Why are large banks looking for top models?

Banks regulated at their core 

Financial Services e.g., Banks, Insurance, Funds, Payment are heavily regulated business lines that require licencing, regulatory reporting, mandatory staff organization and adaptation to specific, complex, and constantly evolving rules mostly to protect client’s assets. Banks have the unique challenge to have their core business of lending being constrained by prudential rules (Basel II, III, IV) that impose compulsory financial KPI in their equity level (CET1, Tier 1 capital, Total capital ratio), liquidity (LCR) and balance sheet leverage (TLAC, MREL). Specific to banking business is the need to weight in 3 drivers to make informed decision on opening or extending a client relationship: profitability, risks, and capital.

With current low rates environment that cuts down profitability, banks are vigilant in capital mobilization. Hence, they are organized to assess the risk exposure they are willing to take with clients (corporations, SMEs, Non-Profit organization, retail) and the subsequent Risk Weighted Asset (RWA) client activity generates in the banking activities: loans, savings, investments, guarantees, hedging and trading activities. By anticipating 2025 “regulatory inflation” that will sharpen RWA levels and thus capital requirements, banks will have to be even stricter in their risk appetite management even if it means reconsidering business ambitions, geographic reach, and service segments.

Strategy follows capital 

Thanks to Avertim pan-European footprint in Belgium, we are good at spotting contrasted strategies and decisions of Tier 1 banking groups driven by regulatory capital requirements. Let’s consider the three strategies of ING, Deutsche Bank and BNP Paribas for 2024;

• ING decided to close its long-established retail business in a series of countries (Austria, France) where operations were sub-scalen, while devoting resources to the countries where the bank concentrates most of its €310 bn of RWA e.g., The Netherlands (24% of total), Belgium (17%) and Germany (15%).
• As a real challenger of US-based titan banks (JPM, Citigroup), Deutsche Bank  is cutting down its business: market exit (Global market equities sales & trading), resizing (Global market fixed income rates) and IPO (Asset Management). Outcome of this business transition between 2018 and 2021 is a drop in RWA by €46 bn. Objective by 2025 is freeing-up more capital to shareholders and mitigate regulatory RWA inflation (estimated at +8% ceteris paribus).
• As the EU market continues to consolidate, BNP Paribas objective is to gain market share in revenue, fuel profitable growth (revenue growth greater than costs), control RWA increase (revenue growth greater than RWA rise) and balance RWA between its businesses. End 2021, BNPP had €714 bn in RWA.

So, it seems that determining the right RWA for a loan or a portfolio of business is becoming a competitive advantage. What is the appropriate approach and the potential positive outcome for the bank?

Case studies of capital savings & optimization thanks to better model management

• ING succeeded in reducing the RWA on its government’s debt portfolio by 4% in selecting the right model to optimize capital management. Ultimately, ING shaved €8,3 billion of RWA and saved €664 million of regulatory capital ready for reallocation to other client portfolios;
• In 2019, following European Central Bank audit on models (TRIM program), ING was compelled to implement mandatory fixes that resulted in rising RWA by €5,2 billion. Indeed, ECB considered that some models used at ING tended to reduce unwarranted variability of RWA;
• At a Tier 1 Banking group, we found out that 25% of the corporate counterparties were wrongly classified (for instance Non-Profit organization classified as Hedge funds) which generated an over-consumption of regulatory capital by 1,5% for the sample. Ongoing analysis on other segments.

Avertim recommended approach to ID a top model

1. Step 1: RWA target definition (portfolio, segment, business etc)
   • Interview with the business, risk, finance, and capital management to confirm objectives and constraints.
   • Selection of a list of potential models to target (“buy-list”).
      
2. Step 2: Models mapping and qualitative assessment
   • Interviews with stakeholders to assess each buy-list model consistency (data, process, tools, controls, governance) and outcome (assumptions, bias, calculation engine).
   • Data extraction analysis to confirm pain points (accurate calculations, missing data points).
      
3. Step 3: Decision on the target model
   • Formalization of a model book with multi-criteria assessment and recommendation.
   • Formal decision on the target model during a committee: model and roll-out strategy.
   • Recommendation on the target model operating model (architecture, connection between tools, automation, controls).
      
4. Step 4: Target models lineage & documentation
   • Design of the target model.
   • Implementation support and testing.
   • Documentation of the target model for audits & reviews.
   • Model roll-out and post-implementation support.