Fixed-rate Housing Loans: Monetary Policy Transmission and Financial Stability Risks


Photo: Andrew Merry – Getty Images

Abstract

Fixed-rate borrowing increased significantly during the COVID-19 pandemic, which has delayed the effect of the higher cash rate on borrowers’ cash flows. A key issue for the economic outlook, and by implication financial stability, relates to the ability of borrowers with fixed-rate loans to adjust to substantially higher borrowing costs when their fixed-rate mortgages expire. Borrowers with fixed-rate loans have had a considerable period to adjust their finances to prepare for the increase in their mortgage payments and many appear to have similar savings to borrowers on variable rates. However, on some metrics fixed-rate loans have higher risk characteristics than variable-rate loans. With many fixed-rate mortgages expiring in the period ahead, the Reserve Bank will continue to closely monitor the implications for household consumption and financial stability.

Introduction

During the COVID-19 pandemic, the value of fixed-rate housing loans increased substantially, peaking at almost 40 per cent of outstanding housing credit in early 2022, or roughly twice their usual share from prior to 2020 (Graph 1). Many borrowers also fixed their interest rates for longer periods than is typically the case. Lenders lowered their advertised fixed rates below variable rates to compete for borrowers (Graph 2). Strong competition on fixed-rate loans was made possible, at least in part, by lenders’ ability to obtain low-cost term funding via the Reserve Bank’s monetary policy response to the pandemic, which included the introduction of the Term Funding Facility, the three-year yield target and forward guidance (RBA 2022a). Competing vigorously on fixed-rate loans also enabled lenders to attract new borrowers without reducing their reference rates for variable-rate loans (which would have lowered rates for their existing variable-rate customers). New fixed-rate lending slowed sharply from late 2021 as new fixed rates rose relative to variable rates, along with market yields around the period when the yield target ended. By mid-2022, new fixed-rate lending had declined to around 5 per cent of total new lending.

Graph 1

Graph 1: A two panel graph from 2019 to 2022 showing the share of new lending originated at a fixed rate (stacked by its original maturity) on the left panel and the share of outstanding credit with a fixed rate (stacked by its residual maturity) on the right panel. The left panel shows that new fixed-rate lending increased sharply from 2020 to late 2021, especially at terms greater than two years. The right panel shows that the outstanding share of credit with a fixed rate increased from 20 per cent in late 2019 to almost 40 per cent in late 2021. It has since declined to around 30 per cent by the end of 2022, with around half of the outstanding loans due to expire within a year.

Graph 2

Graph 2: A single panel graph from 2005 to 2022 showing interest rates on new housing loans. One series shows the average new variable rate and the other shows the advertised fixed rate among the major lenders for a term of three years. Both series have tracked each other closely over history, with fixed rates typically a little above variable rates. However, from 2020 to 2021 fixed rates fell below variable rates for the first time (consistent with the large increase in fixed-rate lending). Since late 2021 fixed rates have increased once again above variable rates.

Most borrowers in Australia who fix their mortgage interest rate do so for three years or less. This means that the fixed-rate term on most loans taken out during the pandemic has expired recently or will do so over the coming two years. One-quarter of fixed-rate loans outstanding in early 2022 have now expired; most have rolled on to a variable interest rate, rather than re-fixing at a higher rate. Another 40 per cent of fixed-rate loans outstanding in early 2022 will expire by the end of 2023 and a further 20 per cent by the end of 2024. This equates to 590,000 loan facilities in 2022, 880,000 in 2023 and 450,000 in 2024. The profile of expiring fixed-rate loans is similar across the states and territories and between capital cities and regional areas.

The analysis in this article draws largely on the Bank’s Securitisation dataset, which covers around one-third of outstanding housing credit (Fernandes and Jones 2018), liaison with major banks and survey data on household balance sheets.

