Executive summary

The shift in the interest rate environment has helped less significant institutions (LSIs) to improve profitability and capital levels over the past two years thanks to higher net interest margins. At the same time, there are persistent issues in parts of the LSI sector, such as pressure to modernise banking business models, decreased credit demand, increased funding costs and an expected deterioration in credit quality. Regarding the latter, real estate and small and medium-sized enterprise (SME) lending portfolios – both relevant areas for LSIs – have started to experience rising non-performing loan (NPL) levels and thus higher provisioning needs. All these threats will likely lead to somewhat weaker profits in the coming years.

Considering that LSIs are largely retail or diversified lenders and thus subject to high and increasing competition in the light of technological innovations, the positive developments on the profitability and capital side might support the sector’s investments in digitalisation. In the same context, the market consolidation trend for these business models is expected to continue.

This report examines the structural features of the LSI sector, its key developments and major supervisory activities aimed at addressing challenges from the authorities’ perspective.

The LSI sector represents a roughly stable share of 16% of the euro area banking market (excluding financial market infrastructures^[1]). Germany continues to account for the dominant share of that market with 59.6% of entities and 62.7% of the €4.9 trillion in total assets. Despite remaining relatively stable overall, the LSI sector continues to consolidate notably. Between the end of 2014 (inception of the Single Supervisory Mechanism, or SSM) and the end of 2023, the number of LSIs fell from more than 3,100 to 1,928. Germany, Austria and Italy accounted for the bulk of this reduction – largely driven by mergers of smaller savings and cooperative banks.

LSI profitability has experienced a major upswing over the last two years. The average return on assets (RoA) increased to 0.5% at the end of 2023, compared with 0.6% for significant institutions (SIs). For LSIs the return on equity (RoE) increased to 4.8% (SIs: 9.3%) and the cost/income ratio decreased to 59.4% (SIs: 57.0%). The key driver of the profit recovery for LSIs was the increase in net interest income (NII) on the back of higher interest rate levels. While in 2022 some LSIs still experienced significant downward adjustments in the valuation of bond portfolios, this effect faded in 2023. However, LSI profitability in the medium term is expected to shrink as higher interest rate levels start to negatively influence credit risk and credit growth, while funding costs will likely increase.

Heading towards 2024, there has already been a noticeable deterioration in LSI credit risk indicators. After trending downward for a number of years to a low of 2.1% in the first quarter of 2023, LSIs’ aggregate NPL ratio^[2] at SSM level increased slightly to 2.4% by the fourth quarter of 2023. Although such levels are not yet alarming, this is a sign that the effects of higher interest rates and weak macroeconomic circumstances on credit quality are starting to materialise after several years of a rather benign lending environment. For the time being, NPL ratio is further on rise also in 2024. Segments that have grown strongly over the past couple of years, such as real estate-related loans, are subject to particular scrutiny from banks and supervisors alike.

As strong profit growth exceeded asset growth, this enabled an increase in LSIs’ total capital ratio of 0.3 percentage points to 19.2% in 2023 (SIs: 19.8%). When looking at capital composition, Common Equity Tier 1 (CET1) ratios remain significantly higher for LSIs (17.9%) than for SIs (15.8%). In addition, the average risk-weighted asset (RWA) density is also higher for LSIs (51.1%) than for SIs (33.8%).

LSIs’ liquidity positions have benefited from an increase in deposits since 2021, especially those coming from non-financial corporations (NFCs) and households, against the background of an increase in central bank rates. Furthermore, the amount of issued debt securities has increased, whereas central bank funding has fallen by 72% or €156 billion following the repayment of central bank liquidity. The liquidity coverage ratio (LCR) stood at 217% (SIs: 164%) in the fourth quarter of 2023, well above the regulatory minimum target of 100%. Liquidity stress as observed in some cases outside the euro area has not materialised among banks under European banking supervision. Nonetheless, supervisors are paying more attention to possible idiosyncratic risks.

Given that key risks to profitability in conjunction with operational capabilities remain, and credit risks are also starting to materialise, the ECB and national competent authorities (NCAs) set their key supervisory priorities mostly in the areas of credit risk, business model sustainability, operational resilience and governance. In view of these priorities, a number of SSM-wide initiatives have been carried out in cooperation between the NCAs (responsible for the direct supervision of LSIs) and the ECB (responsible for LSI oversight) over the past two years. Details of those joint supervisory activities are outlined in this report, also emphasising the necessity for supervisors to remain flexible and risk-focused.

Technical clarification

Please note that the LSIs in Member States participating in the SSM vary greatly in terms of number, size of assets and business models. This has implications for the comparability of LSI country aggregates. The cut-off date for this report was 7 July 2024 for LSI data unless otherwise stated; any information received after that date may not be fully reflected. The sample of LSIs refers to banks at their highest level of consolidation, excluding branches and – unless otherwise indicated – excluding FMIs. Furthermore, banks that were reclassified from SI to LSI or vice versa between 31 December 2021 and 31 December 2023 are excluded for consistency reasons from a time series perspective. For information on supervisory activities, the analysis focuses only on information reported by NCAs, bearing in mind that not all dimensions of supervisory activities can be reflected and different interpretations may apply. For more details on technical assumptions, please refer to Annex D. Confidentiality rules protecting dissemination of individual bank data are in place.

1 The SSM approach to LSI supervision

1.1 The organisation of banking supervision in the SSM

The role of the Single Supervisory Mechanism (SSM), which comprises the ECB and the national competent authorities (NCAs) of participating Member States, is to ensure that EU policy on the prudential supervision of credit institutions in the euro area is implemented in a consistent and effective manner and that credit institutions are subject to supervision of the highest quality. The SSM’s three main objectives are to:

• ensure the safety and soundness of the European banking system;
• increase financial integration and stability;
• ensure consistent supervision.

The ECB exercises oversight over the functioning of the system based on the procedures and responsibilities set out in the SSM Regulation^[3] and the SSM Framework Regulation^[4] and conferred upon the ECB and NCAs for significant and less significant institutions, respectively. The ECB and the NCAs perform their tasks in close cooperation, taking into account their economic and organisational specificities. While the ECB is responsible for direct supervision of significant institutions (SIs) via joint supervisory teams composed of ECB and NCA experts, the NCAs retain responsibility for supervising less significant institutions (LSIs). The national supervisors plan and carry out their supervisory activities for LSIs using their own resources and decision-making procedures^[5].

1.2 Oversight of LSIs by the ECB

The ECB in its oversight function is responsible for the effective and consistent functioning of the SSM. The ECB needs to ensure that high supervisory standards are consistently applied to all banks in the euro area, including LSIs, while considering the different features of the national banking systems and the respective supervisory approaches.

Proportionality plays an important role in this context, which is why, jointly with NCAs, the ECB has developed a classification regime for LSIs. This aims to ensure that the approach to supervising and overseeing LSIs is proportionate and commensurate to the risks that individual institutions pose. LSIs are currently classified on the basis of their impact on the financial system and their risk profile. Since 2022 impact criteria^[6] and risk criteria are assessed separately. High-risk LSIs are classified based on the existing risk assessment by the NCA, LSIs’ compliance with capital and liquidity requirements as well as their potential crisis status. The classification for high-risk LSIs is updated more frequently (quarterly), and the outcome remains confidential for financial stability reasons. High-impact LSIs are determined once a year for each of the countries participating in European banking supervision. The classification is subject to an annual review in cooperation with NCAs, and the resulting list is published (see published list of supervised banks). For classification as high-impact, criteria such as size, business model, importance for the economy, relevance of cross-border activities, and whether a bank could become potentially SIs in the near term and are used. At least three LSIs per country should be classified as high-impact to ensure minimum coverage, although exceptions are possible. Small and non-complex institutions (SNCIs) cannot be classified as high-impact unless they are the largest LSI in a jurisdiction where all LSIs are SNCIs.

The resulting classification has implications for effective supervision and determines which notifications NCAs have to submit to the ECB and whether this is done ex ante or ex post. In addition, for all LSIs, a multi-year Supervisory Review and Evaluation Process (SREP) approach applies, meaning that the supervisor is allowed to tailor the depth of the analysis of each risk for each year, hence determining which risks should be assessed in more depth. For high-impact LSIs this entails annual assessment of each risk Supervisory capital stress tests for high-impact LSIs should be conducted at least every second year.

1.3 Organisation of LSI supervision activities

While NCAs are responsible for direct supervision of LSIs, the ECB mainly provides support through its oversight function. The ECB is also responsible for deciding on common procedures for all LSIs (as well as SIs). These include acquisitions of qualifying share­holdings, granting of banking licences (and licences for investment firms that qualify as credit institutions) and the withdrawal/lapsing of these licences, as well as decisions on the significance of banks and requests related to passporting.

