Home ownership remains high on the political agenda after a decade in which it hit its lowest rate in a generation. At just 64.9 per cent in 2020/2021, home ownership in England sits at levels not seen since the mid-1980s and is well below its 2003 peak of 70.9 per cent. If the rate of ownership returned to the level of the early 2000s, 1.4 million more families would now own their own homes.
The challenges for young people looking to buy their first home exacerbate intergenerational tension at a time when the political environment is highly polarised between young and older voters. But the government has no convincing plan to revive home-ownership rates, and the Labour Party will struggle to win over aspirational voters without a strong policy offer to make it easier for people to buy a home of their own. This report proposes a set of reforms to fill that gap.
Conventional wisdom looks to housing supply to increase home ownership. But there is no compelling reason to think that raising the rate of supply will change the pattern of who owns property. Well-established empirical evidence shows that significantly higher supply will have only a modest impact on house prices over a generation. And, in any case, house prices themselves are not the most important factor in determining home ownership. Almost all first-time buyers (FTBs) rely on mortgage finance to buy, so what matters for raising levels of home ownership is who is able to borrow.
The availability of and access to housing finance has been a key factor in explaining trends in home ownership since 2008. Home ownership has risen rapidly when lenders have been willing to take on the risk of lending to FTBs and fallen sharply when sentiment or the regulatory environment have changed. Tighter lending criteria and more realistic pricing of risk following the global financial crisis of 2008 increased the deposit requirements faced by FTBs and new regulations have limited the amount they are able to borrow.
Lending to FTBs inevitably involves risks. But wild swings in the cost and availability of high loan-to-value (LTV) loans currently lead to volatile home-ownership rates. Since the financial crisis this has resulted in growing intergenerational unfairness. Many more families have been trapped in the private rented sector, including three times as many families with children than in the mid-2000s, even though their rent often outstrips the mortgage payments they would make. Meanwhile those fortunate to buy their first home have typically faced much higher interest payments than wealthier buyers with bigger deposits.
It doesn’t need to be like this. We can raise and stabilise home ownership by ensuring lending to FTBs continues at a sustainable and affordable level across the credit cycle, managing the risks of lending more effectively.
Two Barriers to Home Ownership
There are two types of risk for lenders considering extending mortgage finance to an FTB. One is credit risk: the risk that lenders will incur losses if adverse events mean borrowers are unable to make payments so that the debt cannot be fully repaid from the sale of the property. Many other Organisation for Economic Co-operation and Development (OECD) countries make widespread use of mortgage-insurance systems to stabilise the availability and price of high LTV loans.
The second is interest-rate risk: the risk that mortgage payments will become unaffordable because of rising interest rates. This limits the size of loans that lenders are prepared to extend to borrowers, since borrowers must not only prove that they can afford the initial repayments, but that repayments would remain affordable if interest rates were to rise. Long-term fixed-rate mortgages are used in many countries to eliminate this barrier.
In the UK, these risks are left with lenders and borrowers. Lenders mitigate credit risk by limiting lending for high-LTV mortgages, particularly when those risks are perceived to be high and, if they do lend, charging a premium. Meanwhile, in boom times, risky lending can become excessive. Interest-rate risk is managed by limiting the amount borrowers are allowed to borrow to ensure that repayments remain affordable even if interest rates were to rise. These restrictions on borrowing leave many prospective FTBs unable to get on the property ladder.
Which of these two risks is the bigger barrier to home ownership? Our analysis of the available data suggests that most young non-homeowners are some way from overcoming the deposit and income constraints needed to assuage lenders’ fears. This static analysis suggests that increasing the availability of 95 per cent LTV mortgages and allowing people to borrow at higher income multiples appears to have only a limited benefit in terms of home ownership.
However, there is evidence that many prospective FTBs can raise a deposit from other sources if necessary. Of those who purchased a property between the 2014–2016 and 2016–2018 waves of the Wealth and Assets Survey (WAS), only 7 per cent appeared to have sufficient savings to afford a deposit in the earlier wave. On average, they were £10,000 short of the savings required to put down a 10 per cent deposit on the average FTB property in their area. Easing the deposit constraint would therefore likely allow many who do not have sufficient savings to afford a 10 per cent deposit to get on to the housing ladder, whether through savings or with the help of their families. Boosting home ownership without jeopardising financial stability will therefore require policies both to increase the availability and lower the price of high-LTV mortgages across the cycle through greater use of insurance, and to encourage greater availability of long-term fixed-rate mortgages.
