whose real engine is not premium underwriting — it is a ₺402bn private-pension asset book that earns steady fee income, plus a capital-light bancassurance channel through Akbank. Management-reporting ROE stayed at 63–64% through the Turkish disinflation cycle, but the market tends to treat AGESA like a cyclical insurer and misses two things: roughly 40% of technical income is credit-linked life sold with Akbank loans (highly scalable, low capital), and expenses are falling twice as fast as premiums (expense ratio 48.5% → 39.2% in five years). The right watch-list is pension AuM, credit-life market share, and the expense ratio — not combined ratio, which does not meaningfully exist in this business.

1. How This Business Actually Works

AgeSA does not earn money the way a traditional P&C insurer does. It has three stacked profit engines that share one distribution asset: the Akbank bancassurance channel plus an 890-person direct-sales force.

The economic engine in one paragraph. Credit-linked life insurance is sold alongside Akbank consumer and commercial loans — short duration, very low capital, and the loan contract does the distribution work for free. Private Pension (BES) earns a management fee on AuM; the 30% state contribution is the customer-acquisition hook, and stickiness is enforced by a 10-year vesting rule, so once a participant is in, the fee annuity is extremely durable. Return-of-Premium & Savings is the discretionary savings product — bigger ticket, longer duration, and the real beneficiary of Solvency II VNB growth. The G&A base (₺6.3bn in FY25; sales personnel 34%, HO personnel 33%, IT 12%) is largely fixed, so incremental premium drops through at high margin once the expense ratio crosses the ~40% threshold — which it just did.

Why returns are structurally high. AgeSA carries no motor, no catastrophe, no large-claim tail. Life claims are mortality-driven and well modeled; credit-life claims track unemployment and are re-priced annually. Balance-sheet risk is asset-side (TRY government bonds and deposits in the shareholders' portfolio; customer-owned pension funds are off-balance-sheet) rather than liability-side. That is why the business prints 60%+ ROE at 200%+ solvency — the capital intensity is less than a quarter of what a diversified non-life insurer needs for the same revenue.

2. The Playing Field

AgeSA sits #1 among private-sector companies in both pension AuM and total Life & PA GWP. The BIST peer set is mostly non-life carriers; the closest true comparable is Anadolu Hayat (ANHYT), a similar life + pension JV tied to Türkiye İş Bankası. Türkiye Sigorta (TURSG) is the state-owned non-life giant; Aksigorta (AKGRT) is the sister Sabancı/Ageas P&C carrier that now shares a CEO; Anadolu Sigorta (ANSGR) and Ray Sigorta (RAYSG) are pure non-life. Allianz Yaşam and Garanti Emeklilik are unlisted but are the only real pension-AuM threats.

What the table reveals. Among listed BIST insurers, AGESA combines the highest ROE with the lowest capital intensity — the signature of a fee-and-distribution business, not an underwriter. ANHYT has similar structural economics but runs a larger share of traditional savings life and less bancassurance-originated credit life; its ROE is lower because it does not lean as hard on the bank's loan flow. TURSG, ANSGR, AKGRT and RAYSG each face motor and catastrophe loss ratios that clip ROE into the 30–45% band even in a good year. The unlisted peers that actually matter — Allianz Yaşam ve Emeklilik and Garanti Emeklilik — are bancassurance twins tied to partner banks and are the only firms with a structural shot at AgeSA's #1 spot. Both have taken that top spot at different points in the past decade when the partner bank out-grew Akbank.

AgeSA is the outlier in the upper-right: the highest ROE despite the highest expense ratio, because the denominator of ROE (equity) is unusually small for a business of this cash-flow quality. Cutting the expense ratio further — the stated 2025/26 plan — is pure operating leverage.

3. Is This Business Cyclical?

Yes, but not in the way most insurance businesses are. AGESA has three distinct cycles operating at different frequencies, and only one of them is the classic underwriting cycle.