To perform their LSI supervisory duties, NCAs tend to split their activities into off-site and on-site work. Off-site supervisory activities are generally broader in nature and include different types of bank-specific or horizontal assessments, meetings with management or supervisory bodies and issuance of decisions. The SSM has no predetermined minimum engagement per activity, but NCAs usually set a minimum frequency for most activities. The actual frequency of activities may differ to reflect ongoing needs and address risks in a focused manner. Meetings with the management body and to a lesser extent also the management board in its supervisory function are standard tools for NCA supervisors. In addition, supervisors typically meet on at least an annual basis with the internal audit functions as well as external auditors. On-site inspections enable supervisors to gain a very intrusive and in-depth perspective into supervised banks, typically focusing on specific topics or risk areas.

All the findings from these activities feed into SREP assessments, which are usually conducted annually by most NCAs.

Staff resources for LSI supervision vary greatly between SSM countries and depend strongly on the characteristics of the national banking system (the number of LSIs and size of the LSIs and the sector, the presence or absence of institutional protection schemes (IPSs), special business models, crises cases, etc.). For some countries, staffing is often below one full-time equivalent (FTE) per LSI for both on-site and off-site staff, especially countries with numerous entities. The median staffing is around 1.9 per LSI.

2 The structure of the LSI sector

Most LSIs are rather small in size, pursue “standard” retail or diversified lender business models and operate in highly competitive markets. Against this background, there is a very limited number of new banking licences, which mainly reflects the emergence of entities with a stronger fintech component. As a result, the total number of LSIs has steadily declined over the last two years, standing at 1,928 ^[10]entities as of the fourth quarter of 2023. Hence, the consolidation trend continues to be one of the most prominent features, driven mainly by the high number of LSI mergers in the countries with the largest populations of LSIs.

2.1 Scope of the LSI sector

Continuing on from the previous report in 2022, the total number of LSIs has further declined over the last two years, standing at 1,928 entities as of the fourth quarter of 2023. It is important to mention that Germany still accounts for roughly 60% of total LSIs in the SSM and hence has a strong influence on the LSI metrics used in this report (Chart 1), with Austria and Italy adding another relevant number of LSIs to the total sample.

Chart 1

Number of LSIs and average size in 2023

(left-hand scale: total assets, EUR billions; right-hand scale: number of LSIs)

Even though at first glance the LSI aggregate is dominated by the three aforementioned countries, this does not entail homogeneity. In fact, national LSI sectors display diverse characteristics not only in terms of the number of local entities, but also in terms of business models (Section 2.3) and overall/individual entity sizes.

As of the end of 2023, the average size of an LSI was close to €2.5 billion (SIs: €242.5 billion). However, the dispersion is material not only across countries, but also within them, which underlines the importance of proportionality (Box 1).

Chart 2

LSI entity sizes in 2023

(EUR billion)

The reduction in the number of entities between the fourth quarter of 2021 and the fourth quarter of 2023 was predominantly caused by mergers, largely of smaller savings and cooperative banks in Germany (-8%; 104 entities) and Austria (-11%; 42 entities). This trend reflects a continuous consolidation, with 88 mergers in total having been completed (Germany: 42; Austria: 31).

Additional changes came from the lapsing and withdrawal of licences, but also from new licences being granted. In the past two years:

• the licences of ten LSIs in seven different jurisdictions were withdrawn;
• the licences of 11 entities (also including subsidiaries) lapsed;
• new LSI licences were granted to 16 credit institutions in nine countries;
• ten financial holdings and branches in total were set up within the SSM.^[11]
• Following the review and assessment of significance:
• the ECB assumed direct supervision for AS LHV Group (criteria “economic importance”) as of 1 January 2023. In addition, Revolut Holdings Europe UAB (criteria “NCA request”) and Wüstenrot Bausparkasse Aktiengesellschaft (criteria “size”) underwent direct supervision as of 1 January 2024. Furthermore, Swedbank Baltics AS and CrelanCo SC (both ad hoc reviews following mergers) at the beginning of 2022, and NatWest Bank Europe GmbH, RBS Holdings N.V. and RBS International Depositary Services S.A. as of 1 January 2024 were classified as significant and became subject to the ECB’s direct supervision^[12];
• reclassification to less significant was applied to Bank Degroof Petercam NV and its subsidiaries (ownership change) on 25 February 2022 and C.R.H. – Caisse de Refinancement de l’Habitat (not meeting any of the significance criteria for three consecutive calendar years) on 1 January 2022.
• As a result, the total number of LSIs declined by 8% or 160 entities in two years.

Chart 3

Changes in total number of LSIs – 2023 vs 2021

(absolute number of LSIs)

2.2 Market share of LSIs

Since December 2021 the balance sheet of significant and less significant banks in the SSM has continued to expand, reaching €30.9 trillion in December 2023 and driven equally by LSIs (+3.2%) and SIs (+3.4%). LSIs’ growth has been steadier than that of SIs, with a rise of 2.2% in 2022 (SIs: 2.9%) and only a slight moderation in 2023 to 1.1% (SIs: 0.5%). As a consequence, the share of LSIs in total assets has remained broadly stable at around 16%. Including FMIs, the share would increase to 18.5%.

Chart 4

Market share of SIs and LSIs by country

(percentages of total assets)

A static snapshot shows that the share of LSIs’ assets in the domestic banking market varies materially across countries, ranging from less than 6% in France, Belgium and Greece to more than 40% in Lithuania and Latvia, where LSIs occupy a significant share of some business models (Chart 4). A dynamic approach reveals that in the last two years the share of the LSIs has grown in Lithuania (37.0%) and Luxembourg (14.5%). In Lithuania, this has been driven by the rise of “neobanks” – more digitalised banks which nonetheless provide traditional banking services. The emergence of these special types of entities was triggered by changes in the regulatory environment (introduction of the credit union reform and a specialised banking licence with a lower initial capital requirement). In Luxembourg – due to relocations, mergers and acquisitions but also market exits – there has been an overall contraction in the banking sector, with a decline of around 10% in the total assets held by LSIs, and a significantly more pronounced reduction for SIs. In Latvia the share of banking sector total assets held by LSIs has decreased notably (-12%) mostly due to the exit of entities.

More generally, LSIs have expanded their market share in overall loan business compared with SIs. An analysis of individual portfolios highlights that LSIs’ growth has outpaced that of SIs in businesses with NFCs and household portfolios. Cash balances at central banks have fallen following the repayment of central bank liquidity programmes (such as targeted longer-term refinancing operations, TLTROs).

2.3 LSI business models

The classification of LSI business models is based on a largely quantitative methodology that was developed mainly for SIs and was then adjusted for some qualitative aspects to cater for the lower level of data availability for LSIs.^[13]

Chart 5

LSI business models

(number of entities; percentage change since the fourth quarter of 2021)

“Retail and consumer credit lenders” is the dominant business model with 60% of LSIs belonging to this group. Diversified lenders are the second most important group. While savings and cooperative banks exist in many jurisdictions, their structure and organisation differ across the SSM. For example, in France cooperative banks are consolidated into significant groups and are treated as SIs, whereas in some other countries several LSIs are organised as decentralised savings and cooperative banks and often just covered under a joint, country-specific IPS^[14]. In Finland, they are systematised in two deposit institution amalgamations^[15] and constitute a relevant share in the country LSI market (22%). Emerging market lenders, a distinctive business model only in LSIs, are present mostly in Malta and the Netherlands (around 14% of countries’ LSIs). There are also clusters of some business models in particular countries, with custodian and asset managers being more prominent in jurisdictions such as Luxembourg, Belgium, France and the Netherlands (Chart 6).

Around 40 LSIs have a strong fintech/digital component in their business model, with around half of those being “neobanks” that operate mainly or exclusively online. Within this dynamic Fintech ecosystem, LSIs can be further grouped in different clusters depending on their business model: be it retail, payment services, SME lending, banking as a service, or brokerage. More than half of these LSIs were established in a handful of jurisdictions (Germany, Netherlands, Luxembourg). These LSIs are usually quite small in terms of asset size (from a few hundred million EUR to a few billion EUR), but can have a large footprint in terms of number of customers (millions of customers across the EU for some LSIs) or in terms of payment volumes (hundreds of billions of EUR per year processed for some LSIs).

As shown, the number of LSIs has decreased over the last two years. Through the lens of the business model classification, this reduction has mainly been observed among retail and diversified lenders as well as custodians – in absolute as well as relative terms in each category. As retail and diversified lenders are the least profitable banks (see Section 3.2), some of these entities might be less viable and hence be forced more strongly into mergers.

Across countries major differences do exist, with LSIs in some countries largely belonging to the same or similar business models, while in other countries we find a quite broad range of different business models serving their market.

Chart 6

Structure of the sector of LSIs by business model across countries

(percentages by number of entities)

3 Key developments in the LSI sector

LSIs have largely navigated well through the turbulences in recent years. Despite a boost in profitability thanks to increased interest rates, threats to the financial sector remain in the form of increased uncertainties from global geopolitical and macroeconomic tensions, but also high competition and pressure to modernise traditional (non-niche) business models. Sufficient capital buffers therefore remain important to address current and upcoming challenges for LSIs.