A Comprehensive System of Mortgage Insurance
The UK has a mortgage guarantee scheme (MGS) that will remain in operation until the end of 2022 and which is closely modelled on the previous Help to Buy mortgage guarantee that ran from 2013 to 2016 (note that this is different from the Help to Buy equity loan scheme for new build properties). But the scheme has not had a material impact on home ownership: loans under the scheme only account for around 5 per cent of FTB mortgages issued, and relative costs of high-LTV mortgages have remained high, well above the cost of participating in the scheme. The MGS remains a peripheral part of the market, limiting its impact in an environment where lenders are risk averse and regulators are cautious.
It is difficult to argue that regulations on mortgage lending introduced since the financial crisis are responsible for the huge reduction in high-LTV lending. Limits on lending amounts – through the 4.5 times income “flow limit” and affordability tests – affect the number of people who qualify for a mortgage and hence demand. But there is no obvious reason why they would affect the take-up of 95 per cent LTV mortgages but not 90 per cent or 85 per cent products, whose availability has been much less volatile. Moreover, the disappearance of high-LTV products occurred in the wake of the financial crisis and well before the new regulations came into force. Higher costs cannot be explained by additional capital requirements for high-LTV mortgages introduced as part of the Basel regulations either. It appears that the supply of high LTV loans has fallen faster than demand.
Instead, less concrete factors such as risk aversion among lenders and a cautious regulatory environment seem to be to blame for the decline in high-LTV lending seen since the financial crisis. Lenders seek to balance their books between high- and low-LTV lending to avoid becoming too exposed to swings in the property market, particularly when risks are perceived to be high. Regulators also put pressure on lenders to avoid overexposure to the high-LTV market. This has limited the supply of high-LTV mortgages, even with the MGS in place.
In other countries where mortgage-insurance schemes are much more mainstream or compulsory, the lower riskiness of insured mortgages is reflected in their cost and availability as this has become accepted among lenders and regulators. This dynamic is unlikely to unfold in the UK unless mortgage insurance is both permanent and compulsory for high-LTV lending.
We therefore propose that the scheme be made permanent and compulsory for all mortgages above 80 per cent LTV. The risk of a voluntary scheme is that little will change: few lenders might participate, causing high-LTV lending to remain scarce and costly. Meanwhile, a temporary scheme deployed only when risk is high is likely to be more expensive than a permanent one that operates across the whole business cycle. The combination of permanent and compulsory insurance will prevent adverse selection into the scheme, both in terms of which lenders choose to enter and when they choose to participate. The result should be a significantly lower insurance premium compared with what has been charged under the government’s recent schemes. This would ensure a steady stream of affordable FTB lending at times of economic uncertainty, but also temper excessive lending during boom times.
A further problem with the MGS is that, owing to its design, lenders typically struggle to claim the capital relief that should accompany insurance, hence raising borrowing costs for FTBs above what they should be. To tackle this, the Credit Risk Mitigation framework should be simplified to allow lenders to claim capital relief on insured mortgages. This would lower the cost of a 95 per cent LTV mortgage by around 20 basis points.
Fostering Long-Term Fixed-Rate Mortgages
Greater availability of long-term fixed-rate mortgages would also be desirable. This would help those in steady but lower-paid employment to borrow more to move into home ownership, which would in many cases be cheaper than paying rent. It would have other advantages too. More risk-averse consumers would likely welcome the chance to remove interest-rate risk, and macroeconomic stability might be improved if the transmission mechanism of monetary policy worked less through the housing market and more through business-lending and exchange-rate channels.
The UK is unusual among advanced economies in not having long-term fixed-rate mortgages as a mainstream option. But mortgage markets in those advanced economies where long-term fixed-rate mortgages are the norm are very different. Introducing the sort of institutions that exist in these countries would be highly disruptive with uncertain benefits. The existing market works well for many borrowers who would prefer to fix their interest for five years or less at a lower rate, rather than paying a higher rate to insure against interest-rate risk more fully. Moving wholesale to a different market structure would not be in the interests of many consumers.
