After my rant yesterday about APRA’s change to the LGD floor – a subject I am yet more convinced on – I was reminded of where APRA are doing it right.
I was browsing through the latest edition of the Global Risk Regulator (sorry – behind a paywall – and a steep one at USD960) when I came across an article on how the German regulator is “trailblazing” in Europe in terms of bank liquidity management by allowing banks to hold liquid assets not by a formulaic method, but by calculating their requirements for themselves. Of course, there are still lower limits in place (the German regulators probably could not help themselves) but these are at least set by the banks and approved by the regulators.
This may be trailblazing in Europe (the German formula for liquidity looks positively Byzantine) but APRA have been using this method for years – as a quick look at APS 210 and its predecessor, Prudential Statement D1 (from April 1998), will confirm. This is one where I feel APRA has it right – if the bank can model liquidity and convince the regulator that the model is sufficiently conservative, then they should be able to use the model. If not, and many smaller ones do not have the necessary expertise, then a simple formula that conservatively covers the risk is sufficient. In the case of APS 210 the formula is to have at least 9% of liabilities (as defined in the standard) in what are known as HQLAs (high quality liquid assets – also defined in the standard). The rest of the standard is principles based – so discussions on maturity mismatches and policy structures can be done bank by bank, taking into account their uni