While balance sheets grew at a moderate pace from December 2021 to December 2023 (+3.2%), the loan business expanded significantly, especially in 2022, at the expense of cash balances at central banks and other demand deposits. At the same time, LSIs’ profitability experienced a strong upswing on the back of higher interest rates and consequently a rebound in interest income; an interim hit to securities valuations in 2022 was partially offset following subsequent reclassifications and pull-to-par effects. Increases in administrative expenses remained contained in nominal and relative terms, which was also true for impairment and provisions, at least nominally, whereas the relative increase in impairment and provisions was more noticeable. Against this background, capital levels strengthened slightly and remained comfortable on average.

Despite the re-emergence of liquidity risks among some banks in the United States and Switzerland in early 2023, there are currently no signs of structural liquidity issues in the euro area LSIs and banks have coped well with the reduction in central bank funding.

Challenges remain such as adequate loan and deposit pricing in an uncertain macroeconomic environment and a potential further rise in non-performing exposures going forward. Moreover, new risks are emerging such as increased operational risks in conjunction with IT and cyber risks as well as environmental, social and governance (ESG) challenges. LSIs need to improve their risk-bearing capacity and risk management (see also the LSI thematic review on governance^[19], Section 4.3.3).

3.1 Balance sheet composition

The aggregate balance sheet of the LSI sector from December 2021 to December 2023 grew by €153.6 billion or 3.2% to €4,913 billion (SIs: +3.4%). The LSIs growth was driven by growth in the loan portfolio of 8.5% (SIs: 6.5%), specifically loans to NFCs (8.9%; SIs: 5.6%) and households (8.1%; SIs: 3.9%).^[20] However, lending dynamics have weakened substantially recently, which might weigh on bank profitability going forward. The reasons for this include the monetary policy tightening that started in 2022 as well as the general macroeconomic outlook, which has prompted banks to tighten their lending standards. The tightening of lending standards continued throughout 2023 – at different magnitudes across countries – adding to the substantial cumulative tightening since 2022, which is now at historically elevated levels.^[21] Together with weak loan demand, this has translated into a sharp decline in loan growth to firms. Even so, loan growth has been a lot stronger for LSIs than for SIs. This might be an indication that LSIs followed less strict lending standards, which in the medium term might be dragging on profitability. While loan business has grown remarkably, cash balances at central banks and other demand deposits (-16.9%) have significantly declined.

Chart 7

Changes in total assets (December 2023 vs December 2021)

(breakdown of changes in total assets, EUR billions)

Asset growth and composition differ across countries in the SSM.

At country level, LSI asset growth showed significant heterogeneity over the course of 2022-23, largely driven by idiosyncratic developments. In some countries, there was a considerable decrease in total assets – such as France (due to a former SI being in run-down mode), Luxembourg (due to the relocation of entities as well as downsizing and a few market exits), Cyprus (following the cleaning of balance sheets) and Spain (i.a. due to a decline in two sizeable entities). At the same time, there was strong growth in other countries, often due to the emergence of new entities with fintech-driven business models, with Lithuania being the clearest example. There was also strong growth in Estonia (again due to fintech and digital business models), Croatia (coming from bank-specific but also more general growth, in particular in household and consumer loan business), Greece (largely coming from the rapid growth of one entity and mergers which led to further consolidation of the LSIs), Bulgaria (given increased lending activity in the residential real estate (RRE) and household sectors) and Ireland (bank-specific growth), all of which experienced more than 20% growth in total assets of their LSIs.

The composition of assets differs across countries as well, but loans to households and NFCs remain the core for most. Overall, about two-thirds of assets in the European LSIs comprise loan exposures, which is quite common across countries as the share of loans and advances varies mostly between 40% and 75%. Exceptions are due to the predominance of certain business models, like custodians, asset managers or fintech entities being more successful in acquiring funding rather than providing loans. Loan growth has been remarkably strong in Estonia (+66%, dominated by one bank) and Ireland (+58%, again largely dominated by one bank), while it has contracted noticeably in Cyprus (-15%). The amount of non-loan exposures, such as debt securities, again varies significantly across countries. While on average the share of debt securities in total assets is 14%, in some countries these play a more prominent role for the investment mix of LSIs (such as in Portugal, Italy, France, Spain and Latvia). The amount of cash holdings is elevated in Lithuania (driven by fintech entities, where investment business is lagging behind collection of funds), Luxembourg (coming from the strong asset management and custody business at several LSIs) and to a lesser extent in Latvia (resulting from a slowdown in loan business), Cyprus and Ireland (possibly due to its hub nature for non-euro area banks).

Chart 8

Composition of assets of LSIs: fourth quarter of 2023

(percentages)

Household deposits remain the main source of funding, especially after the repayment of central bank funding. As shown in Section 3.5, European LSIs generally rely heavily on deposits from households as their key funding source, accounting for around 47% of total LSI liabilities and equity, further increasing in 2024. The market share for the second most relevant funding source – NFCs – recently declined from 15.2% to 14.5%. Consequently, asset growth has been funded predominantly by the collection of additional deposits (+1.6%; SIs: -1.1%), largely from households (+5.3%; SIs: +5.4%) and from NFCs (+5.3%; SIs: +8.4%), which – together with a material increase in interbank funding – has more than compensated for the reduction in central bank funding, which has fallen by 72% or €156.4 billion. At the same time, the amount of equity has risen by 11% or €51.9 billion. The remaining gap to asset growth (€153.6 billion or +3.2%) has been covered by the growth of debt securities (+€19.9 billion).

Chart 9

Composition of liabilities of LSIs: fourth quarter of 2023

(percentages)

As for assets, we also see significant differences across countries for liabilities. Lower shares of deposit funding are observed in some countries such as France (coming from the business models and size of French LSIs, which are diverse and do not systematically target retail customers), Luxembourg and Ireland (both are strongly funded via financial counterparty deposits). Central bank funding usually plays a minor role, except for Italy (where investing funds in government bonds has been a popular alternative to placing them as deposits at LSIs) as well as Slovakia and Cyprus (because of the prominent use of central bank liquidity programmes). Since December 2021 the materiality of central bank funding has fallen noticeably in Portugal, Italy, Austria and Spain due to the repayment of central bank liquidity programmes. In countries with well-established covered bond and securities markets, debt securities issuance often plays a considerable role, as observed in France for some of the largest LSIs, Finland and to a lesser extent also in Austria at least for larger LSIs, the Netherlands and Ireland. In Finland there has even been a further increase in debt securities funding, which was already a material funding source for the Finnish LSIs.

Looking ahead, as relative growth in deposits has been stronger for SIs than for LSIs and hence LSIs have slightly lost market share here, this might result in the need to offer more attractive deposit rates, translating into higher funding costs and dragging on profitability.

3.2 Profitability

RoA for LSIs recovered from 0.3% in 2021 to 0.5% in 2023 (SIs: 0.6%), while RoE increased from 3.5% to 5.2% (SIs: 9.3%). The development of profitability between December 2021 and December 2023 was driven by a significant increase in NII offsetting major losses in trading income.

Chart 10

Change in year-end results for LSIs: December 2021 vs December 2023

(EUR millions, percentages)

The ratio of NII to total assets, supported by the increase in interest rates, rose from 1.2% to 1.8%; in relation to operating income, it increased from 51% to 60%, which allowed banks to widen their net interest margin from 1.3% to 2.0%. Starting in the fourth quarter of 2022, both interest income and expenses increased significantly. Given that the relative growth in interest expenses was twice as strong as for interest income, the increase in NII slowed down significantly and almost stabilised in the second half of 2023.

Trading income recorded a major loss in 2022, driven by value adjustments to investment portfolios. This was mainly due to German LSIs, given their dominance in the European LSI sample and the implications of German GAAP for the valuation of banking book securities. In other countries, LSIs experienced less of this effect as losses on bond portfolios did not immediately pass through the profit and loss account. The total trading loss for LSIs amounted to €11.0 billion in 2022, which normalised in 2023 to a profit of €4.1 billion. The rebound was again most pronounced in Germany, supported by the reclassification of securities and pull-to-par effects.

The temporary hit from securities valuations was also very visible from the number of loss-making entities: in total, there were 92 loss-making entities in the LSI sector in 2023 compared with 121 in 2021. However, the number jumped temporarily in the first quarter of 2022 to more than 800 entities, as in particular in Germany, Estonia, Austria, Spain and Italy the share of loss-making entities peaked between the first and the third quarter of 2022.