However, we find that there are both demand- and supply-side barriers to the greater use of long-term fixed-rate mortgages. These have arisen largely because the regulatory environment has tended to focus on the dominant bank-lending model, inadvertently making it harder for a more diverse market to emerge. Three changes could help to broaden the range of products available to FTBs.
First, on the demand side, affordability restrictions designed for short-term lending are a significant barrier to longer-term fixed products. Limits on issuing mortgages representing more than 4.5 times the borrower’s income constrain demand from one of the key groups who would be likely to opt for a long-term fixed-rate product so that they can borrow more. The rationale behind this limit – that high loan-to-income (LTI) mortgages risk becoming unaffordable if interest rates rise – does not apply to mortgages where the interest rate is fixed for the full term, so there is a strong case that long-term fixed-rate mortgages should be excluded from this limit. At current interest rates, standard affordability calculations would imply a limit of seven times income for long-term fixed-rate products.
A second demand-side barrier is the fear among brokers and lenders that they could stand accused of mis-selling long-term fixed-rate mortgages. If interest rates subsequently fall, choosing a long-term fixed-rate mortgage might appear a poor decision in retrospect. And for those borrowing at high LTVs, it is typically cheaper over the lifetime of the mortgage to initially choose a shorter-term fixed-rate mortgage and then refinance at a lower LTV later on. But in neither case is it necessarily bad advice to go for a long-term fixed-rate mortgage.
Choosing the certainty of a long-term fixed-rate mortgage is a good idea for those who want to insure themselves against the possibility of higher interest rates. Similarly, choosing a long-term fixed-rate mortgage even at a high LTV may be a good choice if it is the only way of entering home ownership. A long-term fixed-rate mortgage might work out cheaper than renting and does not preclude refinancing on to a cheaper rate later on. Regulators should make clear that they are neutral about short- and long-term fixed-rate mortgages and that advising consumers to choose a long-term fixed-rate mortgage in these circumstances would not constitute mis-selling.
On the supply side, banks are not well placed to offer long-term fixed-rate mortgages because they rely on short-term deposit funding. To overcome this “maturity mismatch”, mortgages in other advanced economies are securitised and purchased by life insurers and pension funds. If this happened in the UK, it could open the way to more and cheaper long-term fixed-rate products coming on to the market. However, these assets are likely to be less attractive to UK life insurers.
UK life insurers can obtain very favourable capital requirements through the “matching adjustment” if the term structure of their assets matches that of their liabilities. However, prepayment risk – the risk that borrowers will pay back the mortgage early rather than holding it for the full term – makes it difficult for such mortgage-backed assets to qualify for the matching adjustment, particularly as there is no track record on the prepayment rate for these assets in the UK.
To encourage the development of this market, regulators could reform Solvency II regulations so that it becomes possible for assets with prepayment risk to qualify for the matching adjustment if additional capital were held against such a risk. A further barrier to these products could be removed if data on prepayment rates from other countries could be accepted as a way to gauge the likely extent of prepayment.
A Better-Functioning Housing Market
A potential criticism of these measures is that they will simply boost leverage and increase house prices across the board, leaving FTBs no better off. This is misconceived for a few reasons. First, outside financial bubbles, house prices are determined by economic fundamentals such as interest rates and rent levels. Lending follows these market signals rather than affecting prices without reference to any underlying economic value.
Second, the mortgage-insurance premium should have a calming effect on the market in boom times, preventing risky lending from becoming excessive, or excessively cheap, as happened in the run up to 2008. Overwhelmingly, the effect of mortgage insurance would be to lower the cost of high-LTV loans compared with those for wealthier borrowers, because this is where credit risk is most acute. Similarly, long-term fixed-rate mortgages would only increase the borrowing capacity of a small number of FTBs who are currently constrained by the 4.5-times-income limit.
Consequently, these policies should redistribute credit towards FTBs, rather than increasing the volume of lending. But to ensure these reforms do not present any risk of raising prices, it may be desirable to complement their introduction by offsetting fiscal or regulatory measures on rental property or second homes.
It should be possible to return safely to the pre-crisis home-ownership levels of close to 70 per cent without jeopardising financial stability. But doing so will require careful and conscious reform of how our mortgage market functions so that home-ownership aspirations are no longer exposed to volatile risk appetites in the financial sector. More broadly, helping families to realise their ownership aspirations need not come at the expense of much-needed and complementary reforms to the social and private rented sectors, so that everyone is freer to choose the right tenure for their needs.
