Possible house price drop poses some risk for Canadian, Swedish, Australian banks House prices and household debt have risen to unprecedented levels in Canada, Sweden and Australia. This increases the risk that adverse economic developments could trigger a house price drop, leading to higher loan losses for banks. We have recently taken rating actions on banks in these systems to reflect this risk1,2,3 , although they remain among our most highly rated banks globally4 . While our base case is that house price growth will slow in Canada and Australia, the market may already be turning in Sweden. We believe that a more substantial correction would lead only to limited losses on mortgages, as the banks benefit from a range of safeguards. However, the banks would be exposed to second order effects, as falling house prices would likely weigh on consumer sentiment, amplifying the economic slowdown and pushing up losses on corporate and consumer loans.
House price rise increases sensitivity to downturn. Between 2000 and 2016, house prices rose by 144% in Sweden, by 115% in Canada, and by 113% in Australia. Household indebtedness has risen sharply over the same period. Mortgages account for 63% of the banking system’s total loans in Australia, 48% in Sweden and 39% in Canada.
Mortgage losses likely to be limited. In all three systems, protective features built into the mortgage market and banks’ underwriting practices, along with full recourse loans, would limit mortgage losses in the event of a house price correction (similar to the 2008 house price decline in Denmark).
Banks exposed to second order effects. The material economic slowdown that would likely accompany any substantial house price correction would lead to higher losses on consumer loans, commercial real estate loans, and loans to consumer-exposed corporates. Consumer loans are a particular vulnerability for banks in Canada, where 30% of household loans are unsecured.
Profits would likely be sufficient to absorb loan losses. We expect banks in these systems to absorb loan losses through earnings in most scenarios. Any impact on their capital levels, which we currently assess as strong to adequate, would likely be limited. In Canada, loan losses could increase to 2.2% of gross loans from 0.4% in 2016, to 1.8% from 0.2% in Australia, and to 1.4% from 0.1% in Sweden, before banks’ capital were affected.
Wholesale funding potentially vulnerable. Reliance on confidence-sensitive wholesale funding is greatest for banks in Sweden and Australia. We expect covered bond funding, which ranks higher in insolvency proceedings and is collateralised, to be more resilient than unsecured wholesale funding.
“Mortgages losses likely to be limited”. Sure, thanks to the government nationalising the lender’s mortgage insurers. The public’s losses are not likely to be limited.