Why mortgage stress has risen

Digital Finance Analytics (DFA) has released its mortgage stress figures for May, which shows that around 41% of households remained ‘mortgage stressed’, up significantly from the pre-pandemic level of 32.9%:

Australian mortgage stress

Mortgage stress has risen massively during the pandemic.

Martin North explains the rise in mortgage stress as follows:

Our approach to measuring stress is unique in that we examine household cash flow – money in and money out. Given that many households saved hard last year though the heights of COVID, it is not surprising to see many now draining down those savings, by spending more. This means that their cash flow will in net terms be negative for now, and so will register as stressed. That said, if spending continues unabated financial difficulties will eventuate.

In addition we continue to see more households reaching for credit (from Buy Now Pay Later, to Pay Day loans) as well as equity release from property. In fact the latest hikes in perceived values has led to a run of refinancing, to try and pay down debt, or to provide funds to offspring for property purchase via the Bank of Mum and Dad. Again these one-off moves can adversely impact household stress measures in our methodology.

And we also note that many prefer not to accept the truth that some households are not home and clear in terms of their finances, given the uncertain part-time work, multiple jobs and zero hours contracts which many are on. But we continue to analyze households in net cash flow terms. If more funds drain away, compared with income, they are classified as stressed.

As I have noted previously, the increase in mortgage stress during the pandemic does not gel with the macroeconomic data and seems implausible.

First, mortgage rates have tanked, as illustrated clearly in the next chart from CoreLogic:

Average mortgage rates

Average mortgage rates down sharply since start of pandemic.

Second, household disposable income has risen sharply on the back of massive stimulus, experiencing its strongest annual rise since 2008:

Real household disposable income per capita

Real per capita household disposable income has broken a decade of stagnation.

Third, The ratio of debt repayments – both principal and interest – to household disposable income has fallen to a 17 year low, according to the Bank for International Settlements:

Australian household debt repayments

Australian household debt repayments have fallen to a 17 year low compared to disposable income.

Finally, Australian households have built up a massive war chest of savings, which augers against the ‘mortgage stressed’ theme:

Australian household savings

Australian households saved a record $200 billion in the year to March 2021.

According to the above explanation by Martin North, DFA’s mortgage stress data measures “household cash flow – money in and money out”. So, “given that many households saved hard last year though the heights of COVID”, shouldn’t stress levels have fallen, since more money was coming in than out? Instead, mortgage stress rose significantly by around 8%.

What am I missing here?

Unconventional Economist
Latest posts by Unconventional Economist (see all)


  1. Averages lie?
    Household income includes those without mortgages such as young people living with mum and dad, that received a pay rise via stimulus payments.
    People who borrowed at much higher rates being gifted another interest rate cut that they didn’t need.
    I know a few people who are well off and still withdrew $10k-$20k from super.
    No using home as ATM to fund expensive overseas trip, etc.

  2. PalimpsestMEMBER

    It’s an interesting set of charts. The DFA top curve shows reduced per household debt, while the lower curve shows an increase in stress. There hasn’t been enough property turnover to make the purchase or upgrades of property the key factor (lower rates but larger mortgages), although it will be a factor.
    The income flow as a factor has been a good indicator in the past, but no indicator handles all market conditions reliably. That suggests it might be misleading and take a couple of months to stabilise.
    One other possibility is that it’s correct. That’s a feasible scenario if mortgage stress and household savings are spread unevenly (as they are). This scenario would require a dissection of what “annual household net savings” really consist of. If it’s just the aggregate personal savings account surplus then some business people did remarkably well while others struggled. If it includes property and share valuation then the missing factor becomes much easier to identify (and more uneven still).
    The one that is hard to reconcile with increased stress is debt repayment to household income. The restoration of work, and drop in underutilisation might flow through to these figures soon. For the sake of these poor people I hope so.