Editor’s Note: We are grateful to Lloyds Banking Group for their support towards the research presented in this report, as part of Lloyds Banking Group’s support for broadening access to housing. The analysis and conclusions of this report are solely the responsibility of the Institute.
The rapid decline in the rate of home ownership since the start of the century has focused much policy attention on how the government might raise it. At just 64.9 per cent in 2020/2021, home ownership in England now sits at around the level previously seen in the mid-1980s and is significantly lower than its peak of 70.9 per cent in 2003. If the rate of ownership were returned to its early-2000s level, more than 1.4 million more families would own their homes.
Figure 1 – Home-ownership rates in England since 1980
Source: English Housing Survey
The challenges for young people looking to buy their first home exacerbate intergenerational tension at a time when the political environment is highly polarised between younger and older voters. But ideas for how to get more families on to the housing ladder are thin on the ground.
The government has no convincing plan to tackle the problem. The Labour Party also needs a convincing proposal for making it easier for people to buy a home of their own if it is to win over aspirational voters. This report proposes a set of reforms to fill that gap.
Policy discussions typically focus on high house prices as the underlying cause of falling home ownership and identify increasing housing supply as the key policy response. High prices are inevitably a barrier to ownership because, for any given LTV, they require buyers to amass a bigger deposit.
However, a wide array of evidence – including that used by the Department for Levelling Up, Housing & Communities – shows that even significantly higher levels of housing supply would have only a modest impact on house prices over a generation.[_] Moreover, house prices themselves are not the most important factor in determining who owns property. As we showed in the first report in this series,[_] the availability of and access to housing finance have been more important drivers in home-ownership trends over the past 40 years because almost all FTBs rely on mortgage finance to buy. Home ownership has risen rapidly when lenders have been willing to take on the risk of lending to FTBs and fallen sharply when sentiment or the regulatory environment have changed, particularly in the wake of the financial crisis.
The acceleration of home ownership during the 1980s, when the rate jumped by more than ten percentage points in England, was driven not just by fiscal policies such as the Right to Buy and mortgage-interest relief at source (MIRAS), but by an expansion of the availability of credit to prospective FTBs. Foreign-exchange controls were lifted, which allowed foreign banks to compete in the UK as well as permitting UK banks to access overseas funding. Liquidity restrictions on banks were eased, allowing them to massively expand their share of the mortgage market from playing a marginal role to being the dominant source of mortgage finance. Meanwhile building societies, having dominated the scene before the 1980s with more than 80 per cent of mortgages, had to be allowed access to wholesale funding to remain competitive in the new environment.[_]
Figure 2 – Deposit requirements jumped after the financial crisis and remain high
Source: CML, UK Finance, TBI calculations
Similarly, the sharp drop in home ownership following the global financial crisis of 2008 was largely caused by more restrictive lending practices for FTBs rather than higher house prices. Despite price falls between 2008 and 2012, home-ownership rates declined as the heightened risk for lenders led them to radically tighten their lending criteria. The median LTV for FTBs dropped from 90 per cent in 2007 to just 75 per cent in 2009 (Figure 2). This caused a sharp jump in FTB deposit requirements as a multiple of income, from around 40 per cent to 100 per cent even as house prices fell.
More recently, three major regulatory changes appear to have slowed the recovery in lending volumes to FTBs. First, Basel 2 rules came into effect in early 2008, requiring lenders to hold more capital against higher-LTV loans. This typically required lenders to hold four to five times more capital against high-LTV compared with low-LTV loans,[_] and, combined with more realistic pricing of risk following the experience of the financial crisis, led to a large gap opening up between the interest rates on high- and low-LTV mortgages.[_] This “spread” arises because of the perceived riskiness of lending, the additional cost of capital and the extra administrative cost of processing and managing a higher-LTV mortgage.
In more recent years, availability of high-LTV mortgages has been aided somewhat by government mortgage guarantees. The Help to Buy mortgage guarantee was in place between 2013 and 2016 and a similar scheme was introduced in 2021 to ensure continued availability of mortgages at 95 per cent LTV during the pandemic. Combined with higher taxes on landlords, this may have supported the slight recovery in home ownership after 2016.
