Financial Stability Review – April 2017
Characteristics of Highly Indebted Households
The aggregate household debt-to-income ratio has increased further in recent years, rising from already high levels. This has raised concerns about the household sector's resilience to unexpected declines in income or asset prices. Debt-to-income ratios vary substantially across households; roughly 30 per cent of households owe no debt, while some other households have debt-to-income ratios that are well above the average.
One important issue for financial stability is the number and characteristics of highly indebted households. These households are most likely to have difficulty repaying debt and sustaining consumption if their circumstances change. If the number of highly indebted households is large, their response to adverse income shocks could amplify and propagate an economic downturn, declines in housing prices and losses at financial institutions. This box outlines some of the characteristics of highly indebted households.
Household-level information on the distribution of debt, income and wealth is available in the Household, Income and Labour Dynamics in Australia (HILDA) Survey, which collects data on owner-occupier mortgage debt every year and provides a detailed breakdown of total household debt every four years (most recently in 2014). These data indicate that the share of households with very high debt-to-income ratios was little changed between 2006 and 2014 (Graph C1). In contrast, the proportion of households with more moderate debt-to-income ratios increased steadily, while the share of less indebted households declined. However, because the latest HILDA data on total debt are for 2014, it is not possible to discern whether these trends have continued more recently. The aggregate ratio of debt to income was overall little changed between 2006 and 2014, but has since risen (Graph 2.5).
There are several ways to classify household indebtedness. The approach used here is to classify ‘highly indebted’ households as those with debt-to-income ratios in the top 10 per cent of indebted households in a given survey year. Over the 10 years to 2014, most households in this group had a debt-to-income ratio above 550 per cent and as a group these households accounted for around 35–40 per cent of total household debt.
While the median debt-to-income ratio of the top 10 per cent of indebted households increased sharply from around 600 per cent to over 750 per cent between 2002 and 2006, it was little changed between 2006 and 2014. Highly indebted households steadily became more leveraged between 2002 and 2014, with the median debt-to-asset ratio of these households rising from around 50 per cent to just under 60 per cent.
The type of debt owed by highly indebted households was different to that for less indebted households. Less indebted households mostly had owner-occupier housing debt, while highly indebted households owed a rising share of ‘other’ property debt, most of which was investor housing debt (Graph C2). Between 2002 and 2014, the average share of highly indebted households' total debt comprised of ‘other’ property debt increased from around 20 per cent to nearly 30 per cent.
The HILDA Survey also includes information on the repayment behaviour of households. As might be expected, highly indebted households were less likely to be ahead of schedule on both owner-occupier debt and ‘other’ housing debt than less indebted households (Graph C3). Part of the reason is that they tended to have newer mortgage debt that had less time to amortise. Nonetheless, the share of highly indebted households ahead of schedule on their owner-occupier debt rose between 2006 and 2014, which suggests that, like other borrowers, these households took advantage of low interest rates to pay down debt at a faster rate. The share of highly indebted households ahead of schedule on ‘other’ property debt was lower, and increased only a little between 2006 and 2010. This is unsurprising given the greater tax incentives to retain debt on rental properties than on owner-occupier properties.
Other Household Characteristics
Around one-quarter of highly indebted households were in the bottom two quintiles for income and wealth, which is a higher share than for less indebted households (Graph C4). Lower-income households tend to devote a higher share of their income to essential living expenses (leaving less income for debt servicing), while households with lower wealth may be less able to sell assets to resolve any difficulties in servicing their debt. Nonetheless, a relatively small share of the total stock of debt owed by highly indebted households was owed by lower income or lower wealth households. In 2014, for instance, households in the bottom two income quintiles accounted for only 12 per cent of total debt owed by highly indebted households.
Using HILDA, it is possible to track highly indebted households over time. Households that were highly indebted in a particular year tended to reduce their debt-to-income ratios substantially in subsequent years. For instance, in 2002 households that were highly indebted had a median debt-to-income ratio of 600 per cent, yet by 2006 the median debt-to-income ratio of these particular households had fallen to below 350 per cent and their debt-to-income ratio had fallen further by 2010 (Graph C5, left panel).
The HILDA Survey also includes questions on financial stress experienced by households over the previous year. There was a broad-based decline in the share of households experiencing episodes of financial stress between 2001 and 2015 (the time span available in the HILDA Survey). Nonetheless, households that were highly indebted in a particular year had a greater propensity to experience financial stress. For instance, households that were highly indebted in 2002 were more likely to experience at least one incidence of financial stress in all other years compared with households that were less indebted in 2002 (Graph C5, right panel). The result also holds true for other cohorts. This suggests that a greater share of highly indebted households face financial difficulties and are more likely to be vulnerable to events that affect their ability to repay their debt, such as income declines or increases in interest rates.
Overall, these data highlight that highly indebted households can be more vulnerable to negative economic shocks and pose risks to financial stability. In particular, highly indebted households are less likely to be ahead of schedule on their mortgage repayments and they are more likely to experience financial stress, hence could be more vulnerable to adverse macroeconomic shocks. The consequent effects of this stress on the broader economy may be exacerbated by the disproportionately large share of investor housing debt owed by highly indebted households. As discussed in the ‘Household and Business Finances’ chapter, heightened investor demand can contribute to the amplification of the cycles in borrowing and housing prices, particularly when this investment is highly leveraged. Nonetheless, HILDA data also show that much of the debt held by highly indebted households is owed by households with high income and wealth, who are typically better placed to service larger amounts of debt.
This classification is not intended to suggest that a debt-to-income ratio below or above a certain level is sustainable for a given household. A number of factors influence the ability of a household to service their debt at a given debt-to-income ratio. For instance, higher-income households may be able to afford to devote a larger proportion of their income to debt repayments after meeting basic needs, while lower-income households could find it difficult to service debt at relatively low debt-to-income ratios. 
‘Other’ property debt also includes debt owed on owner-occupier holiday and second homes. 
Instances of financial stress that are reported in HILDA include: being unable to pay a bill, mortgage repayment or rent on time; being unable to heat the home; and asking for financial help from friends, family, or community organisations.