Housing and household leverage

Housing accounts for the largest share of household assets. It serves two important functions for households: acting as a store of wealth, and producing a flow of housing services that households consume.

Since 1997, household leverage has increased markedly from debt equivalent to around 0.8 years of gross disposable income to around 1.5 years of income in 2008 — household leverage has since stabilised at around this level.12 Australia’s trajectory and current level of household leverage is similar to that in some other advanced economies.13

Since the Wallis Inquiry, over 90 per cent of the increase in household credit has been due to borrowing for housing; investor housing credit has accounted for one-third of the increase in total housing credit over this period.14

Housing finance

There is little evidence of a shortage of housing finance. Although lending conditions tightened following the GFC, this only partially reversed the easing of lending conditions that occurred over the preceding couple of decades. Over that period, product innovation increased the range of, and households’ access to, loan products. In general, households are not constrained from borrowing amounts that they can reasonably be expected to repay.15

Demand for housing finance largely reflects purchases of existing (rather than new) housing by households wanting to upgrade or downgrade, by first home–buyers or by investors.

Since 1997, around 10 per cent of the flow of housing finance has been for constructing new dwellings.16 New dwelling construction arguably has not been sufficient to meet population growth over the past decade.17 The Australian housing market has a number of longstanding structural features that inhibit supply responsiveness to demand-driven price rises. These include regulatory and zoning constraints, inherent geographical barriers and the cost structure of the building industry.18 These issues are out of the scope of this Inquiry.

The increase in housing debt over recent decades has been facilitated by households’ capacity and willingness to borrow. Growth in household incomes has allowed households to service larger loans for a given interest rate. Australia’s move to a low-inflation and low–interest rate environment facilitated a one-off shift in the level of household debt. To some extent, households’ greater capacity to borrow has contributed to the increase in housing prices relative to income since 2000.

Taxation and regulatory treatment of housing

Households’ appetite for housing debt also reflects the favourable treatment that the tax and transfer system applies to housing. Returns on owner-occupied housing (including imputed rent and capital gains) are exempt from tax, although this is not unusual by international standards. This makes housing a very attractive vehicle for saving. In addition to the more favourable tax treatment, individuals have an extra incentive to put more of their wealth into their primary residence because of the means test for the Age Pension, which excludes the primary home. This leads to higher allocation of wealth to housing and, for some, an inefficient level of consumption of housing services.

The tax treatment of investor housing, in particular, tends to encourage leveraged and speculative investment in housing. Investors are attracted by the asymmetry in the tax treatment of expenses and capital gains on investor housing. Investors can reduce their tax liabilities by deducting borrowing costs and other related expenses against total income at the individual’s full marginal tax rate.19 However, nominal capital gains, when realised, are effectively taxed at half the marginal rate.

This regime has been in place since 1999; between 1985 and 1999, real capital gains were taxed at the individual’s full marginal tax rate.20, 21 The change in capital gains tax arrangements reduced the tax burden of holding an investment property for shorter periods and has contributed to the growth in investor housing credit and investors’ shift to more leveraged investments over the past 15 years (Charts 3.4 and 3.5). Because of these tax arrangements, owners of residential property have an incentive to repay their mortgage as slowly as possible to maximise the tax deductions they can accrue. Loans with interest-free periods help to maximise these tax deductions in the early years of a loan, although these loans also give borrowers more flexibility with repayments. The tax system, therefore, encourages individuals to take on more risk, which does have implications for risks to lenders.

Chart 3.4: Housing finance (flow of housing lending commitments)

This chart shows the flow of housing finance for investor and owner-occupied housing from 1989 to 2014, as an index with the base being Dec 1998. The chart demonstrates the faster rate of growth in investor housing credit compared to owner-occupied housing credit over the past 15 years.

Source: RBA.22

Although these tax arrangements apply to other leveraged investments, such as equities, investors perceive housing investment as less risky than leveraged securities investment; mortgages are less likely to be subject to margin calls, and there is arguably a widespread and falsely held view that housing prices never fall.23

Chart 3.5: Proportion of landlords with net losses24

This chart shows the proportion of landlords with net losses between 1993 and 2010. The proportion has increased from around 50 per cent to around 70 per cent.

Source: ATO.25

Some submissions argue that the current risk-weighting arrangements favour housing loans, or loans secured by housing, at the expense of business loans. Since the current risk-weighting scheme was introduced in 2008, the share of credit used for housing has increased. However, this has been the trend for the past 20 years or so.26 In addition to this trend, the GFC weighed on business conditions and investment, which reduced business demand for credit. As conditions improved, low market interest rates made the bond market a relatively attractive source of funds compared to bank loans, particularly given the more favourable covenants on bonds compared to bank loans.

The broader effects and risks associated with housing and household leverage

One implication of higher household debt is the extent to which housing finance may crowd out finance for other activities. The banking system generally does not face binding constraints on balance sheet growth from prudential requirements. However, at times, lending by individual institutions may be constrained by such requirements. In addition, banks may choose to restrict lending, overall or to specific sectors, to manage their exposures.

