The Three-Pillar System: Why Germany Still Has 1,400 Banks

The reason a German mid-sized manufacturer can still get a relationship loan from a banker who knows its factory is that most German banks are not trying to make money for shareholders. The three-pillar system is the single most distinctive feature of German finance, and it explains outcomes that look inexplicable from an Anglo-American perspective. This case study belongs to the banking pillar of the Germany Company Stories hub.

How does the savings bank pillar actually work?

Each savings bank is a public-law institution associated with a municipality or district, operating under a regional principle that restricts it to its own area. It cannot compete in another savings bank's territory, and it is legally oriented toward serving its region rather than maximising returns.

Profit is retained rather than distributed to shareholders, with a portion typically directed to municipal purposes and community activities. The absence of a return-on-equity target changes pricing behaviour fundamentally.

The pillar is federated rather than centralised. Individual savings banks are legally independent, and they share central institutions, technology platforms, asset management and a joint liability scheme that protects institutions rather than only depositors.

The consequence is a network with the scale advantages of a very large bank in technology and product manufacturing, combined with local decision-making on credit. That combination is the structural reason for the depth of German regional lending.

Why does this produce such good credit availability?

Because lending decisions are made by people with local information and a mandate to serve the region rather than to optimise a risk-weighted asset ratio from a distant head office.

A regional savings bank lending to a machining company in its district has decades of information about the owner, the workforce, the customer base and the property. That information is not in any credit model and it produces better decisions in exactly the cases where models fail: small firms, cyclical sectors and transitional periods.

The system also lends counter-cyclically to a degree commercial banks do not. During downturns, institutions without shareholder pressure to protect quarterly returns can continue supporting borrowers, which materially reduces the amplitude of regional credit contractions.

The cost is concentration. A regional bank in an automotive supplier district holds correlated exposure across dozens of firms facing the same structural decline, a risk examined in the succession analysis. The regional principle that produces the information advantage also prevents diversification.

What went wrong with the Landesbanken?

They were the pillar's attempt to build wholesale banking capability and it failed expensively. The regional state banks served as central institutions for savings banks and as state development banks, and when their explicit state guarantees were phased out in the mid-2000s they lost their funding advantage.

The response of several was to seek yield in structured credit and international wholesale markets, activities in which they had neither expertise nor competitive position. The financial crisis produced severe losses, state recapitalisations and, in several cases, resolution or forced merger.

The structural error is instructive: institutions with a low cost of funds and no commercial expertise entered markets where the only available edge was risk-taking. When the funding advantage disappeared, only the risk remained.

The pillar has since consolidated substantially, and the surviving institutions have narrowed toward serving the savings banks and regional corporate clients, which is what they were designed to do.

Does the system prevent necessary consolidation?

Within pillars, no; between pillars, almost entirely. Savings banks have merged extensively with each other and the number of institutions has fallen steadily for decades, and the same is true among cooperative banks.

What cannot happen is a commercial bank acquiring a savings bank, because the legal form does not permit it. That closes the route by which banking sectors in other countries consolidate, and it means Germany retains far more institutions than any comparable market.

The political defence is that the pillars provide genuinely different services and that removing them would concentrate credit decisions in a handful of institutions with no regional presence. The economic critique is that the system carries excess capacity that suppresses returns for everyone.

Both are correct. The system trades sector profitability for credit availability and regional stability, and whether that is a good trade depends on whether one is a bank shareholder or a regional manufacturer, which is why the Commerzbank contest generated the political reaction it did.

How does this affect financing options for companies?

It makes bank debt cheap and abundant for firms with local presence and physical collateral, and it makes equity and growth capital comparatively scarce. That trade-off shapes the entire German corporate landscape.

A machinery manufacturer with a factory and a twenty-year banking relationship can finance capital equipment at attractive terms without dilution, which is why so many German firms remain family-owned across generations.

A software company with no collateral, negative cash flow and intangible value cannot access that system at all, because relationship lending is secured lending. This is the structural reason German technology companies raise growth capital abroad, as the BioNTech case illustrates.

The policy implication is that the strength and the weakness are the same institution. A banking system optimised for lending against machines will underfund businesses whose assets are people and code, and no amount of exhortation changes a credit model built on collateral.

Is the system under threat?

From two directions. Interest rate cycles compress margins on deposit-funded lending, and regulatory capital requirements designed for large international banks impose fixed compliance costs that small institutions absorb disproportionately.

Digitalisation is the more consequential pressure. Retail customers increasingly select on application quality rather than branch proximity, which erodes the local advantage that justifies the branch network, and building competitive technology requires scale that individual institutions lack.

The pillar's answer has been shared central platforms, which works for infrastructure but not for product velocity. Decisions requiring consensus across hundreds of independent institutions move slowly.

The likely outcome is continued consolidation within pillars, producing fewer and larger regional institutions that retain the mandate but lose some of the local knowledge that justified it. That is a gradual erosion rather than a collapse, and it will be visible in credit availability to small firms long before it appears in any statistic about bank numbers.

How do cooperative banks differ from savings banks?

In ownership rather than in behaviour. A cooperative bank is owned by its members, who are typically also its customers, and each member holds a vote regardless of the size of their shareholding. Surplus is returned through pricing, service and modest dividends rather than distributed to external shareholders.

Like savings banks they operate a federated structure with shared central institutions for wholesale services, asset management, insurance and technology, which delivers scale without central control.

The practical difference for a borrower is minor. Both prioritise relationship lending, both take a long view of a customer relationship, and both are constrained in the size of exposure they can take to a single borrower.

The historical origin differs and still shapes them: cooperative banks emerged from agricultural and artisan credit associations, which is why their strongest positions remain in rural regions and in trades, while savings banks are municipal in origin and stronger in towns.

What is the joint liability scheme and how does it work?

An institutional protection arrangement in which member banks support each other so that no individual institution fails, rather than only compensating depositors after a failure. It operates within each pillar.

The effect is that a small savings bank carries the effective credit standing of the network rather than its own, which is why regional institutions can fund cheaply despite limited individual scale.

The corresponding cost is mutualisation of risk. Well-run institutions subsidise weaker ones, and the incentive to take risk is theoretically increased by the knowledge that the network will absorb the consequences.

In practice the discipline comes from within: the associations exercise substantial supervision over member institutions, including intervention in management where performance deteriorates. It is a private supervisory system operating alongside the public one.

How does the system handle a failing institution?

Usually by merger within the pillar before failure becomes visible. The associations monitor member performance and arrange combinations with stronger neighbours, which resolves problems quietly and without depositor involvement.

That mechanism explains why the number of institutions has declined steadily for decades without any of the public failure events seen in other banking systems. Consolidation happens continuously and invisibly.

The limitation is that it works for idiosyncratic problems rather than for correlated regional ones. If an entire region's industrial base deteriorates simultaneously, the neighbouring institution available to absorb a weak bank faces the same exposure, which is the scenario worth watching in automotive supplier regions.

What would change if the pillars were merged?

Sector profitability would rise and small-firm credit availability would fall. Those two effects are inseparable, because the low returns are a direct consequence of institutions lending on terms a profit-maximising bank would decline.

The experience of countries that consolidated regional banking into a small number of national institutions is instructive: branch networks contracted, credit decisions centralised, and lending to small businesses in peripheral regions declined measurably.

Whether that trade is worth making depends on what one believes the banking system is for. Germany has consistently answered that it is infrastructure for the real economy rather than an industry to be optimised, and the three-pillar structure is the institutional expression of that answer.

Frequently Asked Questions