Unicaja Pays for Loyalty Before It Collects

A 95% payout and a bottoming customer spread are real; the fee bill from Unicaja's primary-bank push is the tell.

UNI.MC · Earnings Call · 2026-10-30

Unicaja Pays for Loyalty Before It Collects

The third-quarter report from Unicaja Banco looks, at first pass, like a clean beat-and-raise bank story. Net income reached EUR 543 million for the first nine months, up 8% year on year. Adjusted return on tangible equity rose to 12.2%. CET1 stood at 15.5%, and management announced a second interim dividend of EUR 0.0642, taking 2026 interim dividends to EUR 382 million, 126% above last year. The headline capital-return message is loud: a 95% payout and a second interim payment that makes the ordinary dividend almost an afterthought. But the interesting line in this call is not the dividend. It is banking fees. Fees fell 3.8% in the quarter, and management was unusually direct about why: the bank has been rolling out additional loyalty programs that cut transactional fees for stickier customers. The CFO put it plainly: “Fees fell 3.8% in the quarter, mainly because of lower banking fees following the additional implementation of loyalty plans at the beginning of the quarter.” — Pablo Gonzalez, CFO · 2026-10-30 That is a deliberate P&L cost, not a demand shock.

The fee line is the tell

Unicaja is trying to convert transactional customers into primary-bank relationships. The evidence is in the customer metrics: high-engaged customers have risen from 32% to 41% since the end of 2024, and the share of clients benefiting from loyalty pricing has gone from 22% to 53% in less than two years. Management frames this as an investment, and the call language is explicit about the timing mismatch.

These programs include lower transactional fees for more loyal clients, and as you can see in the bottom right, the percentage of clients benefited has grown from 22% to 53% in less than two years. This is a long-term strategy similar to the investment that we are implementing in other segments and businesses that has a short-term negative impact in P&L, but will pay back in the future.
Pablo Gonzalez, CFO · 2026-10-30

This is not a brand-new theme. The prior-year call already warned that banking fees would be challenged in 2026. The Q2 2026 call made the same point in customer language: “We believe in getting new customers so that Unicaja becomes the main bank for them. And this commitment is not a short-term commitment, but a mid- to long-term commitment.” — Isidro Gil, Chief Executive Officer · 2026-07-31 What is new in Q3 is that the cost is now visible at the group fee line, even as the value-added side grows. Non-banking fees, mutual funds, and insurance are up almost 12% year on year; assets under management rose 10%; mutual funds rose 14%; net inflows reached EUR 1.6 billion, a 15% share that leaves Unicaja as the second-largest asset manager by year-to-date inflows. The net inflows engine is working. The problem is that the old transactional fee pool is being deliberately drained faster than the new fee pool is filling. That is why the guidance sounds softer than the underlying business. Total fees grew only 1% in the first nine months, and management now talks about flat-to-low-single-digit fee growth, with the quarterly trajectory flattish. The loyalty programs are structural, not a one-quarter promotion. The market will have to decide whether that is a smart customer-acquisition cost or a permanent fee reset.

The NII inflection is real, but the sensitivity is a moving target

The more bullish part of the call is net interest income. Customer spread bottomed in Q2 2026 at 226 basis points and rose to 228 basis points in Q3. Net interest income grew 3.5% quarter on quarter, and net interest margin rose 7 basis points to 180 basis points. The prior Q2 call had already called the floor: “I think that this is the floor, the basis. And from there, we're going to build up.” — Pablo Gonzalez Martin, Chief Financial Officer · 2026-07-31 Q3 confirms that claim. Management expects the customer spread to keep improving for at least the next year, driven by loan yield repricing from fixed-rate loans and a slower repricing of deposits. But the call also contains a subtle change in the rate-sensitivity story. The first 12 months are still described as low-sensitivity because the bank hedged much of its exposure in 2023. The CFO said: “For the next 12 months, we expect to have a growing NII, but with a very low interest rate sensitivity. Then, after the 12-month period, the sensitivity goes up slightly above 5% for the following 12 months.” — Pablo Gonzalez, CFO · 2026-10-30 That is a more concrete framing than the vague "low single digit" language used earlier in the year, and it matters because it tells investors this is not a pure rate-hike trade. It is a structural repricing trade. The debt portfolio is the swing factor. The ALCO book is roughly EUR 29 billion to EUR 30 billion, with a short 2.6-year duration, 82% public debt, and 84% in amortized cost. Its yield has risen to 2.8% from 2.6%. Higher rates have caused negative mark-to-market valuation adjustments that took 9 basis points off CET1 in the quarter, but management argues the same move improves future NII as the portfolio reinvests. The CFO was almost cheerful about it: “we will have even higher NII evolution than initially expected for the coming years. The full impact of this evolution will be felt in 2028.” — Pablo Gonzalez, CFO · 2026-10-30 In bank accounting, a capital hit today can be an earnings tailwind tomorrow. Unicaja is asking investors to look through the former to the latter.

Capital return is the anchor; M&A is the unanswered question

The capital story is strong enough to support the payout. CET1 fell 28 basis points to 15.5%, mostly on higher risk-weighted assets from loan growth and the valuation adjustments. MREL stood at 26.6%, with an MDA buffer above 620 basis points. LCR was 239%, NSFR 151%, and the loan-to-deposit ratio 71%, leaving an excess retail funding position that supports the ALCO portfolio. The Pillar 2 requirement fell again for 2027, the third consecutive annual reduction. The global bank-capital conversation is currently dominated by IRB model changes and risk-weight density. Unicaja's version is simpler and more local: a lower Pillar 2 requirement, DTA deductions, and a very comfortable capital position. That strength is why the bank can pay out 95% of earnings and still argue it has excess capital. But it also invites the question every analyst asked: what next? The WiZink potential operation was raised again, and management again deferred to the April CNMV disclosure. The prior Q2 call had the same exchange. The bank also said it is not looking at SRTs in the short term and has not decided on any early retirement scheme. Capital return is clear; capital deployment beyond the dividend is not.

What changed and what did not

What changed is the commercial momentum. Business volume growth accelerated to almost 5%, above the 3% guidance. Performing loans rose 5% year on year, with corporate lending up 11% and corporate loans driving the mix. New loan production reached EUR 8.1 billion, 12% above 2025. NPLs fell 15%, the NPL ratio stood at 1.8%, coverage improved to 85%, and cost of risk stayed at 19 basis points. The Spanish economy is doing the bank a favor, and management is leaning on that. What did not change is the cost and fee pressure. Total costs grew 5%, in line with guidance, as Unicaja keeps hiring and investing in technology. The loyalty programs are not a one-quarter event. The M&A question is still unanswered. And the bank is still asking investors to accept a short-term fee hit for a long-term relationship payoff. The bull case is straightforward: a bottoming customer spread, an NII inflection into 2027-2028, a 95% payout, and a CET1 ratio that is among the strongest in the sector. The bear case is equally simple: if deposit costs reprice faster than loan yields, or if loyalty customers do not convert into enough fee-paying activity, the fee drag becomes permanent and the payout is doing all the work. Q3 did not settle that debate. It just made the trade-off visible.