How much capital do savings banks really save with the IRBA?

In recent years, German savings banks (Sparkassen) have increased their capital positions considerably – yet their buffers above the minimum requirements have improved only slightly. Based on the 2021 to 2025 disclosure reports of 30 large savings banks¹, we analysed how their capital ratios have developed: Common Equity Tier 1 (CET1) capital rose by around a quarter at the median, the CET1 ratio by only just over one percentage point. An institution that really wants to improve its ratio has two options – raise more capital or tie up less of it. The latter can be achieved through smaller optimisations (such as extending the share of real estate exposures eligible for preferential treatment, or using the property value) or, as a bigger step, through the Internal Ratings-Based Approach (IRBA). What concrete benefits does the IRBA bring a savings bank?

The status quo: capital is growing, headroom is not

Savings banks retain their earnings, and they do so successfully: every savings bank analysed with a complete time series built up CET1 capital between 2021 and 2025, by just under a quarter at the median. Little of this shows up in the capital ratios, because the total risk exposure amount (TREA) to which every ratio refers grows as well: across the savings banks that report under the standardised approach throughout, own funds rose by around a quarter from 2021 to 2025, the total risk exposure amount by just under a fifth.

At the same time, the requirement has increased – the overall capital requirement comprising Pillar 1, the Pillar 2 add-on and capital buffers rose at the median from 11.5% to 12.3% over the period under review. In addition, the introduction of CRR III has raised capital requirements from 2025 onwards. The capital headroom – i.e. the difference between a savings bank’s total capital ratio and its capital requirement – is therefore practically the same today as it was four years ago. Four years of retained earnings, precautionary reserves and individual capital measures have, in the end, not increased it.

A further change becomes visible in the reports for the first time in 2025. Since 1 January 2025, the revised Capital Requirements Regulation (CRR III)² has applied, and with it a new Credit Risk Standardised Approach (CRSA). In the disclosure reports as at 31 December 2025, the same business volume ties up more risk under the standardised approach than in the previous year – at almost all of the savings banks analysed. We have described the individual changes behind this in the article “CRR III – the new CRSA”.

The total risk exposure amount grew by 9.6% in 2025, business volume by only 2.3% – the latter in line with the Deutsche Bundesbank’s total assets statistics for the savings banks as a whole⁴. The increase therefore stems predominantly not from additional business, but from how that business is measured; how much of it is attributable to the new risk weights and how much to business mix and rating migrations cannot be separated on the basis of the disclosures alone. For capital planning, this means that the standardised approach has become more expensive as a starting point.

The obvious route: raising additional capital

To raise its ratio, a savings bank can take on capital, for example in the form of subordinated loans. For an illustrative savings bank with total assets of €8 billion and a total risk exposure amount of around €6.2 billion, moving from 14% to a strategic total capital ratio of 16% means a need for €124 million in additional own funds. To be competitive and to raise capital on a larger scale, subordinated loans require a spread of 1% to 2% over the risk-free rate. However, they are not recognised in full as own funds over their entire term. The ongoing cost of increasing regulatory capital by this amount through subordinated loans is therefore typically above €2 million per year. Covering it with Tier 1 capital is considerably more expensive still and would be closer to €6 million per year.

The alternative: reducing capital requirements

For the illustrative savings bank with total assets of €8 billion, this can be worked through. If retail business and corporate loans are moved to the IRBA, the models roughly halve the risk exposure amount of these portfolios. In the long run, not all of this reduction comes through, because the output floor caps the reduction of the total risk exposure amount – at 45% in 2026 and at 27.5% from 2030. This is exactly the limit reached here: the total risk exposure amount of our illustrative savings bank falls from €6.2 billion to €4.5 billion. Measured against the strategic target ratio of 16%, the same assets then tie up €274 million less capital, and the total capital ratio rises from 14.0% to 19.3% without a single euro of new capital.

What happens to this €274 million is up to the savings bank. The first use is growth: how much new business one euro of free capital can support depends on the risk weight of that business. For Schuldschein loans with a risk weight of around 70%, it amounts to around €2.4 billion of additional volume; for the mixed existing portfolio of our illustrative savings bank, around €3.9 billion. With net margins of 0.30% and 0.45% respectively, this corresponds to additional income of €7 to 18 million per year (before taxes).

