Financial stability my butt

Advertisement

The September 2011 edition of the RBA’s financial stability review came out last week and as usual I like to take a look at the section on household and business balance sheets. I have previously analysed much of this information as part of recent statement of monetary policy, however the latest FSR contains some updated data.

I have previously stated my concern about the RBA position on the financial stability of the private sector. My main issue is that in all of the recent literature produced by the RBA I cannot find suitable recognition of the fact that private sector debt issuance and private sector asset values are intrinsically linked. As far as I can tell much of the risk analysis performed by the RBA seems to ignore this point and at a time of historically low credit issuance this is a major concern.

The main crux of my concern is covered in the following paragraph in reference to the 2009 first home buyer cohort.

The issue I had with that analysis was that the RBA seemed to be saying that the cohort who took on “riskier loans” will not get into trouble because house prices will continue to rise. This may end up being the case in the long run, but what the RBA doesn’t seem to be taking into account is the “cause and effect” in their analysis. The reason house prices continued to rise was because of the demand of the first home buyers who took on those risky loans, that is, the risk mitigator mentioned by the RBA is intrinsically linked to the cause of that same risk

Advertisement

More recently while analysing a speech by the RBA’s Malcolm Edey I stated:

As we have discussed many time previously on MacroBusiness, one of the major reasons that asset prices grew was because of the available credit to drive those values upwards. Tighten that credit stream and the value of the asset that support the debt position is likely to fall. As the effects of the stimulatory environment wear off and , as mentioned in the speech, the economy re-adjusted to a less expansionary position, then the debt to asset ratio of the private sector has the potential to worsen. Given that Malcolm Edey himself is stating that the economy is about to enter a phase of lower credit expansion, I suggest therefore, that the true test of APRA’s prudence is ahead of us, not behind. A slower growth environment is very much the thing that is going to test whether the bank’s equity-to-asset ratios can actually support their loan books. As I have mentioned in a couple of my previous posts it is not until we see debt growth fall back to a more sustainable level, that is in line with wages, that we can really measure the stability of the economy.

Given we are already seeing falling housing prices with debt growth at 6% YoY, Mr Edey may have gotten a bit ahead of himself.

In recent weeks we have seen some evidence that the effects of slowing credit issuance are now making their way into the broader economy in what seems to be a capitulation of the labour market after many months of labour hoarding. It obviously must be noted that unemployment figures are prone to error, but the trend in the data is quite strong and other leading indicators support this trend.

Advertisement

The new FSR household and business balance sheets section begins with:

The household sector is continuing to consolidate its financial position. Over the past year, the household saving rate increased further and the debt-to-income ratio declined slightly. Given that household net worth declined in the wake of renewed volatility in global financial markets, the prevailing mood of caution appears unlikely to lift in the near term. While households in aggregate are managing their debt levels well, the mortgage arrears rate drifted up over the first half of the year. However, this mainly relates to loans taken out prior to 2009, when banks’ lending standards were weaker; newer loans are performing well despite the increase in interest rates last year. The business sector is also experiencing mixed conditions: the mining and related sectors continue to benefit from the resources boom, while other sectors, including retail, are facing headwinds from subdued domestic household spending and the high exchange rate. Measures of profits and business confidence have therefore diverged between sectors. Having deleveraged considerably, the business sector is in a better financial position than it was several years ago, but its demand for external funding remains weak.

It must once again be noted that the household savings rate is poorly named. It is in fact the ratio of household income that isn’t spent on consumption items. You cannot conclude from the ratio that people are actually “saving” the money. In fact the data suggests they are paying down debt, which in itself sounds positive but in the absence of an offsetting in-flow from another sector ( government and/or external ) it is actually deflationary when the private sector attempts this en masse. This issue is more apparent when there is already a high-level of household indebtedness and therefore no non-asset based savings buffer. With the government aiming for a surplus budget combined with Australia’s external sector deficit this “tightening of the belt” is actually having an adverse effect on the private sector. We have recently seen this manifesting in a number of areas including sluggish retail trade and house price deflation.The issue with that last point is that a large amount of household wealth is actually tied up in housing, which means that while the private sector is paying down debt in an attempt to secure its finances there is actually a risk that it is making itself poorer by doing so.

Advertisement

I am not going to go through the entire document, I have included it and a supplement at the end of this post. However there are a few sections that I think are important to highlight. Firstly some background.

The financial position of the household sector continues to be shaped by a more cautious attitude to spending and borrowing, as evidenced by the considerable increase in the household saving rate (Graph 3.1). After trending up since the mid 2000s, the household saving rate rose further over the past year, reaching 101⁄2 per cent of disposable income in the June quarter. It is now at levels similar to those last seen in the mid 1980s.