Net fee and commission income (NFCI) at the same time remained stable at 0.8% in relation to total assets; hence, in nominal terms grew roughly in line with total assets and was not a key driver of the change in profit levels. Administrative expenses also remained largely unchanged, even from a medium-term perspective over the last eight years, at around 1.8% against total assets. Consequently, the cost/income ratio for the LSI sector stood at 59.4% (SIs: 57.0%) in 2023, a significant improvement compared with 2021, when it stood at 70.2% (SIs: 64.3%). SIs saw an improvement of only around 7 percentage points. However, net impairments and provisions for LSIs deteriorated on average in the SSM, increasing from -0.06% as a percentage of total assets in the fourth quarter of 2021 to -0.27% in the fourth quarter of 2023. Despite this rise, the level of provisions remains contained.

Profit levels across and within countries as well as the magnitude of recovery differ materially, with small pockets of both high loss-making and very profitable entities in several countries.

Chart 11

Annualised RoE and RoA by country for LSIs

(percentages)

These differences across countries are caused by various factors as for example more structural drivers, such as the specific structure and nature of national banking sectors (e.g. the French banking sector is characterised by numerous rather small entities that might struggle more than others with cost efficiency), banks benefiting from being niche players with specialised business models, enabling them to achieve attractive margins in their areas of activity (e.g. in Luxembourg, the Netherlands and Belgium) or the hub nature of certain LSIs (e.g. in Ireland, where several LSIs act as an entry point to the European banking sector). Also, interest rate fixation patterns (see below) seem to be a structural feature which differs across countries.

Certain macro dynamics of course impact profitability, such as strong loan growth (e.g. in Ireland, Malta, Germany and Latvia), the share of new fintech entities in the start-up phase (e.g. in Estonia and Latvia) or ongoing restructuring efforts (e.g. in Greece, Cyprus and Italy) alter the profitability levels. Finally differences in national accounting frameworks can also have relevant implications, in particular for the timing of profit/loss recognition.

With regard to income components across countries, the general improvement in NII occurred in all countries, although to different extents. National differences were caused on the one hand by different levels of loan growth as well as sometimes specialist niche players. On the other hand, the speed and magnitude of the positive shifts were also dependent on national patterns in the fixation of loan interest rates (Chart 12). Countries with a traditionally higher share of fixed rate loans benefited to a lesser extent from higher interest rates (such as Germany and Belgium), compared with countries where loans are usually contracted with shorter interest rate fixation periods (such as Finland, the Baltic States, Portugal, Cyprus and Spain).

Chart 12

Interest fixation pattern for household mortgage loans

(percentages of total volume)

Not only does NII differ across countries, but the relevance of NFCI compared with NII also varies considerably. In most cases, the share of NFCI in operating income is between 15% and 30%, while it is lower in Estonia and Cyprus and higher in Belgium, France, Lithuania, Luxembourg, and the Netherlands due to LSIs in these countries having very specific business models focusing on asset management or transaction banking.

As a result, at different profit levels, the cost/income ratio also varies materially across countries, with a tendency for more negative outliers (meaning banks with an above-average cost/income ratio) than positive outliers in nearly all countries.

Chart 13

Cost/income ratio

(percentages)

A business model breakdown of LSIs’ profitability reveals that retail and consumer lenders together with diversified lenders are among the weakest performers. Their average RoA against the cost/income ratio is reflective of the strong competition in this market.

Chart 14

Profitability by business model

(percentages)

Overall, the sector has substantially benefited from the increase in interest rates, and hence, the increase in profitability will help LSIs to bring forward balance sheet clean-up and implement structural adjustments to their business, also in the light of innovative technologies they can use to remain competitive in the medium to long term. The repricing of liabilities has been relatively slow, while interest income from non-retail and rather short-maturity business has increased significantly. The question remains whether this effect might be masking more medium-term structural profitability challenges. It is not unlikely that several LSIs could face tougher profitability challenges again as repricing on the funding side gains pace while additional income opportunities remain limited. In addition, a continued increase in loan loss provisions, reflecting the challenging macroeconomic situation for many borrowers, will further weigh on earnings. So far as per second quarter 2024 profitability was still stable, as further increases to NII remained limited, while NFCI was slightly lower.

3.3 Credit risk

Over the observed period, the aggregate LSI NPL ratio^[22] in the SSM was initially following a downward path, reaching a record low of 2.05% in the first quarter of 2023, before the trend reversed and the NPL ratio increased to 2.37% as of the fourth quarter of 2023, 14 basis points higher than in 2021. Consequently, for the first time in ten quarters the LSI NPL ratio moved above that of SIs (Chart 15).

Chart 15

Evolution of NPL ratios

(percentages)

At SSM level, the majority of LSIs (60.1%) continue to have NPL ratios lower than or equal to 2%, although their share has reduced somewhat compared with the fourth quarter of 2021 (64.5%). At the other end of the spectrum, the share of LSIs with NPL ratios higher than 5% has also decreased marginally versus the fourth quarter of 2021 (-0.3%), standing at 9.5% as of the end of 2023. In jurisdictions such as Cyprus, Greece, Croatia, Latvia and Slovakia, the majority of LSIs are in fact high-NPL banks, accounting for a share ranging from 60% to 100% (Chart 16).

Further analysis of developments at country level reveals that between December 2021 and December 2023 the LSI NPL ratio decreased materially in Greece (‑12.9%, because of major growth in the loan book as denominator, licence withdrawal of one entity and mergers), Cyprus (‑6.5%), Latvia (‑3.9%) and Croatia (‑3.2%) (in several cases driven by portfolio sales of individual banks), which was offset by higher NPL levels in Austria, Luxembourg (+0.8%) and Lithuania (+0.7%).

Chart 16

LSI clustering by NPL ratio buckets (excluding cash balances at the central bank)

(percentages)

A closer look at LSIs’ loan book composition sheds further light on key patterns and dynamics with respect to NPL levels and coverage (i.e. the portion of NPLs covered by provisions). Chart 17 illustrates the major counterparty categories for LSIs at SSM level, with RRE accounting for the vast bulk (76.5%) of the €1,591 billion in loans to households, followed by consumer credit (CC) at only 7.4%. SMEs account for a significant share (54.8%) of the €1,163 billion in LSI loans to NFCs, with loans secured by commercial real estate (CRE) constituting another important segment (42.3%)^[23]. The aggregate LSI NPL ratio for loans to NFCs (fourth quarter of 2023: 3.9%; fourth quarter of 2021: 3.5%) is more than twice as high as that for loans to households (fourth quarter of 2023: 1.7%; fourth quarter of 2021: 1.8%).

Chart 17

Breakdown of NFCs and households by segment

(percentages)

The breakdown of the household portfolio (Chart 18) highlights the elevated riskiness of CC as the segment with the highest NPL ratio (5.0%). The comparatively high coverage ratio (43.1%) is consistent with the fact that consumer loans tend to have larger proportions of unsecured exposure. On the contrary, RRE is the segment with the lowest NPL ratio (fourth quarter of 2023: 1.1%) and lowest coverage ratio (21.2%) at SSM level. Focusing on NFC loans (Chart 19), both the SME (fourth quarter of 2023: 4.2%) and CRE (fourth quarter of 2023: 4.1%) segments have seen their NPL ratios increase by around 20 basis points, with the coverage ratio being notably higher for SME (37.9% vs 30.5% for CRE), presumably also reflecting the generally larger degree of collateralisation in CRE lending^[24].

The country breakdowns highlight not only the well-known and sustained cross-country differences in NPL levels (also discussed earlier in this section), but also significant differences in the coverage ratios within the same counterparty segments, e.g. ranging from 3.0% in Belgium to 69.6% in Slovenia for RRE, and from 6.7% in Lithuania to 57.3% in Slovenia for CRE. Apart from idiosyncratic bank/portfolio factors that might affect the averages, especially for jurisdictions with small LSI populations, this can to some degree be explained by differences in the average vintage of NPLs and the level and type of collateralisation but might also reflect cross-country differences in provisioning practices.

Chart 18

Breakdown of household portfolio by segment – NPL and coverage ratios

(left-hand scale: NPL ratios by segment, percentages; right-hand scale: coverage ratios by segment, percentages)

Chart 19

Breakdown of NFC portfolio by segment – NPL and coverage ratios

(left-hand scale: NPL ratios by segment, percentages; right-hand scale: coverage ratios by segment, percentages)

Box 3
Commercial real estate exposures in the less significant institution sector

The commercial real estate (CRE) sector has drawn significant public attention in recent years due to the impact from the shift in interest rates and structural changes brought about by the pandemic and evolving consumer habits, which has led to lower office occupancy and accelerated transformation and digitalisation of the retail business.