Two other measures introduced since the financial crisis limit the maximum amounts FTBs are allowed to borrow. Following the mortgage market review in 2014, the Financial Conduct Authority (FCA) introduced a “stress test” to ensure that borrowers would be able to afford to make their mortgage repayments if interest rates rose by three percentage points above the lender's standard variable rate during the first five years of the loan.[_] Additionally, the Financial Policy Committee of the Bank of England imposed a 15 per cent ceiling on the proportion of each institution's loans that could be issued at or above 4.5 times the borrower's income.
These measures sought to enhance financial stability by limiting the risk of losses to lenders if interest rates rose or adverse events in borrowers’ lives meant they were unable to meet mortgage payments. The result has been much more difficult access to finance for FTBs.
At first glance, this suggests that there may be a simple trade-off between financial stability and home ownership, but this is not necessarily the case. As we discussed in the first report in this series, other advanced economies around the world make use of tools to manage credit risk and interest-rate risk.[_] Many countries, most notably Canada, Australia and the Netherlands, have well-developed systems of mortgage insurance to protect lenders against credit risk. In others, such as Denmark and the United States, long-term fixed-rate mortgages are the norm, protecting borrowers against the risk of rising interest rates.
The UK is unusual in having neither of these features. This leaves the risk of rising interest rates on households and increases the cost to banks of lending to FTBs. As a result, lending at high LTVs is expensive and even unavailable at points during the business cycle. The consequence is that lending to FTBs is volatile and the level of home ownership is lower than it needs to be. Many families who would, in previous generations, become homeowners are left in the private rented sector, facing higher housing costs and with weak security of tenure, despite being able to afford monthly mortgage repayments.[_] There are now more than 1.4 million households in England who would be owner-occupiers if the rate of home ownership regained its early-2000s level.
In this paper, we investigate whether similar measures to those found in other advanced economies could be beneficial in the UK and, if so, how they should be designed. We begin by examining the scale and nature of the problem faced by prospective FTBs. Based on this information and the experiences of mortgage markets in other advanced economies, we identify measures that would increase the availability of high-LTV loans and then consider interventions that would increase the availability of long-term fixed-rate mortgages.
Banks lending to FTBs face two risks. First, since FTBs are typically able to put down only a relatively small deposit towards their property purchase, there is the risk that they will be unable to recoup the loan from the proceeds of the sale in the event that the borrower is unable to make repayments because of unemployment, disability or relationship breakdown. We refer to this as “credit risk”. In the UK, lenders deal with this risk by limiting LTV ratios and imposing higher interest rates on high-LTV loans.
Second, if FTBs borrow close to the limit of their affordability, future interest-rate rises may push them over the limit and leave them unable to make repayments. We refer to this as “interest-rate risk”. In the UK, lenders and regulators reduce this risk by limiting the amount they will lend so that repayments would remain affordable even with a significant increase in interest rates.
Prospective FTBs must, then, accumulate sufficient savings to put down a large enough deposit to assuage lenders’ concerns about credit risk, and have a sufficiently high income not just to be able to afford mortgage repayments under prevailing interest rates, but under significantly higher rates too. But which of these is the more significant constraint to home ownership today? Data from the WAS sheds some light on this question.[_]
Figure 3 – How far away are non-homeowners aged between 25 and 44 from being able to purchase an average FTB property in their region?
Source: TBI calculations using 2016-18 WAS (Note: Assumes minimum 10 per cent deposit and maximum loan of 4.5 times income. In this analysis, “families” are benefit units: that is, an individual plus their partner and dependent children if they have them.)
Source: TBI calculations using 2016-18 WAS (Note: Assumes minimum 10 per cent deposit and maximum loan of 4.5 times income. In this analysis, “families” are benefit units: that is, an individual plus their partner and dependent children if they have them.)
Building on work by the Resolution Foundation,[_] we examine the accumulated savings and earnings levels of non-homeowning families where all adults are aged under 45; we then compare this to the savings and earnings levels they would need to obtain a mortgage for a typical FTB property in their region. Currently, less than 5 per cent of mortgages issued are above 90 per cent LTV, so a 10 per cent deposit can be thought of as the effective minimum for most prospective buyers. We therefore assume a minimum 10 per cent deposit and a maximum mortgage advance of 4.5 times gross earnings.
