Author: Prof. Dr Christian Schmaltz
After years of Pillar 1 attention (Basel III, TRIM), the ECB has shifted its supervisory focus to Pillar 2. In that course, it has published its supervisory expectations on how banks internally deal with their solvency risks (called ICAAP) and with their liquidity and funding risks (ILAAP).
That banks cannot only manage their risks via regulatory ratios only is made clear by articles 73 (ICAAP) and 86 (ILAAP) in the CRD IV: banks need to identify and classify their risks (Risk identification), articulate which risks and how much of them they want to tolerate (Risk appetite), break down this statement into a decentralized limit system, measure, manage, monitor and report their risks.
Thus, although every bank is required to have such ICAAPs and ILAAPs, this principle-based regulation has led to very different ICAAPs and ILAAPs in SSM countries. The ECB ICAAP and ILAAP guidelines seek to reduce that dispersion and to harmonize the different approaches in order to assess them in the SREP on a level playing field. In a first step, only banks directly supervised by the ECB, should comply with these guidelines in their 2019 SREP submission. But I expect these guidelines to be a blueprint for less significant institutions in the mid-term as well.
The gone concern approach that some countries have favored in the past remains there: in the past. The ECB emphasizes the going concern, i.e. the continuity of the business as usual. All ICAAP and ILAAP arrangements seek to ensure the survival of the bank continuing its business. If at one stage the survival can only be ensured by cutting back/ stopping business, the bank leaves the ILAAP/ ICAAP world and enters the recovery world.
The ECB expects to see in their SREP assessments, ICAAPs and ILAAPs that exhibit the following characteristics:
- Board involvement
The board understands the main elements of ICAAP and ILAAP and is aware of their strengths and weaknesses. In concise statements (10-15 pages for each), called adequacy statement, the board explains the ICAAP & ILAAP framework to the supervisor and demonstrates its effectiveness during the year.
Internal risk figures matter and are not only produced because the CRD says so. In particular, it is important to demonstrate with examples where decisions have been taken/ modified/ not been taken precisely because internal risk figures suggested it. The concise statement.
The ECB requires banks to define adequacy twice: as normative and economic adequacy.
The normative adequacy is a 3Y projection of the balance sheet (incl. capital), P&L, risk parameters, and funding under an expected (baseline scenario) and adverse scenarios. Once the future balance sheets are derived, the bank should compute all their external constraints (Pillar 1 ratios, rating covenants, SREP requirements, EU requirements, etc.) on these future balance sheets and ensure that they don’t breach any limit which would affect its ability to do business. So, the normative perspective are regulatory ratios computed on future balance sheets.
By contrast, the economic perspective computes internal ratios on today’s balance sheet.
Typical internal ratios are risk-bearing capacity on the capital side, survival horizon and loan-to-deposit ratio on the liquidity/ funding side. A conceptual overview of the two perspectives for ILAAP (ICAAP would be analogously) is shown in Figure 1.Figure 1 Conceptual overview and interplay between normative and economic perspective in ILAAP
As Figure 1 suggests, the normative perspective should learn from the economic perspective: if the economic perspective reveals that the bank is vulnerable against currency devaluations/ certain sovereign ratings downgrades, etc. exactly such an event should be simulated in the normative perspective somewhere along the 3 years.
The bank should articulate using (soft and hard) limits what it means to be adequately capitalized and having adequate liquidity and funding from an economic and from a normative perspective.
- Risk identification
The ECB expects banks to systematically identify all the risks that they are exposed to pursuing their business model. These risks should be structured in a risk tree with main risks and sub-risks. This risk tree is called risk taxonomy. In a second step banks are expected to define a materiality threshold splitting the risks into material and immaterial risks. All material risks are to be included in ICAAP/ ILAAP, i.e. the board needs to articulate how much of them to tolerate, to define (soft and hard) limits, to manage them, as well as to monitor and report the limit usage. Banks that follow a Pillar 1+ – approach, i.e. that take Pillar 1 RWA and add amounts for risks not covered under Pillar 1, have to pay attention that sovereign and public sector risk weights are usually biased under Pillar 1 not reflecting their true risk. Banks cannot argue risks of their positions are already formally taken into account in Pillar and thus do not need to be added under Pillar 2. Only Pillar 1 risks reflecting economic reality and not political consensus do not need to be added under Pillar 2. Furthermore, banks are expected to look through their (main) participations and connected entities to their underlying assets, risks, funding sources to assess whether losses and outflows could exceed the participation value. Furthermore, step-in risk for non-participations, but connected entities are expected to be considered as well.
- Risk buffers
In contrast to Pillar 1, banks have to have their internal definition of capital, liquidity buffer and stable funding sources under Pillar 2. The ECB expects these definitions to be similar, but not identical to the respective Pillar 1 definitions. Similar, because they should have the same high quality as CET1-, HQLA- and ASF positions. Not identical because the internal view should overcome regulatory provisions and take a more economic perspective taken hidden losses/ reserves into account for ICAAP to take a fairer value point-of-view (although it is not expected to fair value the whole balance sheet for ICAAP purposes).
All material risks should be quantified with regularly and independently validated models. The most common envisaged organizational setup to ensure the independence are different groups for model estimation and validation with common reporting to the same unit head. Only very few banks aspire to have a separation between model estimation and validation up to the board level.
- Stress testing
The normative and economic perspectives are scenario-based distinguishing baseline (expected) and adverse scenarios. Due to the importance of scenario generation, a principle on its own has been dedicated to this element. Like in any other regulatory publication, the JSTs expect to find in their ICAAP/ ILAAP assessment scenarios that test the weak points of the bank. If a bank sources large parts of its funding from capital markets, the bank’s rating is an important prerequisite to continue to do so. Thus, the bank should think under which circumstances it would probably lose its rating and wrap this in a stress scenario.
Although the guidelines are not legally binding, nothing prohibits the ECB to assess banks’ ICAAP and ILAAP against them. Ultimately, SREP scores and capital/ liquidity/ funding add-ons will partially depend on banks’ compliance with these guidelines in the next SREP submission in April 2019.
Among ECB banks, I have experienced a substantial uncertainty on how to interpret certain ICAAP/ ILAAP observations made by their JSTs. This uncertainty is now increased in light of the guidelines.