The second use is pricing. The capital cost of a loan results from its risk weight, the strategic capital ratio and the required return on capital – calculated as a spread over the risk-free rate, because the capital held earns that rate anyway. We assume 6 percentage points, compared with 1 to 2 percentage points for subordinated capital. If the risk weight falls, the capital cost falls with it – depending on the type of business, by 0.10 percentage points for a secured residential mortgage loan with a loan-to-value ratio of 80% up to 0.26 percentage points for a corporate loan without an external rating. Whether the savings bank books this leeway as margin or passes it on as a price advantage is a management decision.

^{Table 1: Effect of partial use of the IRBA in an illustrative savings bank with total assets of €8 billion (own model calculation under the fully phased-in output floor; assumptions in the text)}

Capital relief and the pricing advantage occur side by side. The new business potential, by contrast, uses the same released capital and is therefore an alternative use, not an additional item. This can be summed up in one sentence:

What the disclosure reports show

That leaves the question of whether this is also evident outside a model calculation. The disclosure reports of the savings banks that have already taken the step confirm the order of magnitude. Two of the savings banks analysed switched to the IRBA in 2025. Since then, their risk exposure amount for credit risk has been around a third lower – with business volume unchanged or even higher. So this is not a contraction of the balance sheet, but a different measurement of the same business. Their CET1 ratio rose by just over five percentage points within one year.

The difference in level can be shown independently of size. Risk density relates the total risk exposure amount to business volume and is therefore comparable across institutions. In 2025, the median under the standardised approach is around 65%; for the three IRBA savings banks in the sample – the two switchers and one savings bank that has been reporting under the IRBA for years – it lies between 34% and 41%. Per euro of business, the standardised approach thus ties up almost twice as much risk.

Across all the savings banks analysed, one thing stands out: at almost all of them, the development of the total risk exposure amount has eaten up the capital build-up. On average across the savings banks under the standardised approach, 4.0 percentage points of capital build-up are offset by 3.2 percentage points from the increased total risk exposure amount; on balance, 0.8 percentage points remain. The total risk exposure amount made a positive contribution to the capital ratio only at the two switchers – at 4.9 percentage points, an even larger one than the capital build-up itself.

How much of this the output floor leaves

The output floor limits the capital savings an institution can achieve with internal models compared with the standardised approach. From 2030, it stands at 72.5% of the capital requirement calculated under the standardised approach; the maximum saving is thus permanently 27.5%. Until then, it rises in stages³ – we have described these stages in the article “CRR III – Transitional provisions”.

The floor is not currently binding for any of the three IRBA institutions in the sample. For the decision, however, this is not the relevant situation: the €274 million mentioned above has already been calculated under the full floor. It is the permanent relief, not that of the first year. Anyone deciding today should calculate in both worlds – and treat the higher relief in the intervening years as an additional effect, not as a basis for planning.

Impact and conclusion

For savings banks under the standardised approach, the figures say one thing above all: the increases in capital to date – through retained earnings and in some cases through raising additional capital – have merely met the requirements, but have not created any additional headroom. CRR III has de facto raised capital requirements once again from 2025. Raising subordinated loans is relatively unusual for savings banks and, if done on a larger scale, expensive. By contrast, the IRBA appears comparatively inexpensive; it releases capital and enables growth and more attractive pricing.

The model calculation is based on assumptions about portfolio mix and model impact that each institution must verify for itself; and the empirical evidence so far rests on two switches – a strong but narrow finding that can be re-examined with every further institution that makes the switch. In the authors’ view, the direction is nevertheless clear: the possibility of partial use has made the IRBA considerably more attractive, and the standardised approach has become more expensive, particularly as a result of CRR III.

msg for banking is your experienced partner for impact analyses and IRBA projects – from the initial quantification and the business case to support throughout the supervisory approval audit. As the effect of the output floor already needs to be taken into account in capital planning today, every institution should assess its potential promptly. Get in touch with your contact at msg for banking!