Basel IV does not go far enough
Written by Graham Andersen on 20/10/17, published in AB+F.
The banking world awaits the finalisation of the now termed Basil IV. The most controversial point outstanding and yet to be agreed is a floor on the discount to the standard method of calculating bank capital requirements or risk weighted assets. The floor is to be imposed on those banks using internal risk based (“IRB”) models. Regulators in Europe and the USA are in some disagreement but it does seem that they’ll settle on a discount floor of 25%. or 30%.
Setting such a floor will have significant implications for those 5 Australian banks that now use the IRB method to determine risk weight on assets especially as it effects residential mortgages. Whilst the floor is likely to result in risk weights increasing at the margin and across the board, the failure of Basel IV for the Australian taxpayer is that it does not go far enough in determining appropriate risk weights for residential mortgages on the books of Australian banks especially the major banks.
The banks’ IRB models are confidential and therefore its not possible to do direct mortgage by mortgage comparisons. Nevertheless, its possible to look a few factors to demonstrate the likely effect on increasing bank risk weights. APRA has now set a benchmark for the major banks of a minimum risk weight for the residential mortgage book of 25%. However, if a major bank was to write a standard mortgage with no mortgage insurance at an LVR of 85%, the minimum risk weight under Basel IV could be 37.5% or 56.25% if the LVR was just over 90%. Compared to the 25% average risk weight, these are considerable increases.
Whilst Basel IV can be applauded for increasing the strength of Australian banks and cushioning the financial system against any future requirement for government support, I believe it does not go far enough in identifying loans that should require even higher risk weights than the standard.
There are aspects of residential mortgages that significantly affect the risk and are not considered by the standard method. Aspects that are certainly on APRA’s radar but not formally and publicly identified as risk weight factors that must be accounted for in both the standard method and IRB models.
To nominate the most important mortgage factors, not formally accounted for in the standard method used by Australian banks in determining risk weights; varying periods of interest only, investment loans, new high rise apartments and high risk regions. I have no doubt that these factors are considered by all banks in assessing the loan risk, but not in determining the standard risk weight. Even if APRA, behind the scenes, is requiring larger risk weights for these factors that is just not good enough. If Basel IV sets a floor for IRB banks based on a discount to the standard weights then the standard weights must consider all significant factors to guard against gaming.
More controversial though are macro factors and bank specific factors that also should be considered to both protect against system risk and to increase competition for the benefit of consumers.
“Concentration in the banking book” is a factor that APRA should consider under Basel II requirements in setting or agreeing risk weights. The concentration of residential mortgages on the books of Australian banks especially the majors is extreme by international standards. Yet there is no specific weight for this factor imposed by APRA. The total of Australian residential mortgage loans relative to other loans represents a significant risk for the Australian financial system and this should be addressed in risk weights in a public and formal manner.
“Asset (house) prices” where they are at inflated levels relative to measures such as GDP or household income should be considered in determining risk weights and therefore the level of bank capital. Inflated house prices do not necessarily mean that those prices will collapse but it does mean that if prices do drop significantly loses on mortgages will be much greater and therefore represent a greater risk to a bank’s balance sheet. APRA should formalise a risk weight for asset prices.
“Funding from international sources” needs to be identified as a specific risk factor that must be incorporated into risk weights. APRA has a strong regulatory regime based on a bank managing their short and long-term funding mix as well as the mix of wholesale to retail deposits. International sourced loans represent an additional risk to the total Australian banking system. If international lenders withdraw funds are removed from the Australian system and not just moved from bank to bank within the system. This risk would only be managed by banks sourcing funds from the Reserve Bank and therefore represents a risk to the Australian taxpayer that APRA should consider when determining risk weights for banks.
Basel IV may be an improvement but it’s a long way from final.