Stress-test of home loans approved in 2020-22 | I share results

Should you trust mortgage recommendations?

There seems to be so many experts telling Australians how to borrow – truth is, if they have not met you, how would they know?

There are plenty of recommendations online:

In my efforts to track down a widely dependable mortgage ratio to use when assessing how far a borrower is stretching—and whether they are mortgage stressed—I came up short.

Differences in incomes, lifestyles and asset positions are just a few of the variables that make a “one measure fits all” impossible to apply to a vast number of borrowers accurately.

I tested real loans, real borrowers – I share the results

Curious to see how borrowers I have dealt were fairing, I tested commonly used mortgage stress ratios against real loan applications I have seen been approved over the last few years – from a period 2020 to 2022.

I am sharing the results of my data crunching. It is original, real – it happened to 73 borrower groups.

The interesting – and somewhat staggering results, came from re-testing the same borrower groups as if they applied for loans at 2024 interest rates*.

In this article I will explore the following ‘real borrower’ numbers and ratios against some industry-accepted benchmarks.

Re-testing these borrowers is not a perfect way of seeing where they are at currently – but it does give a guide as to how different these periods are in home lending history.

The popular ratios I tested

30% Repayment ratio
It is widely accepted (and often recommended) that no more than 30% of your household income should go towards mortgage repayments.

Why?

In my article explaining different mortgage benchmarks, I question why 30% is the ‘magic number’ given there is often no reference where it came from. What is worse – many commentators do not reference whether it is a ratio of net (after tax) income or gross (before tax) income – the difference can be huge.

I will share with your how many of the loans I saw approved met or exceeded this 30% repayment ratio.

Then I re-test them as if they borrowed on 2024 rates.

Debt to Income (DTI)
The reporting of debt to income (DTI) multiples to regulators is a way of limiting over-exposure of borrowers – and therefore banks. It is part of APRA’s monitoring of the Australian lending landscape.

I explain the debt to income multiple in-depth in my article on mortgage benchmarks.

I will share with you how many of applications had a debt-to-income ratio over six—the multiple considered “risky”.

More information about my loan applications
To give these ratios context I will also share with you the following loan application information.

• Household Income
• Loan amounts
• Loan purpose

These are borrowers in Australia, using Australian lenders – as it happened.

No theory about it– these loans were approved.

The data – 73 borrower groups

Loans: Approved from 2020-22

Most of these loans were approved during the era of Covid-related low central bank interest rates.

Remember when the Reserve Bank of Australia “promised” to keep the central rate low for three years?

I have been around long enough to see central governments back flip on monetary policy so whenever I explained the low rate situation, I made sure to use the inverted comma two-finger gesture for the word “promise”.

So, borrowers were grappling with certainty (or uncertainty) of:

• Employment and business income
• Housing market
• Rental scarcity

Housing demand went up. So did home and rental values. Borrowers seemingly pushed their borrowing capacity in attempt to keep up with it.

So, I tested these borrowers in two parts:

2020-22 interest rates
This was a unique point in time – record low interest rates. This is the period these borrowers took out their home loans.

2024 interest rates
My ‘What if?’ test
I tested the same borrowers again as if they borrowed in 2024. A much higher, but arguably more realistic interest rate environment.

I used the RBA reference rates for 2024 interest rates to get an indication of how far borrowers might be stretched – 6.28% for owner-occupied and 6.50% for investment*.

Borrowers: By the numbers

I wrote loans for 73 borrower groups over a 34 month period between 2020 and 2022. To understand the type of borrower I dealt with over this period. Here is a snapshot of some findings.

Household Income

• The median household income was $244,840.
• Household Incomes ranged from $80,000 to $998,000.
• The highest income earning household ran a business and the lowest was a tech employee.
• 19% of loan applications used a self employed income as the majority income source – they ran a business.

Loan amounts

• Median loan amount applied for $741,412.
• Range between $300,000 and $3,081,300.
• The lowest loan was an additional loan for a client who wanted to make share investments.
• The highest lending amount was a refinance for a residential property investor who wanted an improved cashflow position.

Loan Purpose

• The main reason for borrowers seeking finance was almost equally split between purchase (48%) versus refinance (51%).
• The predominant purpose of the loans established was for owner-occupation (84%) as opposed to 16% for investment.

I don’t for a minute think these are representative of the average Australian borrower – but they are real, approved loans I arranged between 2020 and 2022.