One financial counterpart to the higher saving rate has been a substantial slowdown in the pace of household credit growth. Growth in household credit continued to moderate over the past year, declining to 4.5 per cent in annualised terms over the six months to July. There has been a reduced appetite for most types of debt. Personal credit outstanding declined over the same period, reflecting a recent contraction in credit card debt as well as the ongoing decline in margin lending. The value of outstanding margin debt has more than halved from its peak in late 2007, as volatility in share markets has made equity investments less attractive. Similarly, annualised growth in housing credit eased from 6.7 per cent over the six months to January to 5.2 per cent over the six months to July (Graph 3.2). Growth rates of both owner-occupier and investor housing debt have moderated so far this year. The flow of new borrowing has also moderated, with the value of monthly housing loan approvals declining by 7 per cent since late 2010.

While mortgage refinancing activity has picked up since early 2011, surveys suggest that this is mainly due to households switching to cheaper loans – amid increased competition in the mortgage market – and consolidating debt, rather than taking out larger loans. As debt accumulation has slowed in recent years, the rate of housing equity injection has increased (Graph 3.3).

This should all be well known to MB readers, Data Sword and Houses and Holes have been doing a great job of presenting all of this data and providing the commentary on the disleveraging public. More from the FSR:

Advertisement

Putting the slow rate of borrowing and solid income growth together, the ratio of household debt to annual household disposable income fell modestly from a peak of 158 per cent in mid 2010 to 154 per cent in the June quarter (Graph 3.6). This ratio has now been broadly unchanged since 2006. After rising through 2010, the ratio of household interest payments to disposable income also declined slightly in the first half of the year, to 11.7 per cent. Despite being around 2 percentage points lower than its September quarter 2008 peak, it is still relatively high by historical standards.

Households might have been motivated to become more cautious in their financial behaviour in part because their net asset position is no longer following its past trend of rapid expansion. Household net worth is estimated to have declined slightly over the first half of 2011, compared with annual average growth of almost 9 per cent over the past decade. A further decline is likely in the September quarter, given that share prices have fallen. The recent weakness has, however, mainly been driven by falls in dwelling prices

You can see from the left hand chart that household indebtedness seems to reach an upper limit around 2006 and has been hovering around that level for 5 years. When the GFC hit in 2008, households actually attempted to de-leverage, in doing so the economy slowed considerably. In order to reverse this the government sector went into deficit and also enacted stimulatory fiscal and monetary policy. This had the effect of pushing household indebtedness back up towards previous levels but you can now see that it is once again falling, which in turn is once again slowing the economy. You can see from Chart 3.7 that the de-leveraging of the private sector during the GFC caused the value of non-financial assets to fall. That is, the private sectors non-financial asset values are dependent on the rate of credit issuance. I have talked about this previously. This demonstrates that any attempt to dis-leverage by the private sector in the absence of government sector support will lead to a loss of net worth. In other words “in attempting to save you actually get poorer”.

This problem is most apparent for those who are most indebted. This is normally new entrants to the housing market, in the case of Australia this is quite a large cohort because the most successful method used in the post GFC period to re-initiate credit growth was a grant that targeted first home buyers. This had the desired effect, but it also created an entire generation of young Australians who are now in quite a vulnerable financial position because, as I said above, asset prices are supported by the rate of credit issuance which has now fallen to 50 year lows. It must also be noted that this cohort took out loans when mortgage rates were exceptionally low and housing market was very liquid.

Advertisement

My long standing belief is that the RBA has consistently understated the risks created by the post-GFC lending environment and this latest FSR certainly doesn’t address my concerns. Below is a section that demonstrates my issue:

Comparing the performance of housing loans across age cohorts, it appears that most of the recent increase in the mortgage arrears rate has been due to loans that were taken out prior to 2009. Loans that were extended towards the end of earlier periods of strong housing price growth and weaker lending standards have generally been the worst performing in recent years. Housing loans made since 2009, including for many first-home buyers, have been performing better than earlier cohorts, despite the fact that these borrowers are typically facing higher interest rates than at origination (Graph 3.10). This likely reflects an improvement in loan quality due to a tightening in lending standards after 2008. In particular, the share of new low-doc housing loans (where borrowers can provide less evidence of debt- servicing ability than normal) has fallen considerably since 2008 (Graph 3.11). The share of new loans with loan-to-valuation ratios above 90 per cent also fell significantly in recent years, though it has edged up over the past year as competition in the mortgage market has intensified.

Even though loan performance deteriorated over the first half of the year, the overall mortgage arrears rate in Australia is still low by international standards (Graph 3.12). Looking forward, the experiences of those countries that currently have high arrears rates, as a result of high unemployment or an excessive easing in lending standards in earlier housing price booms, are unlikely to be the model for future outcomes in Australia. First, housing prices in Australia did not grow especially rapidly in most parts of the country in the period since 2004, although Queensland and Western Australia were exceptions at various stages. The decline in housing prices recently has been modest compared with the sharp downturns seen in some cases overseas. Second, even before their tightening in 2009, lending standards in Australia had not eased as much as in some other countries. The near absence of sub-prime housing loans in Australia relative to the United States is one prominent example. Australian lenders also assess mortgage serviceability at higher interest rates than those prevailing at origination, a practice not always followed overseas. Third, as noted above, a large share of mortgage borrowers in Australia make excess repayments. This increases the resilience of households to shocks, relative to countries where it is less common to do so. As well as providing a cushion against changes in borrowers’ financial circumstances, excess repayments increase the distance between the remaining loan balance and a property value that could be lower in the future, making negative equity positions less likely. Finally, the labour market in Australia is in better shape than in many other countries, and its prospects are also more favourable given the macroeconomic outlook.

