Banco BPI and CaixaBank: Anatomy of an Iberian Cross-Border Bank

BPI is the bank that shows what happens after a cross-border acquisition, once the headlines have gone. Nine years into CaixaBank ownership, it is possible to answer the questions everyone asks when a foreign group buys a national bank: does lending shrink, does the brand disappear, do decisions move abroad, and does the country end up better or worse off? The answers are more nuanced than either the nationalist or the free-market version of the story predicts. This case study is part of the Portugal Company Stories hub.

Where did Banco BPI come from?

BPI began in 1981 as Sociedade Portuguesa de Investimentos, one of the first private financial institutions licensed after the post-revolution nationalisations. It became a bank in 1985 and grew through the acquisition of Banco Fonsecas & Burnay in 1991 and Banco de Fomento e Exterior in 1996 — the latter transaction carrying the Angolan business that would later define its strategic problem.

Unlike Millennium BCP, BPI built a reputation for conservative underwriting. Through the 2008–2014 period it took state contingent convertible capital under the Troika programme but repaid it early, in 2014, and avoided the scale of asset-quality damage that hit peers.

Its identity was also strongly Porto-based rather than Lisbon-based, and closely associated with Fernando Ulrich, who led the bank for two decades and became one of the most quoted business figures in Portugal.

Why was Angola such a complicated asset?

Because Banco de Fomento Angola (BFA) was extraordinarily profitable and extraordinarily risky at the same time. In some years BFA generated a large share of BPI’s consolidated earnings from a market with sovereign risk, currency convertibility problems and a governance environment that European supervisors viewed with alarm.

The European Central Bank ultimately forced the issue by requiring BPI to resolve its large exposure concentration to a non-EU sovereign. The solution was to reduce BPI’s holding in BFA to a minority position, transferring control to Angolan interests, which removed the concentration breach and cleared the path for the CaixaBank transaction.

This sequence is worth studying because it shows supervisory rules driving corporate structure. BPI did not sell down Angola because it wanted to; it did so because holding it made the bank unacquirable and non-compliant.

How did the CaixaBank takeover actually happen?

CaixaBank had been a BPI shareholder for years before it moved to control. The decisive obstacle was the voting-rights cap in BPI’s statutes, which limited any shareholder to 20% of votes regardless of economic stake, and the standoff with Angolan shareholder Isabel dos Santos, who held a substantial block.

Once the cap was removed by shareholder vote and the Angolan exposure was restructured, CaixaBank launched a public tender offer and secured majority control in 2017, subsequently increasing its stake to near-total ownership and delisting the shares from Euronext Lisbon.

The delisting is an underrated part of the story. Portugal’s stock market lost one of its more liquid financial names, continuing a long-running contraction of the Lisbon exchange that has left it with very few large listed companies — a structural issue discussed further across the Portugal hub.

What changed at BPI under CaixaBank ownership?

Commercially, less than critics feared. BPI has remained an active lender, holding around 12% of the Portuguese loan market as of May 2025 and writing €1.9bn of new mortgage lending in the first half of that year. Its private banking arm commands roughly 24% of the sector’s assets under management, which is a disproportionately strong position for the fifth-largest bank.

Structurally, a great deal changed. Capital allocation, risk appetite limits, technology platforms and product manufacturing shifted toward group-level decision-making. BPI gained access to CaixaBank’s insurance and asset-management capabilities and to funding at a group cost rather than a Portuguese standalone cost.

Culturally, the bank retained its brand, its Porto headquarters and its local management structure — precisely the model BPCE has said it will apply to novobanco.

Does foreign ownership reduce lending to local businesses?

The Portuguese evidence is mixed and depends heavily on the cycle. In stress, a foreign parent can be a stabiliser — it has capital that a standalone national bank would not have — or a transmission channel for problems that originate abroad, if the parent retrenches to protect its home market.

During the eurozone crisis, foreign-owned subsidiaries in several peripheral countries reduced lending faster than domestic banks, because group-level deleveraging targets were applied across geographies. Portugal saw some of that. Since 2017, with parents well capitalised, the pattern has reversed and subsidiary lending has been competitive.