It is human nature to compare your own financial position to some of these borrowers I will share with you – but leave it at that. Don’t read into that comparison, don’t act – see your mortgage broker or lender for advice specific to your situation.

I introduce borrowers to licensed mortgage brokers who are not afraid to get into the detail.

Repayment ratio results – Compared to 30% benchmark

At the time of loan advancement both net and gross repayment to income ratios had a median below the common 30% benchmark.

The numbers would be vastly different if these loans were made in 2024. There would be an almost 50% increase in the amount of income required to support the increase on both gross and net repayment ratios.

Gross repayment ratio

Proportion over 30%
The gross income to repayment ratio is commonly reported on and given it is before tax, it results in a lower ratio than after-tax ratio calculations.

The proportion of borrowers with a repayment ratio over the “30% benchmark” was only 8% at origination (2020-22 period) – based on gross incomes.

If these applications were based on 2024 rates, 48% of applications would need to set aside 30% or more of their gross income for mortgage repayments.

Range
The range of gross income vs repayment ratios using actual interest rates at the time of loan approval were 8% through to 40% (median 20%).

When I tested on higher 2024 interest rates the ratio range was 12% to 54% (median 29%).

Net income ratio

Proportion over 30%
Net income deals with the proportion of a households after tax income needed for mortgage repayments.

The proportion of borrowers with a repayment ratio over the “30% benchmark” for net income was 32% at origination. This increased to 81% when re-tested at 2024 interest rates.

Range
If I used net figures the proportions naturally increase:

The net income ratio using interest rates at the time of loan approval ranged from 10% through to 54% (median 27%).

Testing again the higher 2024 interest rate levels saw a significant leap. The range was now 17% to 73% (median 40%).

The repayment to net income ratio makes things very real – especially for those who maximised borrowing capacity at low rates.

What do repayment ratios look like in a real application?

Take one of my borrower groups who were approved for $760,000 home loan.

Household income was $177,000.

Gross ratio
These borrowers had a gross repayment ratio of 19% that increased to a much higher 32% when stress-tested against 2024 rates.

vs

Net ratio
When the same borrower was tested on the after-tax earnings they had at the time of application, their net income to repayment ratio of 25%. After tax household income was approximately $11,230 per month against repayments they took on of $2,842.

If the same borrowers took out this loan in 2024, when rates were around 6.28% would be staring at minimum monthly repayments of $4,694 and a net income to repayment ratio of 42%.

A 65% increase in repayment amount.

The equation for a borrower like this is simple and difficult. You need to find over $1,800 more each month to cover your mortgage repayments. What are your choices?

Earn more, save less, spend less – nothing easy about it.

Interested in seeing where you stand? See a licensed broker.

Debt to income results vs DTI of 6

DTI could be considered a quick sense check of borrowing exposure.

If a Debt to Income (DTI) multiple above 6 is considered risky, where did my borrowers land?

The median debt to income ratio was 4.10 with range between 1.60 and 6.87.

There were only a few applications with a DTI over 6. There can be good reasons for high DTI’s which I explain later.

It is worth highlighting this the DTI measure ignores interest rate.

What do DTI’s look like in a real application?

Take the same borrower group as above – they had a DTI of 4.38.

• The home loan applied for was $760,000
• Total debts $775,500 (after adding a couple of credit card limits)
• Household income of $177,000

The DTI of 4.38 is the result of dividing total debt $775,500 by $177,000.

Under 6 gets a lower risk tick from the bank but there were many other assessment hurdles to overcome before a loan approval was given

DTI vs Repayment ratio

When used in isolation, neither of these ratios, are necessarily going to cause an application to be declined.

Using the 30% repayment ratio and a DTI of 6 as risk benchmarks, I asked how many applications were approved that excessed both benchmarks?

The table above is based on net income (which will exaggerate the repayment ratio) shows only four borrower groups exceeded both.

How does income change debt exposure?

On pure numbers it does – the more you earn the more you can (usually) borrow.

But what does that look like as a repayment ratio or DTI multiple?

Looking for a relationship between household income and repayment ratios – there looks to be a slightly positive relationship.

The higher the household income, arguably the higher proportion of income is allocated to making mortgage repayments.

When it comes to borrowing a multiple of your income, the lower income households seem to borrower at higher multiples than the higher income households.