There are a number of problems with this analysis. Firstly graph 3.11 clearly shows that during early 2009 25% of loans issued to owner-occupiers had an LVR greater than 90%. This was a period in which the government was providing substantial grants and interest rates were abnormally low. I simply cannot understand how the RBA could be claiming that there was a “tightening in lending standards after 2008” given that it is quite obvious from the data that large numbers of very high LVR loans were issued at this time using a government grant as a major component of the deposit.

Advertisement

Secondly the claim that “housing prices in Australia did not grow especially rapidly in most parts of the country in the period since 2004” is absurd. A quick glance at any capital city house price chart will tell you that this claim is totally false. Take Melbourne for example, which has had two periods of 20%+ growth since 2004.

Thirdly, the comparisons to other countries are strawman arguments. Considering yourself to have slightly better lending standards than a random set of mostly unspecified countries, some of which have suffered economic collapse due to their housing market, is not a statement that befits a document of this type.

Fourthly, the statement regarding the decline in housing prices being modest compared to overseas does not stand up to scrutiny. In aggregate this may be true, because Sydney the largest market is still in positive territory. However the Brisbane and Perth markets have fallen nearly 7% and 5% respectively over the last 12 months. Those are not “modest” declines in asset values over that short period of time.

Advertisement

Fifth, although the statements about having a buffer in their home loans is true, it ignores the vintage of the most vulnerable loans. A loan issued in 2009 is at most 36 months old, meaning that the size of the additional payments made would be relatively small in comparison to the total value of the loan. As I have mentioned previously, there are a number of areas in Australia where prices have fallen back below 2009 prices. It is likely that in some areas the fall in asset value is greater than any buffer that could have been created in the loan over the period, so although these people may be ahead on their loan repayments their actual mortgage is underwater. It must also be remembered that there are significant sales costs associated with housing, any analysis of “buffering” must take this into account.

Finally, this analysis ignores the fact that over the last 10 years there has not been a period of sustained house price falls. Statements about employment conditions ignore the fact that employment is a trailing indicator and that employment is tightly coupled to credit issuance and therefore not a risk mitigator.

There is also a statement made in the document about bankruptcies:

Advertisement

As for arrears rates, other indicators of financial stress do not suggest that household financial circumstances have deteriorated markedly. Rates of applications for property possession picked up in most regions in the first half of 2011, consistent with the deterioration in loan performance. They remain below earlier peaks except in Western Australia and south-east Queensland, where the rates of applications for property possession are closer to their recent peaks. The nationwide rate of bankruptcies and other personal administrations declined further in the first half of 2011, and is now well below the peak in 2009, though this also tends to be a more lagged indicator of household financial stress.

Although not the same data this statement does seem at odds with recent information from ASIC about company insolvencies:

The most recent release of ASIC insolvency statistics reveals that 2011 is on pace to setting record highs in the number of corporate insolvencies.

According to Dissolve, a business specialising in liquidations, the number of companies entering some form of insolvency administration in the calendar year to July 2011 is the highest ever.

The number of insolvencies in the month of July was 921. While the figure is down from 1,027 in June, it was still the highest July figure ever recorded. New highs were also set for the months of March, April and June.

Statistics have been kept in the current format since 1999, and 2011 has seen a number of “highest ever” numbers.

Advertisement

I am not sure how to interpret the differences. It is possible that recent house prices rises have created a buffer that have allowed people involved in insolvent companies to liquidate assets while not falling into bankruptcy or , as the document states, this is simply a lagging indicator. I also note that a discussion of rising insolvencies is missing from the business sector section of the document.

In conclusion, the RBA has once again failed to convince me that they are fully aware of the risks presented by the high levels of household debt in this country, and more specifically the risk presented by the first home buyers grant boost. Their analysis once again fails to take into account of the dependency between credit issuance and non-financial asset values in the private sector and the risk presented by an uptick in unemployment after a period of de-leveraging so soon after a period of debt stimulation targeted at young people. The market will behave very differently in 2011 when compared to 2008 because the government’s response to the GFC created a large cluster of people in the market that are at the upper end of the risk spectrum. This is not something that existed when the first GFC hit, and so the effects of a credit-led slowdown in the economy will have a much more immediate negative effect this time around because such a large number of people now have limited capacity to deal with it.

Advertisement