Bankers may benefit from the upside of risk-taking through bonuses, while taxpayers may ultimately bear the costs when things go badly wrong. After the Global Financial Crisis, regulators set out to change how bank executives are paid. This column examines how compensation practices have evolved since then and how they relate to bank risk. Longer deferral horizons, as mandated in the EU, have reduced bank risk-taking, but evidence on bonus caps is inconclusive. The 2023 banking turmoil shows that bank boards and supervisors play a crucial role in ensuring good governance of compensation frameworks.
Bank regulation has long sought to curb excessive risk-taking arising from limited liability, deposit insurance, and implicit ‘too big to fail’ guarantees. Capital and liquidity requirements are the main instruments. The Global Financial Crisis highlighted the need to complement these measures with reforms to bankers’ remuneration. In 2009, the Financial Stability Board adopted principles and standards aimed at aligning compensation with long-term performances (Financial Stability Board 2009a, 2009b).
The banking turmoil of 2023 brought the question back to the fore. Weak compensation practices were identified as contributors to the failures of Silicon Valley Bank and Credit Suisse (Financial Stability Board 2024). Since then, policies have diverged further. Switzerland is preparing stricter rules, a US reform proposal has stalled, and the UK has loosened its requirements.
In a recent paper (Gelos et al. 2026), we examine how executive compensation has evolved since the Global Financial Crisis and whether its design affects bank risk.
Implementation of global standards varies across jurisdictions
The Financial Stability Board standards call for a substantial proportion of variable remuneration to be deferred and linked to long-term performance. In contrast, they do not prescribe a cap on the ratio of variable to fixed pay. The Board did not regard the transposition of its standards in national regulation as a prerequisite for compliance and instead emphasised the role of supervisors in promoting adherence.
Jurisdictions have implemented the standards differently. The US adopted principles-based guidance without binding quantitative requirements. A subsequent proposal to introduce stricter deferral requirements has stalled. The EU went further than the Financial Stability Board standards: significant banks must defer at least 60% of senior executives’ variable remuneration over five years. Variable remuneration is capped at 200% of fixed pay, subject to shareholders’ approval. The UK removed its cap in 2023 and shortened the minimum deferral horizon to four years in 2025.
Following the introduction of EU deferral requirements, the average deferral horizon at EU banks increased by one year and now exceeds that at US banks (Figure 1). And the share of variable remuneration is now much lower at EU banks than at US banks.
Figure 1 Post Global Financial Crisis, EU banks increased deferral horizons and lowered share of variable compensation
Source: Gelos et al. (2026), Graph 2, based on Bloomberg, ISS Incentive Lab, and LSEG Workspace.
Pay is tied to performance, but rarely to risk
We find little evidence that banks adjust their performance metrics to risk.
Return on equity has become the most heavily-weighted individual performance target. Its average weight almost doubled, from 12% in 2011–13 to 22% in 2022–24 (Table 1). Return on equity can be increased by taking on additional leverage, particularly when deposit insurance and implicit state support reduce the cost of bank debt (Pennacchi and Santos 2021).
By contrast, economic value added, which accounts for the economic cost of capital, has almost disappeared as a performance metric. Explicit risk measures, such as non-performing loans and the cost of risk, receive negligible weight.
Table 1 Average weight of selected performance metrics in executive pay at large banks (%)
Notes: Based on a sample of 60 US and European firms with assets above $100 billion; bank observations shown. Percentages do not sum to 100% because not all compensation is linked to the selected performance metrics.
Source: Gelos et al. (2026), Table 1, based on ISS Incentive Lab and LSEG Workspace.
Longer deferrals are associated with lower bank risk
To examine how compensation design relates to risk-taking, we analyse CEO remuneration at 73 US and European banks. We combine compensation data with market-based measures of systemic risk, expected default probabilities, credit ratings, accounting-based risk indicators, and capital ratios.
Our approach exploits the introduction of stricter EU remuneration requirements in 2014. EU banks entered the reform with different deferral horizons and therefore varied in their need to adjust their compensation arrangements. Non-EU banks were not subject to the same requirements. We use this variability in a difference-in-differences analysis to examine whether banks required to extend deferrals by more experienced larger reductions in risk.
The results indicate that they did. For risk indicators for which the parallel-trends assumption is not rejected, longer deferrals are associated with lower systemic risk and improved credit ratings (see red bars in Figure 2). Longer deferrals also lead to higher CET1 and leverage ratios, indicating more prudent capital management.
Figure 2 EU banks that had to increase the deferral horizon of compensation became safer
Notes: The figure presents the estimated effects of the EU reform on bank risk and capital ratios. Ratings and Z-score are inverted so that larger numbers indicate riskier banks.
Source: Gelos et al. (2026), Graph 6, based on Fitch Ratings, ISS Incentive Lab, LSEG Workspace, Moody’s, S&P Global Market Intelligence, and V-Lab.
Our results are supported both by theory (Hoffmann et al. 2022) and by earlier evidence from the UK that mandatory deferrals reduced bank risk (Kleymenova and Tuna 2021).
Bonus caps: Mixed evidence
With the same difference-in-differences analysis, we also assess the impact of the introduction of the bonus cap in the EU. We find no statistically significant effect (see blue bars in Figure 2).
The UK’s removal of its bonus cap in 2023 provides a second opportunity to examine its effects. UK banks that increased variable remuneration beyond the former bonus cap experienced increases in two market-based risk measures, LRMES (long-run marginal expected shortfall) and SRISK, during 2024–25.
Taken together, our findings on bonus caps are inconclusive. The broader theoretical and empirical literature is similarly mixed. A bonus cap can reduce incentives to shift risk onto taxpayers. However, banks may respond by increasing fixed remuneration to compete with institutions not subject to the cap. This provides managers with greater protection against poor outcomes while preserving potential rewards from successful risk-taking, thereby inducing more risk-taking (IMF 2014).
Lessons from 2023: Governance matters as much as design
Credit Suisse and the three US regional banks that failed in 2023 had compensation arrangements broadly comparable with those of their peers in terms of deferral horizons. The shortcomings lay in how compensation schemes operated. Credit Suisse excluded goodwill impairments and litigation provisions from the performance measures used to determine remuneration. As a result, the metrics remained positive during 2015–17 despite the bank reporting cumulative accounting losses.
At Silicon Valley Bank, the Federal Reserve found that executive remuneration was tied to short-term earnings and equity returns without explicit risk metrics. Growing unrealised losses on the bank’s securities portfolio were not reflected in the profitability measures used to determine compensation.
In both cases, the failure of performance measures to reflect deteriorating financial conditions weakened the disciplining effects of deferrals.
Policy implications
Three conclusions follow. First, longer deferrals help increase the chance that managers bear the long-term consequences of their choices and thus adopt a more prudent risk behaviour.
Second, the evidence regarding the effects of bonus caps is inconclusive. Their effectiveness may depend on competition with institutions and jurisdictions not subject to comparable restrictions and on how banks adjust fixed remuneration.
Third, the experience of the 2023 turmoil highlights the role of governing boards and supervisors in overseeing the operation of compensation frameworks, and not only their structure. Deferrals can be effective only if the deterioration of performance is recognised in a timely manner, as risk materialises.
Source : VOXeu