Borrowers with expiring fixed-rate loans face large increases in their repayments

Scheduled loan payments will increase for borrowers when their fixed-rate terms expire, based on current interest rates and assuming the cash rate changes in line with the path inferred from financial market pricing as at 1 March 2023 (the ‘market path’). To date, borrowers have rolled off onto a rate similar to that faced by existing variable-rate borrowers (Graph 3). Many borrowers are likely to have subsequently reduced the initial variable rate they rolled off to by negotiating with their existing lender or refinancing with another lender, especially those of higher credit quality (RBA 2023b; Carse, Faferko and Fitzpatrick 2023).

Graph 3

Graph 3: A single panel graph over 2022 comparing average interest rates on existing variable-rate loans, new variable-rate loans and the variable rate loans roll off to after their fixed rate expires. Shows that the variable rate loans roll off to when their fixed rate expires is similar to the variable rate on existing loans.

How much the scheduled payment on an expiring fixed-rate loan increases depends on the loan’s current fixed rate, the timing of the expiry of that rate and the loan’s new interest rate (Graph 4). Scheduled loan repayments on some of the fixed-rate loans that expired in 2022 increased by up to 50 per cent at expiry, although around two-thirds of loans experienced an increase of 30 per cent or less (Graph 4, top panel, area under orange line). Most of these borrowers’ repayments have increased further since they switched to variable-rate loans, as they have for all borrowers with a variable-rate loan (Graph 4, top panel, blue line). The one-off increase in scheduled payments when the fixed rates on these loans expire is large because fixed rates were very low when most were taken out.

Loans that are yet to roll off their fixed rate will face a larger initial increase in scheduled repayments than those that rolled over during 2022 because the cash rate increased over that year and the market path implies further increases until late 2023. Around 90 per cent of these loans will see their scheduled payments increase by 30 per cent or more (Graph 4, bottom panel, area under orange line) and most will experience the total increase upon the expiry of the fixed rate (bottom panel, orange and blue lines similar). Though these increases are large for many of the loans yet to roll off their fixed rate, they are similar in size to the total increases in scheduled payments for variable-rate loans since the first increase in the cash rate in May 2022.

Graph 4

Graph 4: A two panel graph showing a distribution of increases in the scheduled payments for expiring fixed rate loans (assuming they roll off to variable rates). Each panel has two distributions: one for the size of payment increases when the fixed rate expires and another for the size of payment increases by the end of 2024 (assuming mortgage rates evolve in line with the market path for the cash rate). The top panel shows loans that saw their fixed rate expire in 2022: at expiry most saw their payments increase by 30 per cent or less but by the end of 2024 their payments will have increased by 40-60 per cent. The bottom panel shows loans that have their fixed rate expiring in 2023 and 2024: most will experience a 40-60 per cent increase in their loan payment and most of this occurs when their fixed rate expires.

Borrowers with fixed-rate loans have benefited from a prolonged period of low interest rates, and will have had more time than borrowers with variable-rate loans to prepare for higher rates – including by accumulating savings. By having a fixed rate, around 60 per cent of fully fixed-rate loans (outstanding in December 2022) will have avoided higher loan payments equivalent to more than three months of their new required repayment after their fixed rate expires (Graph 5). Constructing this estimate compared the scheduled payment at the loan’s fixed rate with the scheduled payment the loan would have had each month if it had a variable rate. Loans that have a fixed rate for longer will benefit most.

It is not possible to observe how much of the cash flow associated with experiencing lower loan payments that borrowers on fixed-rate mortgages have actually saved (or will save). As discussed later, borrowers have broadly similar levels of liquid savings – regardless of the type of interest rate on their loan – and have increased these over the past few years. While many borrowers on fixed rates may have saved or be saving in preparation for higher loan payments, some may have used the period of low fixed borrowing costs to consume more than they would have otherwise.

Graph 5

Graph 5: A single panel graph showing the distribution of payments avoided by borrowers on fixed rates since May 2022. The graph compares the loan payments the borrowers have actually made (or will make) relative with a hypothetical scenario where they are variable rate borrowers since May 2022. Around 30 per cent of borrowers have avoided between one and three months of the loan payments they are expected to face when their fixed rate expires. A further 30 per cent have avoided between three and six months, another 30 per cent have avoided more than six months. The remainder have avoided less than one month of scheduled payments at expiry.

