Banks’ fragile solvency

The collapse of Silicon Valley Bank, the rescue of Credit Suisse by UBS and the spread of concern to other banks have stirred memories of heated debates about the regulatory response to the 2007-08 global banking crisis.

That response has left banks with more capital and liquidity than before the crisis, but they remain highly leveraged on an equity-to-assets basis. The devil is in the definitions; common equity tier 1 (CET1) is more stringent than balance sheet equity, but ‘exposure’ for regulatory purposes can be less than total balance sheet assets.

All about the ratio

The devil also lies in the gap between total and risk-weighted assets (RWA), with the latter attracting most attention. A ratio of CET1 to RWA of 15% is officially comfortable, but if total exposure is three times RWA (even using less stringent definitions of equity), you can get to leverage ratios of 10-20 times in old parlance. This is a reminder of how fragile bank solvency can be when asset values are falling.

It has long been my view that equity is the only genuine form of loss-absorbing capital. Anything that could in theory convert to equity will probably be held by investors who prefer bonds and believe they can get to the door first on the call ‘fire’. Since return on equity is an annoyingly popular metric and equity is more expensive than debt, banks remain incentivised to keep it to a minimum.

Under the liquidity coverage ratio, a bank should hold sufficient high-quality liquid assets to survive 30 days of stress. But that simply buys time for the central bank to solve the problem of a run. It does not abolish a problem exacerbated by the ease of electronic withdrawals.

This brings us to a perennial regulatory issue: retail and most business deposits are regarded as ‘behaviourally’ stable. These on-demand liabilities are used to fund loans of much longer duration – the famous maturity mismatch that makes the economic world go round. Deposit insurance – whether £85,000, €100,000 or $250,000 – backs that stability up to a point. The Bank of England is considering increasing the level of insurance but, as its governor Andrew Bailey said recently: ‘there is no free lunch’.

Other contradictions revolve around the concept of ‘high-quality’ assets. ‘Risk-free’ government bonds in the US have gone from yielding very little to around 5% within a year. Yes, they remain liquid, but such a price swing threatens stability.

Hybrid model tested

In accounting terms, the hybrid model under both IFRS Standards and US GAAP, which allows some assets to be held at (amortised) cost while others are fair valued at market prices, is being tested again. Assets carried at cost are meant to be held-to-maturity (HTM) for the purpose of collecting cash flows. It depends on the business model: no bank has a model of selling HTM assets, yet they may be forced to do so under stress.

Hence, the importance of those notes to the accounts that give a fair-value measure of HTM assets. (HSBC’s notes 12-14 are good example of this.) It also confirms the wisdom of applying an expected loss model to loans booked at cost.

With fears of a credit crunch, let’s be thankful for the small mercy that most banks have balance sheets strong enough to withstand the best-known unknowns.