The author has been described by News Ltd as an "iconoclast", "Svengali", a pollie's "economist muse", and "pungently accurate". Fairfax says he is a "Renaissance man" and "one of Australia’s most respected analysts." Stephen Koukoulas concludes that he is "85% right", and "would make a great Opposition leader." Terry McCrann claims the author thinks "‘nuance’ is a trendy village in the south of France", but can be "scintillating" when he thinks "clearly". The ACTU reckons he’s "an enigma wrapped in a Bloomberg terminal, wrapped in some apparently well-honed abs."

Tuesday, June 28, 2011

Is it riskier lending to businesses or households?

Are Australia’s banks overexposed to housing vis-à-vis business lending, as so many pundits claim? Does the regulator madly tilt the odds in favour of the former? And should banks have a softer, social objective to extend more finance to businesses as opposed to residential borrowers, since, it is argued, the latter are ‘unproductive’? I hope to address all these questions today. [Read on below the fold]

Australia’s banking regulator, APRA, is regarded as being a far-sighted and prudent agency in comparison to many of its overseas peers. (I’ve previously made the case that there was as much providence as prescience in Australia.). It is certainly true to say that over the years APRA has placed Australian banks under tremendous pressure to adjust practices it considers to be unnecessarily hazardous.

APRA works hand-in-hand with Australia’s central bank, the RBA, which until 1997 also had responsibility for banking supervision. Following the findings of the Wallis Inquiry, APRA was created, and the prudential supervisory responsibility previously held by the RBA transferred to it. The idea was that the RBA had too much on its plate—too many objectives once one accounted for its central role of keeping prices stable via management of economy-wide interest rates. There is much to be said for this logic, which is actually a rather unique regulatory set-up in a developed world context (most of our contemporaries have their central banks, like the US Federal Reserve, do pretty much everything).

Having said that, the RBA does, in fact, continue to exercise regulatory responsibilities in the banking domain through two chief means: first, the RBA controls the so-called “liquidity facilities” that all Australian banks and building societies have access to, and tap from time-to-time. These facilities allow them to “put” their assets to the RBA in a liquidity crisis and obtain funding from it (ie, the RBA lends to them against this collateral).

The second way the RBA exercises influence over the banking system is through its so-called “financial stability” remit. Inside the RBA there is a team of very highly qualified economists—led by Assistant Governor-in-waiting, Dr Luci Ellis—who spend all their time monitoring the manifold risks that have the potential to threaten the viability of Australia’s financial system.

Since most of this system is dominated by the banks, the RBA’s team does an enormous amount of analysis on different credit risks, default rates, the exposure of banks to the housing and commercial property sectors, and so on. All of this work comes to fruition bi-annually in a detailed, 100-plus page study known as the Financial Stability Review.

Neither APRA nor the RBA get rewarded for banks making big profits. They are typically only recognised when they help us avoid some catastrophe, such as during the GFC. This tends to make both institutions exceedingly conservative.

With that background in hand, we can turn to my first question: what do APRA and the RBA consider to be riskier—housing or business lending? This is actually easy to answer through the formal “risk-weights” that APRA forces banks to apply against different types of loans. The risk-weights in turn determine how much capital, or money, a bank needs to hold against every dollar of residential or business credit they extend. These weights are, therefore, a proxy for the probability of loss when you get into the lending game.

For all non-major banks, business loans attract a 100% risk-weight whereas residential credit gets half this amount (ie, a 50% risk-weight). APRA applies a much lower risk-weight to residential credit precisely because the proven empirical risks associated with this form of finance indicate that it has lower probabilities of loss than conventional business lending. If the opposite were true, and home loans were a more hazardous credit, then the risk-weights would work the other way around.

The major banks get the benefit of even lower risk-weights (and, thus, a lower cost of capital) compared to their competitors because they are permitted to use an “internally-determined” risk-weighting system under what is known as the “Advanced Route”. This means that APRA allows them to set their own risk-weights subject to a range of sophisticated rules.

You can see in my first chart, which was extracted from the RBA’s latest Financial Stability Review, the various risk-weights the major banks apply to different credits. On average, the risk-weight on home loans is around one-quarter the risk-weight on business loans to large companies. Since banks can hold less capital against a home loan, this is, in theory, a relatively more profitable activity for them to undertake. That last statement ignores, however, the impact of bank margins. Margins on business loans tend to be much higher than residential loans (as seen in higher business lending rates) in recognition of the former’s risks and the relatively greater labour intensity associated with assessing business credit risk. When all is said and done (ie, adjusted for margins), the actual returns on the two forms of finance are likely to be similar.