Fixed-rate loans delay the transmission of a higher cash rate to mortgage payments

In 2022, scheduled loan payments in aggregate increased a little slower than in the past because of the higher share of fixed-rate credit and because borrowers fixed their rates for longer than is typically the case. A higher cash rate is still transmitting quickly to most loans, though, because the majority have a variable rate, many have already rolled off their earlier fixed rates and many more will do so in the coming months.

The Reserve Bank raised the cash rate by 3 percentage points in 2022 to 3.1 per cent; over the same period, the average outstanding mortgage rate (for all loans) increased by almost 2 percentage points to 4.7 per cent. If all fixed-rate loans instead paid the variable rate on new loans, the average outstanding mortgage rate would be 70 basis points (bps) higher than it was in December 2022. This gap will slowly shrink as more fixed-rate loans expire and will be around 25 bps at the end of 2023 and close to zero by the end of 2024.

The Reserve Bank monitors the effect of interest rates on scheduled housing loan payments because they directly affect household disposable income; this cash flow channel is an important mechanism for transmitting monetary policy (La Cava, Hughson and Kaplan 2016). Borrowers can service their loans when their required loan payments increase by saving less, drawing on existing savings and wealth, or reducing consumption. If available to them, some borrowers may also choose to increase their income – for example, by changing jobs or working more hours. The higher share of fixed-rate loans (compared with past interest rate cycles) delays the cash flow channel in aggregate because some households will not face higher interest rates for a period. But this will also depend on how fixed-rate borrowers prepare their finances ahead of their fixed rate expiring. For example, if they save a lot more in anticipation of the increase in required loan payments in the future, they may not need to reduce their consumption (further) when their loan payments actually do increase.

Fixed-rate loans have riskier characteristics than variable-rate loans …

The large and discrete increase that borrowers with fixed-rate loans have faced or will soon face in their mortgage payments is one of the factors expected to contribute to slower household consumption in the period ahead. It could ultimately increase the potential for financial stability risks if many borrowers default on their loans (leading to losses for lenders) (Bergmann 2020) or reduce their consumption to the extent that unemployment increases significantly and other borrowers facing unemployment in turn struggle to service their debts.

In general, financial stability risks are more likely to eventuate if there are large numbers of borrowers with risky characteristics, including high levels of debt relative to income and assets, low income levels and low spare income after meeting loan payments and other essential expenses (RBA 2022b). On some of these metrics, borrowers with fixed-rate loans are more risky than those with variable-rate loans (who have seen large increases in their loan payments already and who to date have shown little evidence of increased financial stress). Nevertheless, while fixed-rate loans tend to be newer and to a degree are expected to have more debt outstanding relative to income (since borrowers have not yet had time to pay down their loan) the differences are not large.

Some borrowers could fall behind on their scheduled loan payments if they cannot adjust to higher loan payments. One rough metric for assessing whether a borrower might encounter difficulty servicing their loan is if they spend more than 30 per cent of their income on scheduled loan payments. In practice, however, other factors such as the borrower’s income level and savings are also important. In April 2022, most borrowers with fixed-rate loans spent between 5 and 25 per cent of their income making mortgage payments before they rolled off; only around 10 per cent of borrowers on fixed rates spent more than 30 per cent (Graph 6). But after rolling off, roughly 25 per cent would need to spend more than 30 per cent of their income on loan payments – a slightly larger share than borrowers with variable-rate loans. Much of the increase comes from borrowers in the bottom half of the income distribution, who are more likely to have less spare cash flow and so may reduce their consumption and/or encounter difficulty servicing their debt as they roll off. This is a key group to monitor as fixed-rate loans roll off over the period ahead, especially if this group also has low savings buffers.

