RDP 2020-05: How Risky is Australian Household Debt? 4. Conclusion

The long-run rise in household indebtedness has increased concerns about the risks this poses globally to banking sectors and economies more broadly. Given the current level of household debt relative to income in Australia is high compared to many other countries, and its own history, these risks are often highlighted for Australia. Our exploration of these risks shows three implications.

First, we find that the high level and strong growth over recent decades of Australian household debt relates to a measurement issue and the strength of the Australian economy. In Australia the housing rental stock is mostly owned directly by the household sector and so debt financing is owed directly by households, while in many other countries rental housing is owned by corporations, the government or not-for-profits and so debt financing is owed by other sectors. Falls in nominal interest rates and the deregulation of the financial system (mostly in the 1980s) can account for much of the rise in household debt in Australia and globally. Further, stronger real income growth in Australia compared with many other countries accounts for at least part of the stronger growth, and so higher level, of household indebtedness in Australia.

Second, the distribution of debt across households has contrasting implications for risks to the banking sector and household consumption. We show that aggregate bank losses are higher when risks are concentrated among a small group of borrowers (for example, if households in a particular region simultaneously experience larger falls in employment and housing prices). In contrast, falls in consumption are larger if aggregate wealth effects are dispersed over a large number of highly indebted households. A key driver of this result is that highly indebted households tend to have fewer assets than those without debt (particularly wealthy retirees). An additional factor underpinning this result is that there is a limit to how much households can cut back on consumption in response to wealth shocks.

Third, the consequences of household indebtedness seem more likely to manifest through weaker economic growth than large bank losses. Bank stress tests highlight the resilience of the banking sector to severe scenarios, in large part because debt is not concentrated among households with high loan-to-valuation ratios. In fact, most of this debt is held by households whose risk of unemployment is relatively low, while households that are most at risk of experiencing very large changes in asset values (relative to income) tend to have little leverage. However, we simulate some fairly sizeable contractions in consumer spending when households are faced with extreme but plausible shocks. This is especially so if precautionary savings motives or loss aversion mean that households are more sensitive to falling housing prices than they have historically been to rising prices.

Overall, these findings confirm that Australian household indebtedness has important implications but they are sometimes misconstrued. The level and growth of household debt reflects high incomes and direct ownership of rental housing, not evidence of widespread excessive leverage. Household borrowing is a large share of banks' assets and so is important for their performance, but overall high lending standards, low loan-to-valuation ratios and banks' high level of capital mean they are highly resilient to adverse shocks to households. Rather, the implications for consumption of household indebtedness are an important mechanism to take into account when targeting macroeconomic policy to respond to economic shocks.