Are they stretching further?

Segmenting households by income

I divided up the groups in into households earning above and below $200,000. Let’s call them higher and lower income households.

There were 23 lower income households compared to 50 higher income households.

All four DTI’s over 6 were in the $200,000+ income bracket.

A greater proportion (34%) of applications with higher household income needed over 30% of after tax income for home loan repayments. This was compared to 26% of lower income households.

If we did these same calculations using 2024 interest rates 86% of higher income households would need to cover repayments with over 30% after tax income – compared to a still high 70% of lower income households.

Comparing these real loans to benchmarks

Repayment to income ratios

Based on 2020-22 interest rates, the median gross and net income ratios were under the 30% benchmark rate.

Zeroing in on net income – 32% of households needed more than 30% of household income to meet repayments. And this was at record loan interest rates.

A significant number of loan applications were approved with repayment ratios over the recommendations made by experts.

This emphasises the importance of understanding a borrowers whole financial position. Which is why a mortgage broker is a great place to start you home loan journey. I introduce borrowers to licensed brokers.

If these loans were assessed at 2024 rates, 81% of households would need to make allowance for over 30% of after tax income for home loan repayments.

This gives you an idea how the challenge to meet minimum repayments is very different to a few years ago.

Can that be interpreted as 81% of households are overstretched? I argue that is does not in my article challenging the 30% repayment benchmark.

Observation-wise, I have seen a couple of the more exposed clients (which had investment properties) sell to reduce debt and probably pocket some gains.

Debt to income multiple

DTI is a more static measure than a repayment ratio as it does not change when interest rates do. It gives lenders and regulators an indication of overall stretch for a borrower at loan application stage.

The DTI measure puts all debt limits in one bucket – credit cards, HECS/HELP education debt, home loans, etc.

It does not decipher between investment property debt for wealth creation and a personal loan to help with an overspending problem.

There were a four borrower groups with DTI’s great than 6. What they shared in common was the majority debt for investment home loan debt. This is in keeping with Australian banking regulator APRA observing a higher proportion of investors carrying a DTI greater than 6 compared to owner-occupiers.

From a mortgage broker perspective, a loan application with a DTI greater than 6 will often need strong justification as part of any loan assessment and approval.

When do I see a loan application with a high DTI make sense?

• DTI measure that includes a total loan limit but ignores offset balance, savings or investments.
• A high proportion of investment debt can push a DTI – and borrowing capacity – higher.
• Loan applications with high DTI’s usually have strong and clear strategies around reducing debt – like selling an investment property.

Final word

This set of real home loan analysis shows a large number of borrowers needing more than 30% of their income available for mortgage repayments. Some also had debt to income multiples over six – some had both.

You can assume these borrowers are likely more stretched now than they were when they took out their home loan.

Many borrowers can make adjustments to allow for higher repayments, painful as it may be—think lifestyle.

While this set of home loan data is representative of all borrowers, it does show that “commonly accepted ratios” do not determine if a loan application is approved or declined.

I recommend all borrowers – much like those in this analysis – see a mortgage broker for an assessment of their full financial position.

I introduce borrowers to licensed brokers, here:

Notes to study
It should be highlighted that these home loans were assessed under responsible lending rules and lenders used were all regulated. All home loans were approved.
The comparison to 2024 numbers is academic – I don’t make any assumption that any loans would be approved based on 2024 rates.
It may however, provide an insight into the level of stretch-stress households may be under in this new interest rate environment.

DTI measure
APRA CONSIDERS HIGH DTI’S AS >6
https://www.apra.gov.au/update-on-apra%E2%80%99s-macroprudential-policy-settings
DTI references:
https://www.rba.gov.au/publications/fsr/2021/oct/mortgage-macroprudential-policies.html

Analysis measures
Repayment ratio
In calculating repayments of home loans:

• For interest only loans I have used P&I over term remaining after expiration of Interest only period
• P&I loans used actual repayments.

*Interest rates used.
2020-22 Rates – Used actual rates at loan origination
2024 Rates – Rates 6.28% for owner-occupied and 6.50% for investment (reference RBA)

I am not a statistical expert. Just curious to see where some real borrowers sit against commonly used borrowing recommendations-ratios.

Used excel to generate graphs and lines of best fit to identify trend/relationship

Comparing to 2024 rates is academic and does not infer the same loans would be approved in 2024.