Organizational Culture And Bank Risk

London Financial Regulation Seminar: Organizational Culture And Bank Risk Presented by: Joel Suss (Bank of England) Date: May 18, 2021 Culture Matters, But Is Difficult To Measure "Culture has a major influence on the outcomes that matter to us as regulators." But: "As supervisors, we cannot go into a firm and say 'show us your culture" Andrew Bailey, 9 May 2016 Survey Data On Culture Has Limitations - Traditional sources of cultural information (e.g. interviews, employee surveys) tend to be unreliable - • Impression management • Normative contexts • Non-representative • Talk about culture is not culture Moreover: these methods become unreliable when accuracy becomes most important (e.g., when there is a poor culture) Unobtrusive Indicators Of Culture - An unobtrusive indicator of culture (UIC) refers to a single measure of organizational culture based on data collected without engaging employees - Good UICs use non-reactive data that reveals culture in action (i.e., not mere talk about culture) - Example: employees may say that gender diversity is important, but • How many women are on the board? • How many times does the company refer to 'she' compared to 'he' in their documentation? - Culture is measured using multiple UICs that probe different dimensions of culture Research Questions - - Can we harness data to measure bank culture unobtrusively? • Make use of untapped information • Provide objective and consistent assessments for all regulated banks Does culture affect bank risk? • Empirically test hypothesis that poor bank culture leads to higher risk Data: UICs for UK banks - 6 data sources and 20 individual indicators - Conforming to dimensions in the Organizational Culture Profile - 5/7 dimensions covered: 1. Customer orientation 2. Detail orientation 3. Inclusivity 4. Integrity 5. Risk orientation Customer Orientation - Available publicly on FCA website - 5 indicators derived from customer complaint reports • Volume complaints opened • Complaints outstanding • Complaints Upheld • Complaints closed slowly (> 8 weeks) • Redress paid (f/closed complaint) Detail Orientation - 5 indicators on the quality of regulatory reporting • Plausibility checks • Validation failures • Proportion regulatory return modules late • Average late days reporting • Average early reporting Inclusivity - Data available publicly on FCA register - 4 indicators derived from 'Form As' (applications for CF/SMF) • Gender diversity (proportion female), all authorized individuals • Difference in gender diversity, oversight - executive • Age diversity (standard deviation), all authorized individuals • Difference in mean age, oversight - executive Integrity - 3 indicators from PRA intelligence and operational risk returns • Whistleblowing referrals, volume / quarter • Internal fraud events (proportion total) • Internal fraud cost (proportion total) Risk orientation - 3 indicators from regulatory returns / capital requirements information • Capital buffer relative to peers • Average risk-weight relative to peers • P2A/TCR Culture Aggregates - Aggregate underlying indicators into: - - • Dimensions • Summary score Scale indicators (range 0-1) • Reverse 'negative' UICs • Higher values = better Simple average into dimension and overall scores • Allowing for missing indicators (if less than 50% are missing) Conclusion - - - Culture can be measured quantitatively and unobtrusively • 6 sources of data, 20 individual indicators • Coverage: universe of regulated banks and building societies Culture is important for prudential outcomes • Robust and substantive relationship with bank risk • 2 side increase in culture expected to reduce risk of failure by c.50% Future research: • Introduce additional relevant data • Explore mechanisms for how culture affects riskiness