Graph 6

Graph 6: A single panel graph showing the distribution of loan servicing ratios or LSRs (that is, the share of income a borrower spends make their loan payments) as at April 2022. One line shows the distribution of LSRs for fixed-rate loans before expiry, another shows LSRs for fixed-rate loans based on the scheduled payments they are expected to face at end-2024 (after their fixed rate expires) and another line shows LSRs for variable-rate loans projected at end-2024. The distribution of LSRs for fixed-rate loans prior to expiry is concentrated between 5 and 25 per cent. By the end of 2024 the distribution shifts rightward and is similar for (expired) fixed-rate and variable-rate loans.

Fixed-rate borrowers are more likely to have larger loans relative to their incomes (LTI ratio > 6) or high loan-to-valuation ratios (LVR > 80) than borrowers on variable rates (Graph 7). This is especially so for some fixed-rate loans with low mortgage prepayments. These loans are more risky on average, but in part this reflects their tendency to be newer and so borrowers have had less time to accumulate equity or liquidity buffers. Borrowers with fixed-rate loans are also more likely to be first home buyers, although first home buyers on fixed rates tend to have more mortgage prepayments than other loans with otherwise similar characteristics.

Graph 7

Graph 7: A two panel graph comparing fully fixed and variable rate loans by age (on the left panel), LTI, LVR and whether the loan is to a first home buyer (on the right panel). Full fixed rate loans are slightly younger at the median and have a higher share of loans with LTIs above six times, LVRs above 80 per cent and that are first home buyers.

… but many have built savings buffers to help mitigate risks

One way borrowers can adjust to higher loan payments is by using their savings. Comprehensive data are available on offset and redraw balances of owner-occupier variable-rate loans, which form a large part of these borrowers’ savings (La Cava and Wang 2021). However, assessing the savings buffers of borrowers with fixed-rate loans is more difficult because many hold more of their savings outside of their mortgage.

Most fixed-rate loan products do not have an offset facility and typically restrict mortgage prepayments. Nevertheless, some avenues exist for fixed-rate borrowers to save via their mortgages especially if they have a split loan. While there are no comprehensive data on the savings held by fully fixed-rate borrowers in non-mortgage forms, private survey data suggest that fixed-rate borrowers have similar levels of total liquid assets to borrowers with variable-rate or split loans. This is particularly true among borrowers with lower incomes, who might otherwise be regarded as potentially more vulnerable. In addition, liaison with some banks has indicated that borrowers on fixed rates in general have as many (and in some cases more) savings than other borrowers, in part because they have so far avoided higher loan payments.

Some borrowers on fixed-rates have large mortgage prepayments, especially those with split loans

Some fixed-rate borrowers have a ‘split’ loan with a fixed- and variable-rate component. A borrower may elect to split their loan into a fixed-rate facility and a variable-rate facility, with an offset and/or a redraw feature available to make prepayments on the variable-rate portion of their loan. Information from major banks suggests that over half of their owner-occupier customers on fixed rates have a split loan and that the majority of their loan balance has a fixed rate.

Borrowers with a split loan tend to exhibit similar savings behaviour in their mortgages to those with variable-rate loans, and most have substantial savings buffers that can help them meet higher repayments. Around three-quarters of owner-occupiers with split loans could cover their minimum payment for more than three months if they were to immediately roll off to a variable rate (and 60 per cent could cover their payments for more than a year) (Graph 8).

For loans that have a fully fixed rate, most lenders allow the borrower to make limited prepayments, which they can redraw but only when their fixed-rate term expires. Among the largest 10 lenders, the median prepayment allowance is $10,000 per year of the fixed term. Additionally, borrowers who switched from an existing variable-rate loan to a fixed rate may have made additional payments prior to fixing their interest rate – for example, analysis of a sample of loans rolling onto fixed rates suggests that around half of borrowers had made mortgage prepayments of more than three months just prior to fixing their interest rate. Further, around 15 per cent of fully fixed-rate loans in the Securitisation database have an offset facility with a positive balance from one of the relatively few lenders offering this feature.

Consistent with the limited options available to them to save via their mortgage, only one-third of fully fixed owner-occupiers have excess payments to cover their minimum scheduled payment for three or more months if they were to immediately roll off to the average new variable rate (compared to two-thirds of variable-rate and split loan borrowers) (Graph 8).

