This claim features prominently in the Productivity Commission’s report on “Competition in the Australian Financial System” but the only source I can find is a newspaper article quoting a Morgan Stanley report. Mortgage lending is clearly amongst the more profitable areas of Australian banking but I would be surprised if the real return was as high as the Productivity Commission appears to believe. It is difficult to challenge the number given the absence of detail offered by the Productivity Commission but here is my checklist of things which I think the calculation is missing:
- I suspect the capital part of the ROE calculation is based on a 25% RW but the actual capital an IRB bank is required to hold is higher than this because they also face a capital (CET1) deduction for any shortfall between their loan loss provision and the Regulatory Expected Loss (REL) assigned to these loans. This REL deduction probably adds another 3 percentage points to the effective RW an IRB bank applies
- The 40% ROE is also I suspect based on the current point in time loan loss allowance (say circa 2-4bp per annum). Banks would dearly love to live in a world where the expected loan loss was always this low but the real return on equity for these loans should take account of what will happen to loan losses when the economic cycle turns down. Exactly how bad this loss rate will be is anyone’s guess but I would not be surprised if the loss rates were at least 5 times higher than the current rates and even more if we encounter another financial crisis.
- It is also not clear to me what kind of cost to income ratio is being factored into the calculation
- Charges for liquidity costs associated with funding these assets also not clear
- Finally we have the fact that the overall ROE for banks is nowhere near 40% but residential mortgages dominate their balance sheets.
Tell me what I am missing but something does not add up …