From the excellent George Tharenou at UBS Evidence Lab:
2021 UBS EL Mortgage survey implies material deterioration in loan standards
UBS Evidence Lab surveyed ~900 Australians who took a mortgage in the last year – conducted from 21 June to 5 July, at the start of lockdowns in NSW and Victoria. The 2021 vintage is the 7th year. Overall, it suggests a material deterioration of lending standards. The share of home loans misstated (i.e. not completely factually accurate) increased sharply to another record high 41%, up from 38% last year, & far above 27% in 2015. Interestingly, there was a fall in the number of areas where respondents misstated. But, in contrast, the size of misstatement lifted materially – with a much larger over-statement of income (especially by 25%-34%) and assets; with a greater under-statement of financial commitments and living costs (especially by 25%-34%).
There are 1413 words left in this subscriber-only article.
Start your free 14-day trial today!
Large rise in ease of applications; but mixed results on verification
Overall, there was a large rise in application ease. That said, documentation & verification had a large increase in “much less” and “much more”, suggesting a more targeted risk-assessment. Mortgage rejections ~held. But time for approval lengthened again, albeit likely due to the record volume of loans rather than tighter standards.
Income is up & expected to rise; and savings are high, but few plan to consume
Positively, 43% report household income is above the pre-COVID level, & 55% expect income to improve over the next year. However, while 63% of respondents saved extra since COVID, only 10% are planning to consume all of this; and savings intentions vs 3 months ago are high, with 41% planning to save more and only 9% to save less.
Q2 APRA data showed falls in IO & high-LVR loans, but key is spike in high DTI
Separate APRA data showed higher risk home loans largely improved q/q in Q2, with a decline in the share of Interest-Only (to 17.4%, lowest since 2018), high-LVR ≥80% (to 38.8% after 41.2%), low-doc & outside serviceability (2.9%), & investors (27.6%). This was surprising given UBS Evidence Lab results. We think a ‘truer’ read of rising leverage and risk is the record rise in the share of very-high ≥6x DTI loans (+2.7%pts to 21.5%). Indeed, ABS data shows average owner-occupier loan size jumped by 23%y/y in July; while UBS Evidence Lab data shows guarantor requirement increased to a record 36% share. Amid home prices booming 18.3% y/y (highest since 1989), we think borrowers are ‘chasing the market’ & stretching towards their capacity limit to be able to qualify.
Macro-prudential tightening is a risk that’s not priced into equity markets
In a ‘normal’ environment this backdrop would see a repeat policy response of macroprudential tightening & RBA rate hikes. Indeed, booming house prices recently led to a jump in credit growth, & data suggests material deterioration in lending standards. But, during this cycle the RBA has remained remarkably dovish, saying monetary policy is not a tool to address house prices, and only sustained 10% credit growth would be problematic & at some point then consider intervention. Therefore, as the economy rebounds – but there are still no RBA rate hikes, given contained CPI – it’s likely home prices and credit growth remain strong. Hence, amid ongoing lockdown, and with uncertainty about the pace of recovery, it still seems unlikely there will be policy tightening before May-22. We expect dwelling prices to grow 20% this year, and by at least 5% next year albeit closer to 10% is likely (UBS previously expected ~5%).
Housing credit growth should accelerate to >7% y/y (& total credit growth near 6%). Consequently, there is likely a case for macro-prudential tightening in 2H-22. During the 2017 policy tightening cycle, short interest for banking & housing stocks rose substantially. However, short-interest & volatility is currently at low levels and suggests the risk of macro-pru is not priced into equity markets.