What the last cycle actually showed. The 2021–2024 disinflation environment was the acid test. TRY inflation peaked above 80% in late 2022 and settled near 44% by end-2024; the policy rate moved from 14% to 50%. Three things happened simultaneously: (i) the shareholders' investment portfolio re-priced to double-digit real yields, driving net investment income from ₺200M in FY21 to ₺3.1bn in FY25 — a 15x jump that masked operating volatility elsewhere; (ii) credit-linked life GWP grew at 99% CAGR 2021-24 because nominal loan growth and re-pricing compounded; (iii) RoP & Savings took the longest to respond because customers only rotate savings after real yields stay positive for several quarters.

The gap between Management Reporting and SFRS is the TAS 29 inflation-accounting effect — a roughly 25–30% haircut since it was applied from FY2023. In SFRS terms, FY25 net profit was ₺5.15bn, not ₺6.84bn; dividend capacity is set off SFRS. Read the growth trajectory on either basis and the conclusion is the same: roughly 18x net-profit growth in five years.

The part that is cyclical. Pension AuM growth is a leveraged bet on Turkish household savings and real returns on BIST. When fund performance lags inflation (as in 2022), net inflows slow and fee income compresses. When real returns go positive (2024-25), AuM compounds at 60–75% per year. That is why the market correctly treats AGESA as a TRY-rates story layered on top of a fee business — the biggest swing factor in any given year is shareholders' investment income, not premium.

The part that is not. Credit-linked life and pension management fees are structurally insulated. Even in the worst year on record (FY2020, COVID), GWP still grew and AuM still grew. This is a long-duration secular story — BES participation is 19.4M of a working-age population of roughly 32M, so the penetration ceiling is a long way off.

4. The Metrics That Actually Matter

Ignore combined ratio here. Ignore price-to-book as a lead indicator. The five numbers that explain value creation at AGESA, ranked by relevance:

Why these, not the usual ratios. Loss ratio is close to useless because life claims are modeled, not reported as a spike. Combined ratio is not even reported — AGESA's G&A / technical-profit framing is the right one. P/B can mislead because reported equity is a SFRS number distorted by inflation accounting — a P/B of 2.0x on SFRS book can mean very different things year to year. What actually drives the stock is AuM (the durable annuity metric) and the expense ratio (where the operating leverage is compounding).

The VNB chart is how to think about forward value: every ₺100 of premium written today creates roughly ₺6 of shareholder value after best-estimate claims, expenses, and a risk margin. In FY25, new business alone added ₺8.8bn of value — more than the company's entire SFRS earnings that year. That is the KPI for a life insurer; treat it the way you treat originated-loan NPV at a bank.

5. What I'd Tell a Young Analyst

Three practical rules and one warning.

Track the Akbank pipeline, not just AGESA's releases. Roughly 72–73% of life GWP and the majority of PPS new-business value flows through Akbank. When Akbank's consumer loan growth inflects — good or bad — AGESA's credit-life line inflects four to six weeks later. The same CEO (Fırat Kuruca) now runs AGESA and Aksigorta, and the Bancassurance head covers both; this is an explicit strategy to push Sabancı financial-services distribution through one channel. Watch Akbank quarterly loan data like it is AGESA's own.

The durable moat is the state contribution plus 10-year vesting on BES, not brand. The 30% state match and the forfeiture-on-early-exit structure is the reason pension AuM retention is roughly 95% over a rolling three-year window. No Turkish insurer has a product moat in life itself — the moat is in the pension regulatory architecture and in having one of the largest private retail banks as a captive shelf.

Value the business on distributable earnings, not reported net income. Dividend payout has been run at about 35% on SFRS, and the FY25 dividend of ₺1bn was a 3.3x jump year-on-year. With Solvency II at 204% and core ROE durably in the 60s, management has both the capital and the regulatory headroom to keep stepping the payout up. A shift in dividend policy is the most under-discussed upside here.

The warning. Medisa, the new 100% health-insurance subsidiary (₺1.4bn cumulative investment through 2025), is strategically sensible but will dilute reported ROE for two to three years — new health businesses in Turkey burn cash for at least that long. Management has started reporting consolidated vs. unconsolidated figures for a reason; if analysts use the blended ROE, the number will look like it is deteriorating even while the core life/pensions engine is still compounding at 60%+. Model the two separately.