A second powerful quantification of business versus residential risk is supplied by examining bank default rates (ie, “non-performing” loans) and associated losses (or impairments). Once again, the Financial Stability Review delivers in spades here.

The next two charts cast into sharp relief the fundamental problem bank regulators have with business lending. In the right-hand-side panel of the fist chart, the RBA has broken-out Australian banks’ non-performing loans by type. Non-performance (ie, loans in arrears) is expressed as a percentage of the total loans in that category.


So, we can see that over 3.5% of all business loans (blue line) are non-performing (ie, in default). In comparison, only around 0.7% of all residential borrowers (orange line) are materially behind on their repayments. That is, default rates on home loans are a fraction of arrears on business loans.

Yet the real clincher is in the total dollar value of these non-performing loans. Today Australian banks tend to have a rough 60:40 split in favour of residential/business lending. Before the 1991 recession, the ratio was the reverse, with the majority of bank credit going to businesses. These exposures almost caused the collapse of two of the majors, ANZ and Westpac. Since that time, the majors have acknowledged the higher risks inherent in their business portfolios, and reversed course, with the encouragement of regulators.

Even so, look what once again happened during the GFC. Compare, in particular, the first two panels of the next chart. The far left-hand-side panel shows the dollar value of impaired housing loans (ie, red line). It looks like impairments are a little higher than $2 billion in total, which is trivial when you consider that there are well north of $1 trillion worth of residential mortgages out there. Now contrast this against the middle panel, which shows the banks’ impaired business loans based on the latest data. Total business impairments look to be over $22 billion, or a remarkable 10x higher than residential impairments.


Now ask yourself this question: would you have Australian banks ramp up their exposures to the middle column (ie, business lending) in preference to the first column (ie, residential lending), as folks like Steve Keen suggest would be a smart idea? I don’t think so.

For what it is worth, exactly the same phenomenon asserted itself during the much more acute 1991 recession. My final chart shows the loss rates across CBA’s different credit portfolios in that downturn, when unemployment peaked at close to 11% while mortgage rates hit 17%. Notwithstanding the serious difficulties households were having at the time, the losses on home loans (bottom red line) were a tiny fraction of the losses CBA realised on all other forms of credit.


I want to conclude with some words on the question of whether housing investments are ‘unproductive’. Notionally respected critics employ this argument to suggest that policymakers should discourage it by changing the negative gearing and/or capital gains tax rules. This logic is, however, flawed.

In standard economics theory, the two most basic necessities for a functioning economy and, more specifically, productive labour, are: food and shelter. We can interpret ‘shelter’ as covering both residential property, which is the accommodation required by a functional labour force, and commercial property, which is the accommodation needed for productive businesses.

You can either be a ‘producer’ of food and shelter by investing in, and owning, the underlying assets that supply these services/products by owning farms, homes, or commercial buildings; or, alternatively, you can just ‘consume’ these services/products and not invest in the assets by simply buying food from the supermarket, renting a home, or leasing space in a building.

There is a final option where you can both consume and invest by living on a farm (and eating its produce), owning and occupying your home, or owning and occupying the building in which your business operates as some companies do.

As an investor in agriculture, housing, or commercial property, you derive returns as you would with any other productive asset: via a mix of both growth in the capital value of the asset and income from the sale of the products/services (ie, food, residential accommodation, and commercial accommodation).

In addition to its innate productivity as a form of shelter for workers, academic researchers have shown that housing investments positively impact both labour and business productivity. That is, the higher the quality of housing, the more productive workers and businesses can be. The ABS has also demonstrated that investments in new housing have a very high, 2.9 times economic multiplier. This means that every dollar committed to building homes generates another 2.9 dollars of extra spending across the wider economy.

Finally, the housing market is increasingly fulfilling a new economic purpose: with the advancement of technology such as the NBN, supplying accommodation for businesses as more people work from home. The home is no longer an office for just the self-employed individual, or for secondary employment activity. It is increasingly the workplace of professional service employees in their primary job, who don’t wish to commute long hours to their labour market.