It is important to note that this does not necessarily suggest that fully fixed-rate borrowers are at higher risk of facing repayment difficulties or reducing their spending. Given they are restricted from saving via mortgage prepayments, many fully fixed-rate borrowers are likely to hold their savings buffers in other (non-mortgage) forms.

Graph 8

Graph 8: A single panel graph showing the distribution of months of mortgage prepayments for loans by their type of interest (fully variable, fully fixed or split). Fully variable and split loans have a similar distribution of prepayments: around 60 per cent are more than 12 months ahead and roughly 20 per cent each are between 3 and 12 months ahead or less than 3 months ahead. For full fixed-rate loans the distribution is different: over 60 per cent are less than 3 months ahead and roughly 20 per cent each are 3 to 12 or more than 12 months ahead.

Survey data suggest borrowers have similar savings buffers, regardless of interest rate type

Private survey data suggest that many borrowers with fully fixed-rate loans hold substantial non-mortgage savings (Graph 9). These data are based on a smaller sample than the Securitisation data, but provide a more complete comparison between the savings of borrowers on fixed rates and variable rates (non-mortgage savings are not visible in the Securitisation data). The additional savings are held in (non-offset) bank deposits and other liquid assets (such as managed funds and shares) but exclude funds available for redraw from a loan account. These additional savings show borrowers on fixed rates have meaningful savings buffers that are comparable to variable-rate and split loan borrowers. Borrowers on lower incomes have similar savings regardless of the type of interest rate on their loan.

There are still some borrowers (regardless of the type of interest rate) with low savings buffers that are vulnerable to higher loan payments, especially those with lower incomes. But, based on these data, borrowers on fixed rates appear at least as prepared as other borrowers for the coming increase in their loan payments.

Graph 9

Graph 9: A single panel graph showing the distribution of liquid assets of mortgage holders (expressed as months of scheduled mortgage payments) for loans by their type of interest (fully variable, fully fixed or split). Liquid assets includes balances in deposit accounts (including offset accounts), shares and managed funds. Borrowers have a similar distribution of liquid assets regardless of the type of interest rate on their loan: 60 per cent of borrowers have more than 12 months of liquid assets and roughly 20 per cent each have 3-12 and less than 3 months worth of liquid assets.

One-fifth of fixed-rate loans increased mortgage prepayment by six months or more after roll-off

Once a loan’s fixed-rate term has expired, if it has rolled off to a variable rate most lenders will allow the borrower to make unlimited payments into an offset or redraw facility linked to the loan (they may also be able to redraw any extra payments made before or during the fixed-rate term). At this point, the borrower has an incentive to convert at least some liquid savings they hold elsewhere (e.g. in a deposit account) into mortgage prepayments since they will most likely receive a higher return (and a tax benefit) from doing so. Observing loans after their fixed-rate term expires can therefore provide insights into the non-mortgage savings buffers these borrowers held while they were on a fixed rate.

A material share of borrowers make large transfers into their mortgage after their fixed rate expires. Around one-fifth of fully fixed-rate loans rolling off between February and October 2022 increased mortgage prepayments by more than six months of their new required payment within a few months of rolling off to a variable rate. The share of fixed-rate loans that made large prepayments was twice as high compared with variable-rate loans over the same period (Graph 10). A little under half of these loans did not have meaningful mortgage prepayments prior to rolling off, which suggests that these borrowers held substantial non-mortgage savings to transfer into their loan after their fixed rate expired.

However, two to three months after roll-off, the distribution of buffers among recently rolled off loans remains lower than for variable-rate loans; only around half covered the new scheduled payment for three or more months (compared with around two-thirds of variable-rate and split loans). This suggests that some fixed-rate borrowers are either choosing to hold their savings outside their mortgage or, for some reason, this particular cohort have fewer savings than all variable-rate borrowers.

