Cross-posted from AB+F:
Graham Andersen, the founder of Morgij Analytics, has long been critical of the banks’ refusal to release transactional data to the market. Here he outlines why this absence of data is a problem for APRA when it comes to comparing Australian banks with their offshore peers.
The prudential regulator has released its Insights Issue Two for 2016 with the lead article supposedly providing an insight into “International Capital Comparisons” of our major banks.
Disappointingly, the report provided little new insight but rather offers up a mish mash of previous dubious analysis from July 2015 that uses unsubstantiated assumptions based on a hubris that APRA is the greatest regulator on the planet.
Not surprisingly, APRA has concluded – and the major banks agree – that in accordance with the Financial System Inquiry’s recommendations, they are capitalised in the top quartile when compared to their international peers. But they’re not. Why is this so important?
Australia’s major banks are ‘too big to fail’ and as such carry an implicit guarantee of the Australian taxpayer which is acknowledged by credit rating agencies and the market alike.
In order to protect the Australian taxpayer from the effect of banks retaining the profits but socialising the losses, the FSI recommended and the government accepted that the banks should be ‘unquestionably strong’ and in the top quartile when compared to international peers on capital to risk-weighted assets.
Whether one uses the international comparison generated by the Basel Committee in their Quantitative Impact Study or APRA’s unsubstantiated analysis is critical since they determine how many billions of dollars in capital that the major banks either have to raise or need not raise. It’s no small issue.
One would expect that in the circumstances – and given the risk that the taxpayer carries in the Australian financial system – that APRA would be both precise and conservative in the extreme in analysing and comparing our major banks’ capital to international peers. But they’re not.
Confidence in the system
The 2015 APRA report does not provide specific data on comparing capital calculations and the regulator admits in the report that this is not really possible due to a lack of disclosure of bank internal capital models calculation methodologies and each jurisdictions supervisory powers to apply different assumptions to minimums which have the effect of increasing risk-weighted assets.
Consequently, APRA justifies adjustments based on generalisations and basically hearsay which unsurprisingly benefit the comparison. It’s just nowhere near good enough. APRA must rely on facts and disclose those facts so that the market and taxpayers are both informed and have confidence in the system.
My simple comparison, which is proof of the dubious nature of the international comparison analysis, is to look at the leverage ratio.
If comparatively APRA were a tougher regulator requiring much higher risk-weighted assets for our major banks, then compared to international peers the leverage ratio (ratio of capital to actual assets) would be high up the comparative table. It is not. Australian major banks are about average.
To be clear, the higher the risk-weighted assets, the higher the leverage ratio. So, if you are comparing banks on a like-for-like basis, a regulator that was stricter and required higher risk weighted assets – than a regulator in another jurisdiction – it would be expected that comparatively the leverage ratio would be high because the difference between actual assets and risk-weighted assets is smaller.
Whilst the major banks have been misrepresenting their international capital comparisons for years, it is beyond the pale that APRA supports this continued misrepresentation at great cost to taxpayers and ultimately bank shareholders.
The regulator must prove that their report of July 2015 is unquestionably correct or throw the whole thing out now. Its credibility is at stake here and sticking to an indefensible position only undermines the whole financial system.