In an environment of rising house prices, such constraints may see bank allocate funds towards housing and away from (unsecured) business loans.27 All else being equal, this would tend to increase the price and reduce the amount of business lending.28 Cross-country analysis on the effects of bank credit on economic growth suggests that credit provided to businesses has a larger positive effect on growth than credit provided to households.29

Housing is also a potential source of systemic risk for the financial system and the economy. Since the Wallis Inquiry, the increase in households’ mortgage indebtedness has been accompanied by banks allocating a greater proportion of their loan book to mortgages; the share of loans for housing has increased from 47 per cent in 1997 to its current share of 66 per cent. A large enough disruption to the housing market could have significant implications for household balance sheets, financial stability, economic growth, and the speed of recovery in household spending and broader economic activity following a shock.

As discussed in the Stability chapter, the FSI Secretariat conducted an analysis of a number of scenarios (Box 5.3). One of the scenarios considered the effect of a shock that resulted in a sharp and prolonged fall in house prices. In this scenario, household wealth would contract and there would be broader and, potentially, long-lasting effects on the economy and financial system. A sharp fall in house prices could push some households into negative equity and would amplify financial distress associated with any broader economic downturn. Deleveraging, combined with lower consumer confidence, would weigh on household consumption and broader economic growth. The extent of the damage to households’ balance sheets would determine, to a large degree, the speed of recovery of household consumption.

An extreme shock of this nature would also affect the quality of banks’ balance sheets and their capacity to extend new credit. This would include business lending, particularly for small businesses — which tend to use housing as collateral. Offshore wholesale funding would be likely to become more expensive and some banks might find it more difficult to raise funds, which would exacerbate pressures on the cost and availability of bank credit. Overall, the deterioration in bank balance sheets would compromise the speed of a subsequent recovery in economic activity.

How policy would react to these developments — to boost demand or stabilise the financial system — would depend on the capacity of Government to use available tools to respond. The strength of the Australian Government balance sheet and the level of interest rates would affect the degree to which both fiscal and monetary policy could provide support for the economy.

It is difficult to quantify the likelihood of such a scenario occurring in Australia. Official stress test results in recent years have found Australian banks are likely to be relatively resilient to most shocks. Historically, banks have mainly realised losses on their commercial property loan portfolios. However, the exposure of the financial system to the housing market has clearly increased over time and, in the opinion of the Inquiry, the systemic risk associated with this trend should be further considered.

The Inquiry seeks further information on the following area:

What measures can be taken to mitigate the effects of developments in the housing market on the financial system and the economy? How might these measures be implemented and what practical issues would need to be considered?

12 The ratio of household debt to annualised household disposable income. Reserve Bank of Australia (RBA) 2014, Statistical Tables: Household Finances — Selected Ratios (E2), RBA, Sydney.13 Reserve Bank of Australia 2014, First round submission to the Financial System Inquiry, page 32.14 Reserve Bank of Australia (RBA) 2014, Statistical Tables: Money and Credit Statistics, Lending and Credit Aggregates (D2), RBA, Sydney.

15 Reserve Bank of Australia 2014, Submission to the Inquiry into Affordable Housing, Senate Economics References Committee, Sydney, page 1.

16 Reserve Bank of Australia (RBA) 2014, Statistical Tables: Money and Credit Statistics, Lending Commitments — All Lenders (D6), RBA, Sydney.

17 National Housing Supply Council (NHSC) 2012, Housing Supply and Affordability — Key Indicators, NHSC, Canberra, page 25.

18 Reserve Bank of Australia 2014, Submission to the Inquiry into Affordable Housing, Senate Economics References Committee, Sydney, page 1.

19 Against total income, not only income from the property.

20 On investments held for more than 12 months.

21 At an average rate of inflation of 3 per cent, a property would need to be held for 66 years under pre-1999 tax concessions to get the same capital gains tax concession as the current arrangements provide after holding the property for one year.

22 Reserve Bank of Australia (RBA) 2014, Statistical Tables, Money and Credit Statistics, Lending Commitments — All Lenders (D6), RBA, Sydney.

23 Glenn Stevens has noted that “It is a very dangerous idea to think that dwelling prices cannot fall. They can, and they have.” Stevens, G (Governor of the Reserve Bank of Australia) 2012, The Lucky Country, address to The Anika Foundation Luncheon, 24 July, Sydney.

24 Net rent equals gross rent less rental deductions, which include interest, capital works and other deductions.

25 Australian Taxation Office (ATO) 2012, Taxation Statistics 2011–12, ATO, Canberra.

26 Reserve Bank of Australia (RBA) 2014, Statistical Tables: Money and Credit Statistics, Lending and Credit Aggregates (D2), RBA, Sydney.

27 APRA imposes constraints, in particular by requiring banks to meet certain ratios with respect to capital relative to risk-weighted assets. Banks face internal constraints from desired returns on equity.

28 Chakraborty, I, Goldstein, I, and MacKinlay, A 2013, ‘Dark side of housing-price appreciation’, VOX, 25 November.

29 Beck, T, Buyukkarabacak, B, Rioja, F K and Valev, N T 2012,’Who gets the credit? And does it matter? Household vs. firm lending across countries’, The B.E. Journal of Macroeconomics, issue 1, vol 12, De Gruyter, United States.