Graph 10

Graph 10: A two panel graph comparing mortgage prepyaments for loans with a recently expired fixed rate and fully variable rate loans. The left panel shows the change in mortgage prepayments for loans in the few months after their fixed-rate expires against the change in prepayments for variable-rate loans over a similar period: 20 per cent of formerly fixed-rate loans have made large prepayments into their mortgage (six months or more of scheduled payments) after their fixed rate expires (compared with 10 per cent of variable-rate loans). The right panel shows the distribution of mortgage prepayments for these expiring fixed-rate loans against variable-rate loans: the share more than 12 months ahead on their mortgage remains lower for expiring fixed-rate loans compared with variable-rate loans, despite the higher share making large transfers in.

Conclusion

Borrowers with fixed-rate loans have faced or will face large, discrete increases in their loan payments when their fixed-rate terms expire. Loans that are yet to roll off will face the largest increases, although these borrowers have also benefited the most from avoiding higher loan payments to date and have had more time to prepare for the rise in mortgage payments. Although higher mortgage payments will strain the finances of some borrowers, most are facing higher interest rates from a position of strength, with very low rates of mortgage arrears, a very low unemployment rate and a high rate of participation in the labour market.

The historically high share of fixed-rate lending during the pandemic means that a rising cash rate will take slightly longer than usual to pass through to mortgage payments for all borrowers. However, aggregate mortgage payments have still increased substantially because the majority of households have variable-rate loans and most fixed-rate mortgages are fixed for relatively short periods of time (RBA 2023b).

Aside from encountering a large step up in their loan payments, borrowers on fixed rates tend to have newer loans and, on some measures, more risky loan characteristics than borrowers on variable rates. While the differences are not large in aggregate, more vulnerable borrowers (such as those with lower incomes, more leverage and first home buyers) are more exposed to large increases in interest rates and typically have fewer margins of adjustment to their financial situation. They should, therefore, be monitored carefully for signs of emerging stress.

Borrowers’ savings buffers can help them to adjust to higher loan payments. Many borrowers on fixed rates have built savings buffers to help them adjust to higher loan servicing obligations. In particular, many borrowers on fixed rates have split loans with sizeable prepayment buffers, and one-third of fully fixed-rate borrowers have also accumulated savings in their mortgages. Many borrowers with fully fixed-rate loans are likely to hold substantial non-mortgage savings: a material share of loans make large transfers in to their mortgage after their fixed rate expires and private survey data shows borrowers have a similar distribution of savings, regardless of the type of interest rate on their loan. The Bank will continue to monitor the expiry of fixed-rate loans closely in the period ahead, especially given its importance for the consumption outlook and, by implication, financial stability.

References

Bergmann M (2020), ‘The Determinants of Mortgage Defaults in Australia – Evidence for the Double-trigger Hypothesis’, RBA Research Discussion Paper No 2020-03.

Bullock M (2022), ‘How Are Households Placed for Interest Rate Increases?’, Speech to the ESA (QLD) Business Lunch, Brisbane, 19 July.

Carse V, A Faferko and R Fitzpatrick (2023), ‘Developments in Banks’ Funding Costs and Lending Rates’, RBA Bulletin, March.

Fernandes K and D Jones (2018), ‘The Reserve Bank’s Securitisation Dataset’, RBA Bulletin, December.

La Cava G and L Wang (2021), ‘The Rise in Household Liquidity’, RBA Research Discussion Paper No 2021-10.

La Cava G, H Hughson and G Kaplan (2016), ‘The Household Cash Flow Channel of Monetary Policy’, RBA Research Discussion Paper No 2016-12.

RBA (2022a), ‘Review of the Yield Target’.

RBA (2022b), ‘Box B: The Impact of Rising Interest Rates and Inflation on Indebted Households’ Cash Flows’, Financial Stability Review, October.

RBA (2022c), ‘Household and Business Finances in Australia’, Financial Stability Review, October.

RBA (2022d), ‘Household and Business Finances in Australia’, Financial Stability Review, April.

RBA (2023a), ‘Box A: Mortgage Interest Payments in Advanced Economies – One Channel of Monetary Policy’, Statement of Monetary Policy, February.

RBA (2023b), ‘Domestic Financial Conditions’, Statement of Monetary Policy, February.

Source: Reserve Bank of Australia [2021]