Capital ratios: Closing the audit gap
Capital ratios are the best way to check a bank’s resilience, yet they aren’t audited. ICAEW was asked to consider how this gap might be plugged. Philippa Kelly finds out the results.
Understandably, the financial crisis bailouts and the increased use of contingent convertible debt – which converts from debt to equity if capital falls below a set level – brings greater focus on the capital numbers published alongside audited financial statements. However, risk-weighted assets are not audited.
Half-year results for 2015 confirm that these ratios not only remain crucial for investors but are also used by management to drive business decisions. Return on risk-weighted assets has become a prominent measure of success for banks.
The price of getting it wrong
ICAEW examined the potential for providing assurance on bank capital ratios in 2010. At that time, our stakeholder research with investors, regulators, bank executives and non-executive directors indicated that there was little demand for regular assurance on bank capital ratios or capital information.
However, the Institute was asked by the Prudential Regulation Authority (PRA) in 2014 to revisit this issue. Andrew Bailey, chief executive of the PRA and deputy governor of the Bank of England, said: “We need capital ratios to be credible. We are interested in understanding whether audit of these measures could help contribute to a process of assurance that enhances their credibility.”
There is also increasing demand for external assurance, driven by market and regulatory requirements, for example, in relation to benchmarks and indices. But assurance on capital ratios will not be universally welcomed. Banks face more cost pressures and the cost of assurance is likely to be significant – particularly for complex banks using thousands of internal models. However, that cost is small when set against the price of getting this wrong.
The need to get capital ratios right is reinforced by the new Senior Managers Regime, which makes executives and non-executives more personally accountable than they have been previously. The new criminal penalties and ‘guilty-until-proven-innocent’ approach are focusing the minds of bank executives on the strength of key functions.
In other countries, the requirement for independent scrutiny of capital information has evolved patchily: some have publicly-available assurance reports; some inform regulators alone; and some have no scrutiny whatsoever. However, given the size and importance to the economy of the UK banking sector – and the systemic risk posed to global markets – credibility and reliability are crucial for British institutions.
ICAEW’s Financial Services Faculty issued a discussion paper in July – Reporting on regulatory capital: choices for assurance – looking at how we can make audit of these figures a reality. This paper sets out the potential benefits of assurance and discusses the issues involved in designing an assurance engagement on capital ratios and related information.
Building confidence in results
Following the discussion paper, we will consider what an assurance engagement might look like in practice. Further initiatives will be needed, however, alongside assurance to reach the requisite level of confidence in capital numbers.
It is difficult to compare capital ratios between different banks on a like-for-like basis, particularly those using internal ratings-based (IRB) approaches. That is an inherent feature of a risk-based capital system.
Alternatively, banks could use a standardised approach, where risk weightings are set by regulators. IRB approaches use sophisticated models designed by the banks themselves and based on how they manage different kinds of risk.
Both approaches require extensive judgement by management and considerable amounts of data. Assurance can give users more confidence that banks have calculated their capital ratios and risk-weighted assets on a consistent basis from one period to the next, that any judgements have been applied reasonably and by appropriate people, and that the processes, controls and governance surrounding the calculations are robust.
It can also give added confidence in the quality of the data used in the calculation process, the reliability of data used and reassurance that it has been challenged vigorously. By providing added confidence in these areas, assurance may help eliminate potential sources of inconsistency and make reasons for any differences more visible.
What's your view?
We want to facilitate a consistent way to deliver assurance that meets different users’ needs. We want to know how far assurance should go and we are asking stakeholders: what assurance do you need? People should send their responses by 16 October 2015 to Philippa Kelly