For a CFO negotiating credit lines, the practical implication is to understand the parent’s country-exposure strategy, not just the local relationship manager’s enthusiasm. Limits set in Barcelona or Paris are not visible from Lisbon until they bind.

What does BPI’s private banking strength tell us?

That in a small market, distribution niches can be more defensible than scale. A roughly 24% share of sector assets under management for the fifth-largest bank means BPI’s wealth franchise substantially outperforms its balance-sheet position, built on long-standing relationships with Porto and northern Portuguese business families.

This is also where group ownership adds most value. Wealth management is a product-manufacturing business, and CaixaBank brings fund ranges, insurance products and structured offerings that a standalone Portuguese bank could not economically build.

The risk is the mirror image: wealth clients are relationship-driven and mobile. If integration degrades service quality or a key private banking team departs, share can move quickly in a way that mortgage share cannot.

How does BPI compare to its Portuguese peers?

It is smaller and more conservative than Millennium BCP, less politically exposed than CGD, and structurally similar to Santander Totta — both are Spanish-owned subsidiaries of large Iberian groups operating in a market where their parent is the dominant player next door.

On profitability it sits comfortably in the pack. On asset quality it has historically been the best of the large Portuguese banks, a legacy of its underwriting culture through the 2000s boom that spared it the worst of the post-crisis clean-up.

Where it differs most is optionality. BPI has no independent strategic agenda; its expansion, acquisitions and capital plans are decided as part of CaixaBank’s Iberian allocation. That is efficient, but it means the bank cannot pursue a Portuguese-specific opportunity that does not clear a group hurdle rate.

What does the BPI case predict for novobanco under BPCE?

A likely continuity of brand, network and local management, with progressive integration of infrastructure and capital decision-making. The BPI precedent suggests the commercial franchise survives, the balance sheet gets stronger, and the strategic autonomy disappears.

It also suggests a specific pattern in the numbers: cost synergies materialise slowly, because they come from technology and product manufacturing rather than branch overlap in a market where the acquirer had no branches. BPCE has no Portuguese retail network to merge, exactly as CaixaBank had none in 2017.

The main difference is cultural. CaixaBank and BPI are both Iberian retail banks with similar operating models. BPCE is a French mutual group buying a private-equity-optimised bank, which is a wider gap to bridge.

How did BPI perform through the interest rate cycle?

Strongly, in line with the sector. Portuguese banks benefited disproportionately from the 2022–2024 rate increases because the domestic mortgage book is overwhelmingly variable-rate and reprices within months, while deposit rates lagged. BPI’s conservative loan-to-value profile meant credit losses stayed minimal even as instalments rose sharply.

The subsequent easing cycle reverses that mechanism. Net interest income compresses faster in Portugal than in fixed-rate markets, which is precisely why BPCE cited variable-rate exposure as a diversification benefit when it acquired novobanco — the Portuguese and French rate profiles offset one another.

For BPI the mitigation is fee income, where the private banking and insurance distribution businesses provide a revenue stream that does not depend on the spread environment. That is a structural advantage of belonging to a group with product manufacturing capability.

What is the state of the Lisbon stock exchange after BPI’s delisting?

Diminished. Euronext Lisbon has lost a series of significant names over two decades through takeovers, delistings and relocations, leaving an index dominated by a small number of large companies in energy, retail, paper and utilities. BPI’s departure removed one of its more liquid financial stocks.

This matters beyond market vanity. A shallow domestic equity market means Portuguese companies raise growth capital privately, through bank debt or from foreign investors, which reinforces the pattern of foreign control described in the analysis of Portuguese banking ownership.

It also removes a governance mechanism. Listed companies face continuous disclosure, analyst scrutiny and a market price that disciplines management. A wholly owned subsidiary reports to a single shareholder, and the public loses visibility into how a significant part of the national banking system is run.

Frequently Asked Questions