Chart A

LSI and SI market shares and NPL ratios in the CRE segment

(left-hand scale: LSI and SI market shares in the CRE segment, percentages; right-hand scale: LSI and SI CRE NPL ratios, percentages)

As of the fourth quarter of 2023, total loans secured by CRE^[25] amounted to €2.1 trillion for the whole SSM banking system (Chart A). In terms of market share, the contribution of LSIs was relatively moderate at 23.3% (€492 billion). However, the picture is not homogenous across countries and the LSI share ranges from less than 3% in three countries to more than 30% in six countries. This implies that different jurisdictions are exposed in an unequal way to risks stemming from the LSI CRE segment, which is further evidenced at the bank level: as of the fourth quarter of 2023, there were 213 individual LSIs, for which loans backed by CRE accounted for more than 30% of the total loan book, with the majority stemming from Germany (157 LSIs), followed by Austria (12 LSIs), Italy (10 LSIs) and Cyprus, Luxembourg, Slovenia and Slovakia (1 LSI respectively). Especially in countries where LSIs tend to have a smaller share of the local CRE market, non-performing loan (NPL) ratios for these LSI portfolios are often higher than for their national significant institution (SI) competitors, which might indicate that LSIs are occupying riskier client segments.

The degree to which these CRE exposures are subject to surging refinancing and default risk linked to the changes in the interest rate environment depends, among other factors, on the maturities, interest rate profiles and repayment structures of the underlying loans. Indicative evidence^[26] available suggests that the use of very short maturities (with regular rollover) as well as bullet and balloon repayment structures, implying large instalments due at maturity, is less common in the LSI compared with the SI sector, in particular in jurisdictions with larger LSI CRE exposures. Variable interest rate structures, which have immediate cash flow implications for borrowers upon reset, do however also appear prevalent in the LSI sector, with the exception of Germany. While these observations suggest that, for LSIs as a whole, short-term challenges in CRE lending may be somewhat less pronounced than for larger banks, medium-term and longer-term structural factors like the ones mentioned at the start of this box (also including requirements for properties to comply with energy efficiency standards) imply that CRE lending will remain a key focus point of credit risk management in the years to come.

Box 4
Geopolitical tensions and its impact on the less significant institution sector

1. Heightened geopolitical risks have been a persistent phenomenon in recent years. The general outlook also remains uncertain, with the risk of a sudden further reamplification of geopolitical tensions. The potential repercussions of such events could be wide-ranging. Even though banks in the Single Supervisory Mechanism, including less significant institutions (LSIs), have proven to be resilient so far, both supervisors and banks should remain vigilant.
2. In this context, before the Russian invasion of Ukraine, the LSIs already exhibited limited direct exposures to the two countries, which were concentrated in a few institutions. Since then, LSIs’ total direct exposures to Russian counterparties have decreased by 51%, amounting to €1.48 billion as of the fourth quarter of 2023. Six LSIs with connections to Russia have exited the market, and another six entities have taken actions to close or curtail their Russian subsidiaries. While these developments clearly indicate that LSIs’ vulnerability to first-order implications from the conflict is decreasing, supervisors continue to monitor the situation at LSIs where sizeable exposures remain.
3. With regard to the conflict in the Middle East, which flared up in the last quarter of 2023, available data indicate minor direct LSI exposures to countries in the region^[27], standing at €1.27 billion and distributed among a few entities in Germany, the Netherlands, Luxembourg, Austria and Italy.
4. Apart from risks related to regular banking business, financial market infrastructures with a banking licence have proven to be particularly exposed to geopolitical risks due to the specificities of their business model, including legal and compliance risks related to seized assets as a consequence of international financial sanctions regimes.
5. While the immediate consequences of such geopolitical events could be perceived as contained for LSIs, the second-order effects (e.g. volatility in energy prices and heightened inflation, exposures to affected sectors, increased IT and cyber risk) are major points of attention and are likely to remain in the supervisory focus in the coming years.

3.4 Capital adequacy

The key capital metrics indicate a comfortable position for LSIs on the back of improved profitability.

The total capital ratio (TCR) stood at 19.2% as of the fourth quarter of 2023 (SIs: 19.8%), 37 basis points higher compared to the fourth quarter of 2021 (18.8%; SIs: 19.6%). Capital levels over the course of these two years recovered, reaching their lowest point in the third quarter of 2022 (18.3%; SIs: 18.7%). These fluctuations were caused to a large extent by temporary losses mainly in Germany, as the positive shift in interest rates in conjunction with the requirements of German GAAP resulted in the need to immediately recognise negative changes in market value for current non-trading financial assets. From the fourth quarter of 2022 onwards, this effect started to fade out and the aggregate SSM capital position showed improvements. A decomposition of the different elements of the TCR indicates that between December 2021 and December 2023 the increase in total capital – supported by higher earnings specifically in 2023 – was the main driver of the ratio’s evolution, while growth in total assets and RWA density was only moderate.

CET1 ratios mirrored the developments described above, standing at 17.9% as of the fourth quarter of 2023 (SIs: 15.8%) (Chart 20). The share of CET1 in total capital further increased slightly to 93.3% in the fourth quarter of 2023 from 92.1% in the fourth quarter of 2021. SIs exhibited a higher TCR (19.8%) but lower CET1 ratio (15.8%), reflecting stronger utilisation of and reliance on Additional Tier 1 and Additional Tier 2 instruments compared to LSIs.

Chart 20

CET1 ratio and TCR – SIs vs LSIs

CET1 ratio and TCR

(percentages)

The distribution of CET1 ratios across the LSI population provides further evidence of comfortable capital ratios, with more than 75% of LSIs exhibiting ratios greater than 15%.

Chart 21

Percentage share of LSIs and SIs by capital bucket (CET1 ratio)

(percentages)

A country-specific breakdown of TCR by jurisdiction shows several key differences (Chart 22). First, there is a notable divergence across country medians, ranging from 39.5% (Ireland) to 15.2% (Greece). Second, some jurisdictions such as Slovakia, Malta and France exhibit material within-country variance. That being said, there is no clear-cut correlation between the within-country variance and the number of LSIs or their market share.

Chart 22

Total capital ratio by country

(percentages)

When reviewing the developments at country level over time, several additional aspects are worth highlighting. Owing to the emergence of a number of “neobanks” in the past two years, total assets in Lithuania have increased sharply by 459.3%, with a corresponding (albeit smaller) increase in RWAs (275.0%) and lesser increase in total capital (158.0%). Consequently, the average TCR in Lithuania has fallen by 930 basis points, although it still remains above the SSM average at a comfortable 20.5%. Furthermore, total assets in Estonia have grown significantly by 71.8%, but in this case with no negative impact on the TCR. The two Baltic jurisdictions have some of the highest inflation rates in the euro area, which is affecting nominal growth figures. This has, however, little impact on the SSM aggregates given these two countries’ relatively small share in the LSI market. Further marked increases in the TCR (in the range of 400-700 basis points) have been observed in Cyprus, Greece and Slovenia, but the underlying drivers are different. In Cyprus, this was due to a de-risking of the overall balance sheet, with total assets decreasing by 4.5%, RWAs falling by 14.8% and total capital growing by 17.3%. By contrast, in Greece and Slovenia there was an expansion of total assets and RWAs, which was accompanied by an even stronger increase in capital.

Analysis of LSIs' compliance with the Pillar 2 capital requirements set by NCAs as part of the SREP process (see chart 23) shows generally comfortable headrooms (i.e. excess of the levels of total capital with respect to the relevant requirements) as of Q4 2023. In 8 countries the median P2R capital headroom for LSIs exceeded 10%, while ranging from 8% to 10% in another 10. The median P2R headroom across all SSM LSIs was 8.8%, strongly driven by the countries with the largest LSI populations. The observed distribution around the SSM median shows that for most LSIs, total capital headroom stood in a range of 7% to 12%, with few LSIs falling significantly below that range and a notably higher number materially exceeding it.

Chart 23

Capital headroom with respect to overall capital requirements (capital supply compared to P2R capital requirements)

(percentages)

The overall risk exposure density (risk exposure amount over total assets) has marginally increased by 1.2%, standing at 51.1% as of the fourth quarter of 2023 and hence is still notably higher than for SIs (33.8%). It should be noted in this respect that the use of internal models to calculate risk exposures is much less pronounced among LSIs than SIs, as for LSIs 95.9% (SIs: 40%) of credit RWAs are determined by applying the standardised approach.

The leverage ratio serves as a simple, non-risk-based “backstop” to capital ratios based on RWA and is calculated by dividing Tier 1 capital by the total exposure amount, including assets and off-balance-sheet items without risk-based adjustments. As of the fourth quarter of 2023 the aggregate LSI SSM leverage ratio improved to 9.4% (SIs: 5.8%) compared with 9.1% as of the fourth quarter of 2021. Looking at individual jurisdictions, this improving trend was evident in 14 countries, with Greece displaying the most relevant growth (+4.0 percentage points). In the remaining eight countries, the leverage ratio showed signs of deterioration. This was especially the case in Lithuania (-5.1 percentage points). Croatia (-1.4 percentage points), Latvia (-1.5 percentage points) and the Netherlands (-1.1 percentage points) also saw their leverage ratio decline notably.

Nevertheless, 88.5% of LSIs (SIs: 45.8%) still have a leverage ratio above 7%. Compared with the TCR, the LSI leverage ratio shows smaller variance of medians and a more consistent picture across jurisdictions.

