Financial Stability Review – March 2012 Box B: Home Mortgage Debt: Recent Insights from the HILDA Survey
The Household, Income and Labour Dynamics in Australia (HILDA) Survey interviews the same households each year, mainly between August and November, with the latest published results being for 2010. It therefore makes it possible to trace individual changes in household debt over the past decade. Detailed data on owner-occupier mortgage debt are included in each year's survey; a full breakdown of all forms of household debt is available at four-yearly intervals (2002, 2006 and 2010).
Broadly consistent with the aggregate data (see Graph 3.8 in the ‘Household and Business Balance Sheets’ chapter), the HILDA Survey data on total household debt show that the debt-to-income ratio for the median indebted household increased sharply between 2002 and 2006, and then rose only marginally between 2006 and 2010 (Graph B1). A similar pattern is evident when the sample is restricted to indebted owner-occupiers only. The recent slowdown in growth is observable across the distribution of the debt-to-income ratio.
The moderation in the debt-to-income ratio reflects a shift of many households towards paying down debt rather than accumulating it: the share of indebted owner-occupier households who made substantial principal repayments on their mortgage (of $25,000 or more over the year) was significantly higher in 2010 than the average between 2002 and 2007 (Graph B2, Table B1). A repayment of this size is significant, and well in excess of the additional repayment that would accrue if repayments were held constant in the face of lower interest rates. A $25,000 reduction in debt is equivalent to more than double the annual principal repayment a borrower would make in any of the first fifteen years on the average new mortgage (of around $300,000 at origination). Moreover, the share of households who substantially increased their mortgage (by $25,000 or more) was lower in 2010 than the average between 2002 and 2007.
The change in mortgage repayment behaviour has been broad based across the income distribution, but particularly marked among households in the upper income brackets. There was a significant increase in the share of households in the top 20 per cent of the income distribution who made substantial mortgage repayments in recent years, while the share of these households that substantially increased their outstanding mortgage debt fell. Higher-income households typically spend a smaller share of their incomes on essential goods and services, so they have more scope to pay debt down quickly than do lower-income households. Despite this, the share of lower-income households making large mortgage repayments in 2010 was also significantly above the average between 2002 and 2007.
By age bracket, the share of older households with mortgages (where the household head is aged over 55 years) making substantial repayments increased significantly in recent years and there has been a substantial decline in older households adding large amounts of debt. Despite the fact that younger households (with a head aged 15–34 years) typically have lower-than-average incomes, the share of these households making large repayments also increased significantly. Overall, the change in mortgage repayment behaviour was largely driven by households with a head aged 35–54 years; a lower share of younger and older households have mortgage debt.
The recent repayment behaviour by households has occurred broadly across states. New South Wales borrowers in particular appear to have become more cautious; the share of households making large mortgage repayments was higher in 2010 than in prior years and the share of households taking out substantial amounts of mortgage debt decreased. The share of Queensland and Western Australian households making substantial mortgage repayments also increased significantly.
The recent household balance sheet consolidation is part of a more prudent approach to financing that households have been taking in recent years. Although the household sector remains quite indebted, households appear well placed to manage their debt. Around 70 per cent of owner-occupier housing debt is held by high-income households (in the top 40 per cent of the income distribution) that typically have lower debt-to-income and debt-servicing ratios (DSRs) (Graph B3). Only a relatively small share of low-income households (in the bottom 20 per cent of the income distribution) have housing debt. However, these households tend to be very indebted and have high DSRs. Sometimes this is because they have temporarily low incomes.
Few households appear to be vulnerable to falling into mortgage arrears. While 5 per cent of households who owned residential property and had owner-occupier debt missed a mortgage repayment in the year prior to the 2010 Survey, only a small share of these households were behind schedule on their mortgage. This suggests that most households who miss a repayment are able to resolve their difficulties. The share of owner-occupier households that could be considered to be most vulnerable, that is, with both high DSRs and high loan-to-valuation ratios (LVRs), has decreased in recent HILDA Surveys and is quite low (Graph B4). Part of the recent decrease in the share of households with high DSRs reflects lower interest rates – all other things equal, DSRs fall when interest rates fall – though these measures of vulnerability have been broadly steady at a low level for some time. The measure of debt servicing recorded in the HILDA Survey covers actual repayments made by households and includes excess repayments; some households may have a high DSR voluntarily. While required repayments are not reported in the HILDA Survey, reasonable approximations of the interest component and loan term are available. Based on this information, it seems that roughly one-half of borrowers with DSRs above 50 per cent could be making payments substantially above their likely required repayments, and would be reasonably considered to be less vulnerable to falling into distress.