Greek banks spent three years re-rating out of a crisis discount. On my numbers Piraeus has almost finished that journey — which means the remaining case rests on a single assumption that most investors have never examined: how little Greek depositors demand to be paid.
Piraeus earns a 16.4% return on tangible equity against a cost of equity near 10.1%. Applied through the standard bank valuation identity — justified price to tangible book equals return on tangible equity less growth, over cost of equity less growth — that supports roughly 1.8× tangible book.
The shares trade at 1.72×. The gap is about nine percent: a fair price for a good bank, not a mispricing.
That matters because the consensus case for Greek banks is still framed as a discount story — top-quartile European returns on equity trading below European multiples. That argument was correct in 2023. On these numbers it is now largely spent, and continuing to make it means buying a re-rating that has already happened.
Which forces a more precise question. If the multiple is roughly right, the return comes from earnings. And the largest single swing factor in those earnings is not loan growth, not costs, and not credit quality. It is what Piraeus pays its depositors.
In 2024 the European Central Bank's policy rate averaged 3.58%. Piraeus paid its depositors 0.54%.
That is a pass-through of roughly fifteen percent — among the lowest in Europe, and the single biggest reason Greek bank earnings held up through the hiking cycle while margins compressed elsewhere. In 2025 the figure was 17.8%.
Greek depositors, in short, have not demanded to be paid. Deposits fund roughly two thirds of the balance sheet, so that behaviour is worth more to the income statement than any operational decision management has taken.
Euribor averaged 3.58%. Deposit cost 0.54%. Net interest income of €2,088m — the high-water mark of the rate cycle.
Rates fell, and deposit costs fell with them — from €85.2m in Q4 2024 to €62.6m by Q2 2026. Net interest income troughed at €1,903m.
The model assumes the advantage erodes but does not disappear. Net interest income still recovers to €2,571m as loan growth outweighs the lower yield.
A ten-point move in the beta is worth something close to €70m of annual pre-tax profit — around four percent of earnings, every year, compounding into the terminal value. No other single input in the model comes close to that sensitivity.
If it is structural — a function of an under-banked market, limited competition for retail savings, and household preference for liquidity — the bank is worth more than 1.8× book. If it normalises toward European levels as competition returns, today's price is right, and the buyer is simply paying fair value for a well-run bank.
Greece moved from Emerging to Developed Market status on 21 September 2026, and the event is widely described as a positive for Greek equities. For Piraeus specifically, the near-term mechanics run the other way.
In separate work modelling the reclassification flows, I estimate Piraeus faces roughly €152m of net forced selling — one of the four largest outflows in the market, equal to about four days of average trading volume.
The reason is arithmetic rather than sentiment. Greece carried a 0.90% weight in FTSE Emerging and roughly 0.60% in MSCI Emerging Markets. Its projected weight in the developed indices it is joining is under 0.1%. The money that must leave exceeds the money that must arrive.
Neither effect sits in the valuation above. The analysis rests on earnings and book value, not on index mechanics.
Management targets a CET1 ratio of about 13% and expects to surpass it. With a 16% return on tangible equity and loans growing at five to seven percent, Piraeus produces capital faster than the balance sheet absorbs it.
The ordinary dividend is guided at 57% of earnings. The model returns everything generated above the 13% target plus a 50 basis point buffer through buybacks, which takes total distributions to 71% of earnings by 2030 while holding the CET1 ratio flat at 13.5%.
That is the realistic path. A bank earning sixteen percent and retaining forty-three percent of it would accumulate dead capital, and the market would not pay for it.
One caveat deserves naming. Deferred tax credit was 54% of CET1 at June 2026. That is low-quality capital which runs off over time and is a genuine consideration in how much of the reported ratio an investor should credit. The model treats it as an explicit drag.
There is no EBITDA, no enterprise value, no free cash flow in the usual sense. Interest income is the top line and deposit expense is the direct cost. The balance sheet drives the income statement rather than following it.
Distributions are capped by the CET1 ratio. The model tests it every year: the dividend is the lesser of the target payout and what capital allows.
Most forecasts have a dozen inputs and one that matters. Identifying which, and sizing its sensitivity, is more useful than refining the other eleven.
Piraeus's own bridge from pre-provision income to profit before tax is out by €9.7m in FY2024. Rather than force it, the model carries an explicit reconciling line.
No price target is published here. The analysis says the discount has closed and names what an investor is now actually taking a view on. That is a position, not a hedge.
Integrated bank three-statement model built to CFI structure — assumptions, income statement, balance sheet, cash flow, and supporting schedules for net interest income, asset quality, regulatory capital and the acquired insurance segment. 947 formulas. Historical figures tie to reported on every line; the balance sheet balances in every forecast period; eleven integrity checks, all passing.
Built from the Piraeus investor relations databook and H1 2026 results presentation. Not investment advice. Educational portfolio case study.