One commonly sees a tension between the private sector’s willingness to disclose information and embrace regulation, and government’s desire to impose it. And then one frequently finds another friction between the public’s appetite for information and government’s propensity to provide it. In today’s ever more complex and integrated world, Australia should be thinking of information disclosure and regulatory architecture as not an impediment to business, but rather a key comparative advantage. Unfortunately, few people seem to recognise this.
In early 2009 I was chatting with one of the government’s most senior economic advisors about the not-so-long-lived ‘RuddBank’ proposal. At the time, RuddBank was in full flight.
For those who have forgotten, the government had announced that it would commit $2 billion of taxpayer cash combined with $26 billion worth of taxpayer ‘guarantees’ to buttress the ailing commercial property sector.
The unstated motivation for this was to bolster the major banks, and avoid a repeat of the near-disastrous travails experienced by Westpac and ANZ in 1991 when their commercial property exposures brought both banks to their knees.
But a lot had changed since 1991, with the big banks wisely cutting their portfolio weights to the riskier commercial property and business sectors, which are highly ‘pro-cyclical’ (ie, volatile). At that time, around 60% of bank lending was to businesses with the remaining 40% to housing. Following the searing default rates and ensuing losses on business and commercial property loans during the 1991 recession, which contrasted with the tiny losses experienced by residential borrowers, the major banks reversed these ratios. (Refer to this analysis of CBA’s losses on business and residential lending during the 1991 downturn.)
These very different risk profiles are one of the reasons why APRA applies a much higher 'risk-weighting' (or cost) to business lending vis-à-vis residential credit notwithstanding the recent complaints by one leading ‘business banker’, which the AFR saw fit to voluminously air in two feature-length articles. You might recall that the business banker in question was whining about the fact that APRA imposes higher charges on his products, which are more expensive for banks to offer than residential credit. What was most striking about the AFR’s analysis is that it completely overlooked the Governor of the RBA’s sobering remarks in response to this banker’s claims. When Glenn Stevens was asked why business lending was disadvantaged vis-à-vis residential credit from a regulatory perspective, he responded:
“I think it is an economic fact that lending for mortgages has historically been far safer than for other things. That is why the Basel process downweighs the weighting on mortgages when you are calculating the capital charges. Certain mortgages in America, of course, proved to be much more risky than had been assumed. But, leaving that aside, in this country certainly it has been and remains a very safe form of lending. It is true that for a dollar of capital if you are just earning the same returns you can do more on housing. This is why you charge more for business loans. This is why the business loan has a risk margin built in which is considerably higher than a risk margin on a housing loan, so the banks compensate for that in the pricing…It is true that the capital charge system does discriminate, but that is what it is supposed to do because mortgage lending is less risky and that is why a business loan costs more to get.”
Anyways, back to RuddBank. I was told by the abovementioned official that the GFC had tipped the government’s hand, and they had no choice but to seek to bail-out the commercial property sector in this manner. And I was very confidently advised that this was just the tip of the iceberg—the government was going to be compelled to do a helluva lot more as the crisis unfolded.
While I understood the need for taxpayer guarantees of deposits and bank liabilities, and had been the joint architect of the idea to supply liquidity to the securitisation market, I was suspicious of this public bailout of one small sector. RMIT had put the total value of commercial property in Australia at around $252 billion—that is about 18% the size of the sharemarket, or about 7% the value of all private housing.
So I asked the official what I thought was a very simple yet important question: prior to committing nearly $30 billion of taxpayer cash to bailout commercial property investors, what hard analysis had the government undertaken of the risk that a fall in values posed to the banks? More specifically, I asked whether they knew what the major banks’ current loan-to-value ratio (LVR) was against these assets. If the LVR was very high, you might be understandably concerned that a sudden fall in prices would trigger large loan losses, which could in turn feed into more price falls, and so on. If, on the other hand, the LVRs were mostly very low, the banks would be able to withstand substantial price declines with the equity holders, not lenders, bearing the brunt of the losses as they should rightly do. I was, to say the least, very surprised to hear the official say that they simply did not know. The banks could tell you, I responded—after all, they have the data! Flummoxed, this individual said that they did not have access to this information. What about APRA, I asked. We cannot make head or tail of APRA’s data, I was informed.
Given the wide media coverage of RuddBank, I had already done some preliminary digging and knew that Australian banks’ aggregate loan exposures to commercial property were only about $120 billion of the total $165 billion worth of outstanding debt. You can contrast this with about $1 trillion worth of residential mortgage debt. Speaking to analysts and industry experts I was also told that the banks’ LVRs were low, and likely no greater than around 30-50%. Even in the face of a massive 50% decline in the value of the assets, this suggested that the banks could on average recover all of their money. Of course, there would be bad news for equity investors. But the banks’ risk looked relatively manageable.
From my vantage, making decisions with such poor information is a most undesirable state of affairs. What made this situation even more incongruous was the fact that the Commonwealth had already committed to guaranteeing the banks’ debts in the event that they defaulted on these loans. This was a smart policy decision that I fully supported. Yet when someone offers to guarantee an institution’s borrowings, they normally demand exacting detail on their assets, liabilities and financial activities.