Chart 24

Leverage ratio by bucket and country

a) By bucket

(percentages)

b) By country

(percentages)

3.5 Funding and liquidity

Despite liquidity risk receiving heightened attention in the recent past, the situation in the euro area banking system including the LSI sector remains stable. Reductions in central bank liquidity programs have been more than offset by LSIs via growth in deposit collections, although LSIs seem to have lost market shares in NFC deposits. Despite limited overall liquidity risk, idiosyncratic risk patterns in individual LSIs are monitored by supervisors.

Supervisors and banks alike have placed renewed focus on liquidity risks against the background of non-euro area liquidity stress events at individual banks (see also Section 4). In general, however, at least from a quantitative perspective, liquidity risk remains low in the euro area, as LSIs’ liquidity position in 2023 benefited from a continued increase in deposits. In relative terms, the growth of LSIs’ deposits amounts to €61.4 billion (+1.6%), which is still lower than overall balance sheet growth. Nevertheless – as mentioned above – deposits, especially from households and NFCs in close conjunction with the development of central bank rates, are the backbone of funding for LSIs. This relevance has increased even more since December 2021, as their repayment of central bank liquidity programmes was largely backed by the collection of additional household (+€115.8 billion; +5.3%) and NFC deposits (+€35.6 billion; +5.3%), although in NFC funding LSIs seem to have recently lost market share against SIs (see Section 3.1). Deposits from credit institutions have also seen remarkable growth of €67.3 billion (+16.4%), increasing financial sector interconnectedness. The remaining gap to fund the asset growth (€153.6 billion; +3.2%) has been covered by the growth of debt securities (+€19.8 billion) and equity (+€51.0 billion).

The LCR of LSIs, an important supervisory metric for liquidity risk, has increased to 217.2% (SIs: 164.4%) on aggregate, well above the 100% regulatory requirement.

As for other risk metrics, liquidity risk metrics also differ significantly across countries. More than 80% of LSIs have an LCR of above 150%, hence showing a comfortable buffer above requirements. Just a few banks operate in a narrower band of 100-150%, whereas this is specifically observable in Austria where many LSIs are members of an Institutional Protection Scheme benefiting from a liquidity waiver, i.e. LCR requirements have to be fulfilled at liquidity group level only.

Chart 25

LCR clusters by country

(percentages)

The net stable funding ratio (NSFR) – due to its construction – tends to remain very stable and hence is largely unchanged compared with the fourth quarter of 2021 at 132.5% (SIs: 126.5%). Dispersion across countries remains more limited than for the LCR, with average values mostly between 125% and 170%.

The amount of unencumbered assets is a crucial element of contingency funding. The ratio of unencumbered assets remains high throughout the European LSI sector, according to country averages. The average for all LSIs across Europe is 86.5% (SIs: 82.8%). The differences across countries remain limited, with none of the countries showing exceptional low values for their LSIs on aggregate.

Funding concentration – expressed as the amount of funding at each bank provided by their respective top ten funding counterparts – recently lowered to 18.8% (compared with 19.6% in the fourth quarter of 2021). At country level, we see a high concentration of funding coming from the top ten counterparties in some countries, which seems to be partly driven by strong intragroup links (as in France) as well as asset management footprints (as in Ireland and Luxembourg). At the same time, numerous countries have significantly reduced their funding concentration, also reflecting the repayment of central bank liquidity programme funds.

4 Main LSI supervisory priorities and activities

4.1 Overview of supervisory priorities

Every year ECB Banking Supervision, together with the NCAs, performs a thorough assessment of the main risks and vulnerabilities faced by the SIs under its direct supervision and sets its strategic priorities for the next three years accordingly.

In supervisory practice, these translate into activities in the areas of credit risk, funding risk, business model risk, operational risk, governance risk and climate-related and environmental (C&E) risk.

The SSM priorities also set the tone when NCAs lay down their national priorities for the supervision of LSIs in their jurisdiction. When setting these priorities to their national LSI sectors, country-level idiosyncratic factors are considered, resulting in specific focus topics. As outlined in Section 3, key priorities in LSI supervision across NCAs were credit risk (in all countries, mostly as a first priority), operational risk (in 17 out of 21 countries, often the second most relevant priority), as well as governance (14 countries) and business model risk (13 countries). The overall ranking of the various risk types has remained stable over the last three years.

Figure 1

Ranking of key risks across countries

(ranking of key risks at national level)

a) Credit risk	b) Operational risk
c) Governance	d) Business model

The abovementioned priorities have generally remained key for the next cycle from 2024-26, with climate-related risks becoming the second most important priority for SIs and some new focus activities being mapped accordingly. For 2024 NCAs again very much focused on credit risk, set as the first priority in 12 out of 17 cases, governance, mostly as the second or third priority (15 countries), and operational risk (largely IT and cyber risks; 13 countries).

Off-site supervisory activities on the abovementioned priorities were supported by 218 onsite inspections on credit risk, 113 inspections with primary focus on internal governance as well as 46 missions on operation risk conducted by NCAs in individual LSIs. In total there were 475 onsite inspections conducted over the last 2 years.

4.2 Supervisory activities focusing on financial resilience

Financial resilience remains key, as despite temporary relief, profit generation remains challenging given strong competition paired with changes in the operating environment. The key focus areas for supervisors in 2022 and 2023 were credit and liquidity risk.

4.2.1 Annual credit risk monitoring

The annual monitoring of LSI credit risk metrics represents a framework for the regular, structured and data-driven monitoring of SSM-wide credit risk trends, focusing on analysis of trends and patterns and enabling cross-country benchmarking. It aims to capture the diverse facets of credit risk by providing a perspective on portfolio composition, credit growth, NPL evolution, provisioning and classification. Some of the insights from the most recent iteration are described in Chapter 3.3.

4.2.2 Funding plan assessment

In the light of developments in the interest rate environment and subsequent effects on funding after years of abundant liquidity for financial institutions, SSM supervisors also shifted their focus to liquidity and funding risk. Following cases of extreme liquidity crisis leading to failures of non-EU banks in the first half of 2023, the SSM promptly performed a centralised screening exercise. Reporting data from banks were assessed to identify possible early warning signals, leveraging on indicators such as unrealised losses in bond holdings, funding vulnerabilities related to share of deposits not covered by guarantee schemes, concentration of funding sources as well as growth of business activities. For those banks identified as being potentially vulnerable, the SSM performed a targeted follow-up.

While no immediate threats were identified, the SSM continued its monitoring activities in consideration of the ongoing (at the time) tightening phase of monetary policy as well as possible shortcomings in the funding risk management capabilities of banks. In the second half of 2023 the SSM launched a targeted review of funding risk for a sample of approximately 100 LSIs from 18 euro area countries. The targeted review leveraged, for proportionality reasons, on regulatory reporting and ILAAP submissions. For LSIs not required to submit a Funding Plan according to the EBA Guidelines on harmonised definitions and templates for funding plans of credit institutions^[28], the SSM required the submission of a proportionate subset of the Funding Plan template data points. The exercise was performed by a joint working group, composed of ECB and NCA members. The working group designed a proportionate methodology for LSIs, leveraging on the EBA funding plan template and the ECB guide to the Internal Liquidity Adequacy Assessment Process (ILAAP). The analysis focused both on the risk management capabilities of these LSIs as well as on the soundness of their funding plan projections. Depending on the results and lessons learned from this exercise, further steps may follow.

4.3 Supervisory activities focusing on operational resilience

During the COVID-19 pandemic, banks were forced to improve their operational capabilities at short notice. While banks in the end navigated quite well through the pandemic, entities are under pressure to continue improving their operational resilience. Digitalisation is not only changing the business environment and patterns, but is also leading to an increase in new threats such as IT risks, cyber risks and money laundering. Banks are increasingly forced to develop and facilitate processes using modern state-of-the art technologies and implement suitable entity and governance structures. As a result, the ECB and NCAs are also focusing on IT risk and digitalisation and as well as continuing their efforts around banks’ governance.

4.3.1 ICT risk supervision

The Digital Operational Resilience Act (DORA), which will be introduced in 2025, aims to remove legislative disparities and address uneven national and regulatory supervisory approaches regarding ICT risk (Recital 9 DORA). Prior to its implementation, the ECB prepared an internal overview of the 2023 LSI supervisory practices applied to ICT risk.

The implementation of DORA presents challenges for the current structure, where most NCAs assign a separate SREP score for ICT risk. In most LSI jurisdictions, the key ICT-relevant supervisory products for LSIs refer to SREP decisions and on-site inspections, while enforcement is based on remedial action plans proposed by LSIs following supervisory recommendations and provided there are no administrative penalties. In line with its 2024 supervisory priorities, the ECB will follow up on the supervision of LSIs’ operational resilience readiness. At the same time, both ECB and NCAs are dedicating significant efforts to a capacity building in the LSI community through sharing of experience and trainings. he challenges in building up sufficient IT-related expertise are mostly addressed by (i) recruiting ICT experts, (ii) providing training and/or support to general supervisors, and (iii) establishing a dedicated ICT unit.

