Unipol Assicurazioni S.p.A. 0Q6M

Deep-value analysis — 0Q6M · Unipol Assicurazioni S.p.A. (ISIN IT0004810054)

Analyst: Claude (deep-value agent) · Date: 27.08.2026 · Reference price: EUR … (closing price of the 0Q6M.IL line on 27.08.2026) · Market cap EUR 20.15 bn · 717,206,043 shares · Primary listing: Borsa Italiana (UNI/UNPI, EUR 28.12 on the same date)

---

IDENTIFICATION AND PRICE WARNING — read before any figure

0Q6M is not a distinct company. It is a secondary listing line of Unipol Assicurazioni S.p.A., the same company the tracker also holds a second time under the ticker UNI (sheet Sumar, r332, Exchange=MIL) and which I analyzed on 24.08.2026 in rapoarte/deep/2026-08-24-deep-UNI.md. It is not a homonym collision: it is the same issuer, same ISIN, two columns in the tracker.

The serious problem isn't the duplicate, it's the price. Yahoo publishes two symbols for the foreign line:

Symbol	Market	history()	Price served by quoteSummary	Volume
0Q6M.L	LSE	zero bars — "possibly delisted; no price data found"	EUR 3.839	—
0Q6M.IL	IOB (International Order Book)	full daily series	EUR … (27.08.2026)	23,512 shares
UNI.MI	Borsa Italiana (primary)	full series	EUR 28.12 (27.08.2026)	951,931 shares

The 0Q6M.L line is dead: it has no trading bar, but the API keeps serving a static price of EUR 3.839. I searched for that level in the historical series of the primary listing: UNI.MI closed at 3.838 on ….2020, at 3.840 on 21.12.2020 and at 3.840 on 21.01.2021. The quote is, therefore, frozen for about six years. The real price is 7.32 times higher.

The consequence, and the reason this report exists. The GBL scorecard in the tracker for the 0Q6M column (rows 4-33, "Data Indicatori 26.08.2026") was built on the ghost quote: row 4 says "extremely low P/E, deep-value signal" (P/E 4.39x), row 5 "Trading at ~27% of book value" (P/B 0.27x), row 42 "Graham Number discount ** …**". All three are artifacts of a 2020 price. The correct figures at EUR … are P/E 11.75x, P/B 2.01x, Graham Number discount …. The column's total score — 27/30 — is 7 points above the score of the same company evaluated correctly under UNI (20/30). The final chapter treats the divergence in detail; everything that follows uses EUR ….

What I regenerated: the data pack data-pack-0Q6M-20260827.md had come out completely empty (the empty ticker "0Q6M" does not resolve on Yahoo). I added the entry "0Q6M": ("UNI.MI", ...) to _BVB_ALIASES in deep_data_pack.py, documented with the reason, and re-ran the script — it now returns the balance sheet, ratios, and the FCF bridge. There is no SEC filing: Unipol is not SEC-registered, has no ADR listed in the US, has no CIK. The EDGAR step does not apply and was not run.

---

Executive summary

The thesis. Unipol Assicurazioni is the largest Italian-capital insurance group — #1 in Danni (non-life) and #1 in Motor TPL (RC Auto) — with EUR 17.4 bn direct premium in 2025 (…), ~15 million customers and the densest agent network in the country. Fundamentals are excellent and improving: FY2025 consolidated net profit EUR 1,530 mn (…), Danni combined ratio 92.9% (from 93.6%), Auto CoR 94.8% (from 100.0%), Solvency II 233%; H1 2026 accelerates — net profit EUR 913 mn (…), EUR 1,056 mn (…) with BPER on a full-semester basis, CoR 91.8%, loss ratio 63.9%, Solvency II 259%, consolidated equity EUR 11,813 mn. The company is well run, has a real distribution-and-data moat, and the dividend (EUR 1.12, yield 3.99%, payout 52.5%) is covered three times over by capital generation.

What changed in the three days since the UNI analysis and why it matters. Web research on this run surfaced three events that the 24.08 report did not contain, all material and all pointing the same direction:

1. The extraordinary shareholders' meeting of 30.07.2026 approved a delegation for a capital increase of up to EUR 2.5 bn, exercisable in one or more tranches by 31.12.2027, conditional on IVASS approval, intended for the acquisition of the MPS perimeter from Intesa Sanpaolo. 77.5% of capital present, 99.7% votes in favor. The cooperatives (~49%) confirmed they will subscribe pro-rata; J.P. Morgan SE leads the underwriting consortium. EUR 2.5 bn on a EUR 20.15 bn market cap = up to 19% new shares, depending on the issue price.
2. The consideration for the "Compendio Unipol" (635 MPS branches) rises to up to EUR 3.5 bn — a perimeter with ~EUR 55 bn direct deposits, ~EUR 42 bn loans, 2 mn customers, estimated net profit EUR 400-460 mn (preliminary estimates as of 31.12.2025).
3. The deal is far from certain. Intesa's tender offer on MPS (EUR 30.6 bn, 16 Intesa shares + EUR 1.0 cash per 10 MPS shares, launched 08.06.2026) is contested: MPS management launched a counter-plan (an "80 bn group"), and on 26.08.2026 Intesa filed a complaint with Consob regarding the "issues" of the operations announced by MPS after the offer launch and compliance with the passivity rule. Intesa's shareholder meeting for the supporting capital increase is on 10.09.2026. Without the tender offer there is no compendium, hence no second banking pole.

Estimated value. The FCF bridge from the data pack (CFO 4,593 − capex 539 = EUR 4,054 mn) is arithmetically correct and economically unusable: 4,680 of the 4,593 mn of CFO is the change in working capital, i.e. the accumulation of technical reserves — policyholders' money, not the shareholders'. It would produce an FCF yield of 20.1% for a company with a 52.5% payout and a 3.99% dividend, which is impossible. The definition I use is normalized net income attributable to ordinary shares = EUR … mn (EUR 2.11/share, yield 7.52% — between the market's TTM earnings yield of 8.51% and its forward yield of 8.06%, hence calibrated). Triangulation across five models gives a range of … ... …, with a median at …. Monte Carlo over 20,000 scenarios: median …, P(undervalued) …%, but the 25th percentile at ….

Verdict: HOLD, NOT BUY at EUR …. Buy only below ~EUR … (the EPV level, where you pay nothing for growth). The quality of the business is not in question — the price and the capital structure are. Three concrete arguments: (a) the stock has done 5.44x in three years (EUR 5.162 at 31.12.2023 → … today) while EPS has done only 1.56x — the remaining 3.50x is pure multiple expansion, an engine that cannot repeat itself from P/B 2.01x; (b) a capital increase of up to EUR 2.5 bn is voted and prepared, and at a plausible issue price of EUR 20 it would require EUR 264 mn of incremental owner earnings just for per-share neutrality — more than Unipol's look-through share of the MPS compendium if it ends up inside BPER via equity-method accounting; (c) 23% of H1 2026 profit comes from a EUR 211 mn revaluation of an unlisted stake (SpaceX) about which management stated publicly, on the very day of reporting, that the gain "has since halved" and that the position will not be kept.

---