The government would have been also quite entitled to ask the banks to explain exactly what they were doing with the billions of dollars of capital they were raising via the taxpayer guaranteed loans. Was this being used to lend to businesses and households in a time of tightening credit, as one would hope, or was there a risk that CEOs would seek to stockpile some of the cash to pursue, say, a risky offshore expansion strategy that had almost zero relevance to the national economic imperatives?
When I enquired about this, I was told that the government was getting little new data from the banks even after providing the guarantees. Disturbed by this finding, I suggested that the Treasury and Prime Minister’s office should be seeking at least fortnightly or monthly data feeds that addressed the risks outlined above, and showed policymakers exactly what the banks were doing with the taxpayer-backed money. I was asked whether I could prepare a detailed ‘data request’ for the banks stipulating all of the information that the Prime Minister and Treasurer could reasonably expect to see so long as taxpayers were insuring their loans. Naturally, I was happy to help out. And speaking to various regulators in the period since, I understand that the government took up my advice to the nontrivial consternation of the institutions that were being forced to now transmit much more granular information. The truth is that this is the minimum they could expect. It is but prudent risk management.
This brings me to two other related points. In today’s digitised world, the Australian government should be using its information collection and dissemination abilities as a point of comparative advantage. Yet, as Nicholas Gruen has also observed, it is awfully difficult getting data on some dynamics that could shed valuable light on the way in which the economy is working.
The RBA has recently highlighted the need for monthly inflation figures. Let me offer two further examples. All Australian lenders collect very accurate information on the repayment records of their borrowers. They closely track borrower default rates, since if they miss a repayment it directly impacts the lender’s return. Aggregated across all loans throughout the economy, default rates offer an exceptionally important insight into the sustainability of leverage (ie, the ability of households and businesses to repay their debts). Mounting default rates in the US were one of the first signs of a looming credit crisis. And yet while all this information exists, it is rarely published. Every so often the RBA will oblige with a little chart on system-wide default rates, or perhaps some higher resolution analysis of default rates across different sectors of the market. But for some reason they do not release the underlying data, or publish a regular time-series.
Now I have raised this matter with them directly, and I believe that the very bright folks in our central bank are working on the problem (or encouraging APRA to do so). But since canvassing this with the RBA over a year ago, I have seen no changes as yet.
We do get some default data from ratings agencies like Standard & Poors. The problem is that this information only relates to securitised loans, and therefore covers just 15% of all mortgages. The demise of the securitisation markets during the GFC has also rendered the S&P default rates unreliable since the rise in the seasoning (or age) of the loans outstanding--given there have been far fewer new loans added to the pool--have polluted the time series since the probability of default changes with the life of a loan.
I guess what frustrates me about all of this is that it would be quite straightforward for APRA and the RBA to immediately publish 30, 60 and 90 day default rates on all bank-settled loans in Australia (on both a ‘calendar’ and ‘dollar day’ basis, for the nerds). These data could be further divided up into LVR bands (or deciles) to give one a better feel for the distribution of risk. This would transform the financial market’s understanding of bank risk, and the debt servicing capabilities of businesses and households. Rather than reading the typically misguided missives in the media about household debt, we could actually have an informed debate about the ability of borrowers to meet their repayments.
Another surprising oversight is that, to the best of my knowledge, we cannot get regular data on the number and value of home loans settled across Australia every month. Today the ABS only reports housing finance ‘approvals’. The problem with this is that approvals are quite different from settlements. Borrowers sometimes apply for multiple loans, as occurred during 2009 when banks’ credit criteria were changing and getting qualified for a loan was uncertain. And if lenders do vary their credit standards, there can be significant drop-off rates between initial approval and final settlement. The amazing thing is that lenders know precisely what loans they are settling, since this is the money that goes out the door. Accordingly, the data exists. Finance ‘approvals’, on the other hand, are inherently less certain. Over and above multiple applications, rejections, and withdrawals, there are often large discrepancies between the amount of money a borrower initially applies for and the sum they eventually get.
APRA and the RBA can easily access all of this information. It is time they and the ABS made a better effort at enhancing our understanding of the changing nature of Australia’s financial risks.
Going forward, another idea would be to establish a National Electronic Credit Register (NECR), which is something that I recently proposed.
If you think about it, credit is effectively an over-the-counter (OTC) contract. There is no centralised exchange novating the relationship between the parties as we see, for instance, with securities listed on the ASX. And as we discovered during the GFC, one of the profound shortcomings associated with OTC markets is that they effectively eviscerate transparency. The only people who know what is going on are the counterparties.
Now in Australia, APRA and the RBA collect a great deal of ex post facto credit data. But this is normally aggregated information and does not necessarily tell them anything about the individual loan-by-loan risks. It also does not necessarily furnish them with any insights about the ex ante, or before the event, credit assessment parameters employed by lenders.
The establishment of NECR would presumably be very straightforward. All Australian lenders have electronic lodgment processes and there are standardised communications formats that allow lenders to communicate with one another (in the mortgage market this is known as LIXI (or the “language of lending”)).
APRA, the RBA, and ASIC (to cover non-banks) could, therefore, simply insist that any licenced entity involved in the creation of residential and business credit sends NECR a simple little data packet upon the settlement and, notably, discharge (ie, repayment) of every single loan.
As with the other initiatives suggested above, this could transform our understanding of the financial system's risks, which is presumably important in a post-GFC world.