Figure 2

LSI ICT risk supervision approaches compared to LSI landscapes

4.3.2 Internal governance

As part of its LSI oversight role, the ECB carried out a thematic review of the governance arrangements of LSIs in 2021-22, in cooperation with the SSM NCAs. The thematic review involved data collection from a sample of more than 200 LSIs across the 21 SSM countries and covered a broad range of aspects related to LSIs’ internal governance arrangements, especially the setup and functioning of their supervisory bodies (i.e. the management body in its supervisory function – such as the board of directors or supervisory board), as well as the related national supervisory practices.

The findings from the thematic review revealed issues concerning the boards’ composition, in particular regarding the formal independence, experience and diversity of board members. In addition, supervisory concerns were raised about the functioning and oversight role of supervisory boards. Furthermore, it was found that internal control functions had partially insufficient direct access to the boards, and risk committee structures or alternative arrangements were not always in place.

In 2022-23 the NCAs submitted action plans on how they aimed to follow up on the identified findings for the LSIs in their respective jurisdictions. To facilitate this process, several interactions between the ECB and the NCAs took place, while a workshop with the NCAs across 21 SSM jurisdictions was organised to share good practices when assessing LSIs’ internal governance arrangements. Since then, significant progress has been made by NCAs to enhance their supervisory practices when assessing LSI governance arrangements.

Overall, LSIs are still lagging behind SIs, especially regarding the formal independence of board members as well as diversity aspects, including the diversity of board members’ expertise. To promote further convergence and good practices regarding LSI internal governance across the SSM, the ECB (via its LSI oversight country desks) will continue to closely monitor and follow up bilaterally with NCAs on the progress in addressing remaining gaps and weaknesses concerning LSIs.

4.4 Climate change

Supervision of climate-related and environmental (C&E) risks was a priority area of LSI oversight in 2022 and 2023, in line with the SSM priorities. In this respect, the ECB has launched several initiatives in recent years to take stock of how NCAs have implemented their own C&E risk agendas. Moreover, the ECB offered the possibility for NCAs to opt in for some ECB supervisory exercises, i.e. the 2022 ECB thematic review on C&E risks^[29] and the 2022 ECB supervisory assessment of institutions’ C&E risk disclosures^[30], while providing regular training sessions open to NCA staff.

The ECB guide on C&E risks^[31], published in November 2020, contained a recommendation for NCAs to apply to LSIs the expectations it set out in a manner that is proportionate to the nature, scale and complexity of the activities of the institution concerned. Accordingly, most NCAs have already set expectations regarding C&E risks addressed to LSIs and have performed dedicated C&E risk supervisory assessments to evaluate their respective institutions’ ability to adequately identify and manage C&E risks and their capabilities to steer C&E risk strategies and risk profiles. Such assessments, in line with the 2022 ECB thematic review on C&E risks^[32], have revealed that, compared with SIs, a larger share of LSIs are in the early stages of development. Nonetheless, an increasing number of LSIs have developed good practices in specific areas, such as conducting materiality assessment for credit risk, assigning responsibilities to the management body and strategy setting.

In a few cases, NCAs have set institution specific deadlines aligned with those of the ECB^[33] based on the application of proportionate supervisory action.

The ECB continues to support NCAs by sharing centrally developed methodologies, templates and tools. Given the key importance of these risks in the upcoming Capital Requirements Directive (CRD6/CRR3), the ECB will continue facilitating the mutual learning and exchange of best supervisory practices. Looking ahead, the implementation of new mandates for competent authorities, particularly on transition planning, will represent a key development area for both NCAs and LSIs.

4.5 Other LSI supervision and oversight activities

4.5.1 Fintech and crypto-related activities

Consumers’ continued interest in crypto-assets and related ecosystems warrants heightened supervisory attention as SSM banks continue to be involved. Based on the perceived relevance of crypto-assets and related ecosystems, both a growth and risk^[34] factor for SSM LSIs, the ECB launched a questionnaire in 2023 on LSIs’ crypto-asset activities in order to gather insights into NCAs’ perceived risks, current supervisory and regulatory practices and key players in the crypto-asset market. Key statistics from the questionnaire were as follows:

Based on these statistics, crypto-related exposures are still very limited for SSM LSIs with the caveat that the data collection is not straightforward due to the heterogeneity of reporting standards. The participating NCAs report to be rather concerned about operational risks and money laundering stemming from this asset class and related services, rather than about other traditional risks such as credit or market risk. A high-level overview of NCAs’ risk perception of different crypto-related services (by risk type) can be seen in Table 2.

As a result, the continued monitoring of the crypto-asset class and LSIs’ involvement is warranted and may continue to be part of SSM supervisory practices, specifically in the context of the general fintech landscape and the broader digitalisation of SSM LSIs.

4.5.1.1 Fintech contact group

The ECB and NCAs have established a contact group in which LSI supervisors regularly exchange views and supervisory experience on developments in the LSI fintech^[35] sector. This initiative covers around 40 LSIs that have a strong fintech/digital component in their business model (be it in retail, payment services, SME lending, banking as a service, brokerage, etc.), half of them being neobanks that operate mainly or exclusively online. The risks stemming from the exposure of (traditional) LSIs to the fintech universe are also discussed, for instance the use of online deposit platforms or crypto-related activities.

4.5.2 LSI crisis management

The LSI crisis management framework was developed in the early stages of the SSM and consisted of several Joint Supervisory Standards (“JSS”). The framework played an important role in ensuring effective management of LSI crisis cases for a number of years. Following the ECB Banking Supervision reorganisation and drawing on lessons learned from the previous nine years, more practical guidance and further convergence of practices when dealing with LSI crisis management were needed.

In 2023 an SSM-wide project aimed at reviewing the LSI crisis management framework was jointly led by the ECB, Banca d’Italia and Banco de España with the active participation of all NCAs.

The revised LSI Crisis Management Cooperation Framework (CMCF) is now a simplified framework, with leaner documentation and a more pragmatic approach for the management of future LSI crisis cases. Examples of the changes introduced are (i) a streamlined process for the setup of the Crisis Management Cooperation Group (CMCG), (ii) a clearer definition of roles and responsibilities between the ECB and NCAs, and (iii) stronger involvement and strategic steering by senior management. As well as the revised CMCF, a collection of good practices developed by either the ECB or NCAs were exchanged to provide practical support to NCAs when dealing with crisis cases.

4.5.3 LSI public statistics

Since early 2023 the ECB has released a set of aggregated banking statistics covering LSIs. The LSI statistics complement the existing supervisory banking statistics on SIs that have been published by the ECB since 2016. The LSI statistics offer a unique supervisory perspective and enhance transparency under the SSM.

Mirroring the structure of the SI statistics, the LSI statistics cover key indicators on capital, profitability, liquidity and asset quality. The indicators can be explored visually via an interactive dashboard, which provides an overview of the latest developments in these areas and allows users to analyse time series as well as select either SI or LSI samples. The full dataset for the new LSI statistics, with granular time series going back to the second quarter of 2020, is available via the ECB’s Statistical Data Warehouse and is updated on a quarterly basis.

Supervisory banking statistics are calculated by aggregating the COREP (capital adequacy information) and FINREP (financial information) data reported by banks at the relevant point in time. Entities acting as FMI providers (CCPs and CSDs) have been excluded from the LSI statistics to avoid bias, as their size usually exceeds the significance threshold even though they may not be classified as SIs.

4.5.4 LSI stress-testing

Following the stocktake of stress-testing practices for LSIs performed over the course of 2022, the ECB and NCAs have been working on collecting and exchanging further information. This resulted in a series of ongoing workshops organised with the NCAs to promote good practices, including methodologies and tools, which might be taken into consideration by other NCAs to overcome potential issues. In addition, a quantitative stocktake of several LSI stress-testing metrics, aggregate results as well as publication formats supplemented the 2022 exercise to further enhance knowledge sharing.

4.5.5 LSI SREP methodology updates

Since 2015 the ECB and NCAs have been working together to develop a common SREP methodology for LSIs, based on the EBA SREP guidelines and building on the methodology for SIs and existing national SREP methodologies. The SREP for LSIs aims to promote supervisory convergence in the sector while supporting a minimum level of harmonisation and a continuum in the assessment of SIs and LSIs. NCAs may exercise a degree of flexibility when implementing the LSI SREP methodology to account for national specificities. Another important milestone in ensuring consistency and high standards for LSI supervision is the roll-out of the SSM Information Management System (IMAS) portal^[36] for NCAs, which facilitates assessment on the basis of the common SREP methodology and allows supervisors of LSIs to record this assessment in a single system across the SSM.

