The UK banking system is large and concentrated. The first job of the Prudential Regulation Authority is to keep it safe and sound. Parliament has also given the PRA a secondary objective: to facilitate effective competition both in banking and insurance. Competition is good for consumers and should help the system renew itself for a more resilient future. Last week, the PRA announced an important step to promote competition in the capital framework for banks and building societies. It has authorised 23 new banks since the current regulatory framework came into effect in 2013, during which time challenger banks have continued to compete with the bigger players. These banks have brought new technology, innovative business models and valuable services to many types of customers. The PRA has sought to facilitate newcomers’ entry to the market, setting up a dedicated supervisory unit to support them. However, they do face some difficulties competing with the large incumbents, which came out of the global financial crisis with a greater market share than when they went in. Regulators around the world have spent years transforming prudential standards in order to repair the financial system. This is no time for a retreat. Jeopardising the benefits of a more stable financial system, which supports the real economy to grow and create jobs, would be misguided. Naturally, where features of the new regime do not work properly, we should consider how to fix them – while remaining true to the goal of stability. Take, for example, the risk of a significant difference in capital requirements between the challenger and incumbent banks. Larger firms typically use internal models to calculate their capital requirements. This means that the weights used are sensitive to the specific risks they take. Smaller or younger groups typically use standardised weights. But, as the Treasury select committee has highlighted, there can be large disparities between the two calculations. Research by the Bank of England has shown that the gap between internal models and the standardised approach for low loan-to-value mortgages has given smaller banks an incentive to increase lending for higher loan-to-value mortgages. That somewhat defeats the purpose. Recent initiatives – such as the systemic risk buffer for large lenders – have already narrowed the effect of disparities between the two approaches. The next stage at the global level is the subject of discussions in Basel. Meanwhile, the PRA is making an important and complementary step in the domestic framework. Last week, the PRA published proposals to supplement its current risk-by-risk assessment, used for setting banks’ requirements, with an overall assessment of the total level of capital needed for each firm. As a check, it already makes extensive use of business model analysis and peer reviews. But we can go further. Under the plans, supervisors of a bank using the standardised approach would be informed by benchmarks calculated from the risk weights used by those employing internal models. And they will take into account factors such as the extent to which existing methodologies can lead to smaller banks maintaining capital in excess of a prudent coverage of the risks faced. Where this work shows the PRA can safely lower a bank’s capital requirement, it will do so.