'Risk weightings' = risky times for the banks
Every now and then, a piece of data is presented to me that makes me sit up, and gives me a good slap in the face. It happened this week. This graph, perhaps at first appearances is innocuous.
It's not the light blue of sky and dark blue of ocean, but a chart of Australian Deposit-taking Institutions (essentially banks) equity as a percentage of their assets.
Source: Wayne Byres, Chairman APRA, Fortis Fortuna Adiuvat: Fortune Favours the Strong speech, 5-Apr-17.
What does the chart say?
This chart says that leverage in the banking system is incredibly high, with only around 6 cents of equity for each $1 of assets (which we did know). That is, for each $1 of equity, the banks borrow ~$17 to onlend. This leverage is significantly higher than for non-financial companies (although a direct comparison should not be made).
But furthermore, and surprisingly, this ratio is essentially unchanged over the past decade - even after the regulatory changes / increased capital buffers in recent years.
Higher capital holdings implies lower leverage, so how could this graph be correct?
As highlighted by APRA, it comes down to the composition of lending. Essentially, the banks have been increasing the portion of residential mortgages to their total lending pool (with a lower portion of, for instance, commercial loans). Residential mortgages have been considered, from a regulatory perspective, lower risk than, say, a commercial loan. Therefore, the increased capital requirements imposed on banks in recent years, has been essentially offset by having a higher portion of "lower" risk residential loans.
"Lower" risk residential loans
Increasing exposure to one sector is a risk in itself, especially when that exposure is already high (for instance, as at 28-Feb-17, 70% of the CBA's lending book is in investor- and owner occupied-housing loans). And that is irrespective of the risks within that sector.
But the risks in the housing sector are not 'irrespective', and they are increasing. For instance:
- House prices are clearly elevated and continuing to increase, especially in Sydney and Melbourne
- Wages growth is below trend
- There is a high degree of underemployment in Australia
- Mortgage arrears are rising (albeit still low)
- Funding costs are likely to increase because of increasing interest rates overseas, flowing through to mortgage interest costs here (irrespective of the RBA's cash rate setting)
- Personal debt levels are among the highest in the world.
The system now looks more favourably upon the barista at your local coffee shop having $1.2m of residential mortgages, over an operationally sound cash-flow positive small or medium size business that is seeking a loan to grow its business.
APRA is expected to more specifically articulate the measures it will enact to make Australia's banking system 'unquestionably strong' around the middle of this year. (The definition of 'unquestionably strong' is a whole topic in itself, which we won't delve into in Investment Matters today. However, clearly the view is that the banks are not unquestionably strong now, and increased regulation, and capital, will be required to achieve that goal.)
So we don't yet know the specific upcoming changes which will be enacted by APRA, and we can't yet quantify the impact. However, based on speeches this week and discussion regarding the the Basel V global banking regulations (for which the introduction has been delayed), it is fairly clear that risk weightings of residential mortgages are likely to be increased. Other changes may also be enacted.
What are risk weightings?
Simplistically, risk weighting are used to calculate the amount of capital that needs to be assigned to a loan. The higher the risk weighting, the more associated capital needs to be set aside.
Risk weightings for residential mortgages in Australia have already been increased. In 2015, APRA increased risk weighting for mortgages from approximately 16% to an average of at least 25%, effective from 1-Jul-16. There was also less use of the banks' own risk models for the Big-4 - instead using standard metrics, which further increased capital requirements (the smaller banks used and continue to use a standardised approach).
Increasing the risk weightings further, and adding more refinement (e.g. maybe higher weights for investor loans), would better reflect the increasing risks that exist in the residential mortgage sector.
Impact of higher risk weightings
Increased risk weighting flows though, all else being equal, to more capital being required, lower leverage, and lower shareholder returns.
The impact of the 2015 risk weighting increase was (according to APRA) the equivalent of increasing minimum capital requirements for the major banks by approximately 80 basis points. The exact amount varied for each bank, as their lending composition varies. For the CBA for instance, the impact was 95 basis points, or ~$4 billion of extra capital being required. Consequently, it raised $5 billion of capital in conjunction with its FY-15 results announcement.
If the banks are required to raise capital in response to another step-up in the risk weightings - whether it be (depending on the quantum of additional capital required) via underwritten DRP, or a rights offer - shareholders will be diluted. Dilution means the banks' profit will be shared across a higher number of shares, lowering EPS, and dividends per share. The more additional capital required, obviously the more the dilution.
The banks will likely argue for time, so they can generate the additional capital through earnings over a few years. They shouldn't be given it (and probably won't be).
Our view on investing in the banks
We have not invested in the Big-4 banks on your behalf since 2008. This is not so much because we have seen any imminent danger (which we still don't know). It is more because one of our core investment philosophies is to invest when we see meaningful profit growth upside by doing so, with limited downside risk. In relation to the banks, for a number of years now, we have viewed the profit growth upside on a 3-year horizon as not that meaningful, with more downside risk than we have been comfortable with. We also missed the dividend yield 'roll' that the banks went on for a couple of years, but don't consider yield chasing to be a valid long-term justification for investing.
Furthermore, and really the bottom line is, the weighted average forward P/E of the Big-4 banks is 14.2x. We consider there to be better investing opportunities elsewhere - your equity portfolio currently trades on a forward P/E of 10.5x.
This week has really marked a turning point - a real ramp up - in relation to the articulation and acknowledgement of risks in Australia's banking system by Australia's various regulators. Their articulation of the issues, and their preparedness to do something about it, is clear. This week alone, various measures have been enacted (interest-only loans restrictions, along with more rigorous loan serviceability assessments), and warnings clearly articulated (for instance, the RBA's concern about high and increasing debt levels, combined with moderating wages).
Additionally, more systemic changes - for instance to residential mortgage risk weightings - are likely coming.