The SREP methodology is currently undergoing a cycle of reviews stemming from either regulatory needs (e.g. in order to comply with the new CRD/CRR regulatory regime or the EBA guideline on SREP), emerging risks (e.g. climate risk, IT/cyber risk) or to align any updated SI methodological changes (on business model, internal governance, credit, market or operational risk) approved by the Supervisory Board with the LSI SREP framework.

In addition, and in order to further foster the proportionality principle in SREP execution, an enhanced multi-year approach was developed with a level of granularity and intensity calibrated to the perceived risk and systemic importance of the supervised institution.

4.5.6 Material growth of LSIs

The ECB jointly with NCAs performs a biannual exercise to identify and assess banks that have experienced excessive growth in certain balance sheet positions, such as loans, deposits to households and NFCs or deposits to financial counterparties. The purpose of this exercise is to understand the idiosyncratic as well as system-wide dynamics in order to potentially address emerging risks at an early stage. In this regard, around 160 entities have been identified and assessed based on a comprehensive set of growth dimensions. During the follow-up exercise, NCAs and the ECB work jointly on reviewing the adequacy of the supervisory assessment for these entities as well as the related supervisory activities. In addition, details on national approaches to identify and monitor fast-growing entities are exchanged to help NCAs potentially improve their approaches based on processes and experiences in other countries.

4.5.7 Contact groups and cross-institutional cooperation

The ECB has also established effective and constructive collaboration between the Institutional & Sectoral Oversight Division and the LSI oversight function of the Single Resolution Board (SRB). As part of this collaboration, the ECB participated as an observer in the LSI dry run performed by the SRB in 2023-24, simulating a failure of a fictitious LSI earmarked for resolution.

In order to enhance cooperation among SSM supervisors with regard to links to third countries, there are three Third-Country Contact Groups (TCCGs) coordinated by ECB Banking Supervision, corresponding to China, Turkey and Russia. The latter has been merged with existing SSM expert groups and networks after being expanded to the entire SSM on the eve of Russia’s invasion of Ukraine. TCCGs serve as fora for sharing views on and discussing the developments in third-country jurisdictions as well as their potential impact on the EU financial sector. They are also designed to establish networks among relevant experts across the SSM, enhancing transparency among their members and facilitating decision-making by supervisors in charge of institutions exposed to those jurisdictions. The items covered in TCCGs typically include macroeconomic and financial sector developments, SSM banks’ exposures to such jurisdictions, general trends in the business and operations of third-country banks operating in the EU and updates on bank-specific cases.

5 General conclusion and outlook

The banking sector in Europe continues to navigate through challenging times. After successfully managing the challenges of the COVID-19 pandemic, the sector was immediately faced with major geopolitical tensions in combination with a delayed macroeconomic recovery and disturbances to international trade. The rapid change in the interest rate environment helped to boost profitability, which however might be temporary given the expected increase in provisioning and funding costs.

Initial indications in 2024 from a profitability and capital perspective confirm that the recent positive shift to profitability might be temporary only. Therefore, LSIs need to maintain adequate buffers to ensure that they can master possible increased cost of risk and funding while at the same time continuing to invest in adapting their operational infrastructure to the challenges of the future.

The key priorities for supervisory activities in 2024 therefore remain largely unchanged, namely: credit risk, followed by operational risk and governance. Direct supervision by the NCAs as well as oversight by the ECB will focus on the required changes in these areas and putting in place adequate incentives. Important supervisory activities for LSIs have been and will continue to be brought forward jointly by NCAs and ECB to facilitate the monitoring, benchmarking and improvement of LSI-specific and sector-wide supervisory responses with the overall aim to keep risks to the banking sector controlled and contained.

Annex

A LSIs per country

B List of high-impact LSIs

C Useful links by country

D Classification of LSI business models

Business model classification allows for analysis of profitability, business model viability and structural changes in the banking system. In particular, it permits peer group comparisons and analysis. Classification is primarily based on income-generating activities, where we distinguish between: (i) institutions focused on traditional credit business and related fee-generating operations; (ii) institutions that rely on non-lending-related fee business, such as asset management, mergers and acquisitions, securities and trading; (iii) institutions that are involved in both lending- and non-lending-related fee business. In a second stage, institutions are further distinguished according to their funding strategies, client base and geographical focus.

The 15 categories (including “Not available”) are:

Asset managers: Asset managers invest on their clients’ behalf and asset management fees are the most important source of income. Firms engaged in private banking with a focus on wealth management count as asset managers if they rely predominantly on fee-based income.

Custodians: These banks safeguard financial assets for their clients, fees being their primary source of income.

Corporate/wholesale lenders: Lenders whose main clientele is the corporate and wholesale sector, both as clients and as a source of funding.

Development/promotional lenders: State-owned banks which finance projects governments deem to be of public utility. They are typically large and have a high share of wholesale lending, on which they generate low margins.

Diversified lenders: Institutions with a balanced exposure to the retail and wholesale sectors. In terms of funding, diversified lenders are often mainly financed by their clients (both retail and corporate), although sometimes this is complemented by significant wholesale funding.

Central savings or cooperative banks: These entities provide banking services within the system of savings or cooperative banks, facilitating the flow of funds within the group from banks with excess liquidity to those with liquidity needs.

Retail banks: Focused on lending to retail clients, in many cases with a strong emphasis on residential real estate lending. Generally funded through deposits, with moderate reliance on wholesale funding.

Consumer credit lenders: Also focused on retail clients, these lenders specialise in consumer finance loans. Their funding can be heterogeneous, some relying mostly on retail deposits while others use wholesale funding.

Car finance banks: These entities are linked to car producers. They offer loans to finance car purchases for retail clients, and sometimes also trade financing for their retail network. Funding is mostly wholesale.

Financial market infrastructures (central clearing counterparties, CCPs): These entities offer central clearing services for derivatives and manage collateral, earning fees as their main income source.

Financial market infrastructures (central securities depositaries, CSDs): These offer post-trade services as a central securities depositary, mainly involving settlement, legal transfer of ownership and custody. Fees for these services constitute the main income source.

Emerging market lenders: Institutions that operate in countries classified as emerging markets (EMs). They include both banks domiciled in EMs with predominantly domestic exposures and banks domiciled elsewhere, but with significant EM exposures. They have a similar business model to diversified lenders, but feature higher lending margins and higher risk.

Investment banks: These have a relatively low share of net interest income (mostly wholesale); fees, commissions and trading activities are the main income source.

Other: Banks that operate a niche business model and cannot be classified in any of the other categories.

Not available: Banks that cannot be classified due to data limitations.

E Methodological comments

NCA annual reporting

Over recent months the ECB has been working with NCAs to improve NCA reporting templates and address various methodological constraints identified in previous years. The introduction of operational definitions aims to reduce potential misunderstandings of supervisory concepts. At the same time, reporting by NCAs on their activities and the entities under their direct supervision has been streamlined. However, data collected can only be meaningfully analysed if NCAs have interpreted the information requests in a similar manner. There remains some scope for NCAs to interpret the reporting of information differently, which can create inconsistencies in the data collected. This can affect the comparability of information received.

The survey of supervisory activities on which this report is based cannot provide a comprehensive picture of all supervisory activities conducted by NCAs on LSIs in 2022 and 2023. The analysis focuses only on reported information, and it is important to bear in mind that not all dimensions of supervisory activities can be reflected.

The report primarily focuses on the quantitative dimension of the NCAs’ performance of their supervisory tasks, such as the number of meetings held. It does not detail qualitative aspects related to the methodology being followed, including the intensity of supervisory activities on different priority groups of LSIs.

LSI supervisory reporting

Implementing Technical Standards (ITS) data (FINREP, COREP) are the key source for quantitative references to country aggregates for LSIs throughout this report. The cut-off date has been set at 7 July 2024 for LSI data unless otherwise stated; any information received after that date may not have been fully reflected. The sample of LSIs refers to banks at their highest level of consolidation excluding branches and – unless otherwise indicated – excluding FMIs such as CCPs and CSDs; the size of these usually exceeds the significance threshold even though they might not be classified as SIs. Confidentiality rules protecting dissemination of individual bank data are being put in place. Data on individual subsets, such as country-level information, can only be shared if the sample includes at least three banks and none of these accounts for more than 85% of either numerator or denominator within the perimeter. More details can be found in the methodological note for the publication of aggregated Supervisory Banking Statistics.

The whisker charts display the 10th and 90th percentiles (narrow bars), the median (mark in the box) and the interquartile range between the 25th and 75th percentiles (the thick bar). Outliers below the 10th or above the 90th percentile are not shown in the boxplots.

LSI sectors in Member States participating in the SSM vary greatly in terms of number, asset size and business model. This has implications for the comparability of LSI country aggregates. For these reasons, this report should be seen as a high-level overview of selected supervisory practices of NCAs, rather than a judgement-based comparison with definite findings or conclusions.

A different approach to the number of entities was chosen compared with the ECB Annual Report on supervisory activities.