In practice, the maximum advance will depend on the affordability test, which assesses the borrowers’ ability to make payments if the reversion rate increased by three percentage points. The 4.5 times earnings limit is a reasonable approximation to this; there is also a flow limit on the number of mortgages banks can issue with a loan-to-income (LTI) ratio of more than 4.5. This has reduced banks’ willingness to lend above 4.5 times income.[_] This limit will become more binding if the Bank of England withdraws its recommendation that affordability calculations use an interest rate of the reversion rate plus three per cent as the “stress rate” and the minimum stress rate reverts to the reversion rate plus 1 per cent, in line with the FCA’s guidelines.[_]
On this basis, fewer than 5 per cent of non-homeowners would be eligible for a mortgage, confirming that it is affordability constraints rather than a preference for renting that keeps renters from becoming FTBs. Almost everyone with enough savings to put down a 10 per cent deposit, and enough income to service a mortgage, chooses to purchase a property. Of non-homeowners, 18 per cent have a sufficiently high income to obtain a mortgage but do not have the requisite deposit, while 2 per cent have enough savings to put down a deposit but not enough income to qualify for a mortgage, with 76 per cent falling short on both counts.
This suggests that the deposit constraint is currently the most significant barrier to home ownership. And those who fall short on this front appear to do so by a long way: on average, those with insufficient savings would need to find another £18,500 on average to reach a 10 per cent deposit, and families with insufficient earnings would need to earn £30,000 more.
Figure 4 – What if the constraints are relaxed?
What if the constraints are relaxed?
How could policy change this picture? LTV mortgages of 95 per cent have all but disappeared since the financial crisis. Increasing their availability would reduce the deposit requirement. And long-term fixed-rate mortgages might allow banks to lend a larger LTI multiple as interest-rate risk would be eliminated. A higher LTI multiple of seven, which would still be affordable with a 3 per cent interest rate that was fixed for term, could be feasible in this scenario. On the face of it, easing the deposit and affordability constraints in this way would make only a limited difference to the home-ownership rate. Fewer than 9 per cent of non-homeowners aged between 25 and 44 would be able to purchase the average FTB property in their region even after these changes. The implication is that these changes might increase the home-ownership rate by only one-third of 1 per cent.[_]
However, there is good reason to conclude that these reforms might have a larger effect than we can observe from a static look at these data. First, dynamic effects might be important: if a deposit is within reach, renters may save more rather than resign themselves to a lifetime of renting.
Second, we can only observe the level of savings held by prospective FTBs themselves rather than the full amount of money that they may be able to access. Most importantly, the “Bank of Mum and Dad” is an important source of housing deposits. Using data from the English Housing Survey, the Resolution Foundation reports that a third of FTBs relied on family support for some of their deposit.[_]
Consequently, a lower deposit requirement may put home ownership within reach of others whose family can provide some financial support. Indeed, 20.5 per cent of 25- to 44-year-old non-homeowning households (approximately 437,500) expect to receive an inheritance, and 90,000 of the households expecting an inheritance only fail the deposit constraint. It may well be the case that some of those who expect to receive an inheritance could be in line to receive a gift to help them cross the eligibility threshold (the median inheritance for those born in the 1980s is estimated to be £136,000),[_] especially if policy interventions were to relax the constraint.
The central role of unreported sources of wealth for deposits becomes clearer when we exploit the longitudinal element of the WAS data. This allows us to examine the level of savings of families in the 2014–2016 round who subsequently become homeowners by the time of the 2016–2018 round.[_] If all sources of wealth and income that FTBs use are fully captured in this analysis, then we should see that those who moved into ownership within two years were already close to the threshold for meeting the deposit and income constraints in the earlier period.
But this is not what the data show. Looking at the deposit constraint, the vast majority of those who subsequently became homeowners appeared to have insufficient savings in the initial period, falling short by an average of £10,500. As a group, these soon-to-be homeowners appeared broadly indistinguishable from other non-homeowning households in terms of their savings in the year or so before they bought, yet somehow they managed to find the necessary funds to buy. This cannot be explained by their choosing to purchase a cheaper property than envisaged in these calculations: using the difference between the price paid for their property and their outstanding mortgage to calculate the average deposit they actually put down, we can estimate that they were able to find an average of £32,500 more than their savings two years previously.
When it comes to the income constraint, by contrast, these families do appear to be closer to the line, with more than six in ten having enough income to buy, compared to around two in ten of the non-homeowner group as a whole.
Figure 5 – Many FTBs appeared to lack the savings for a deposit two years earlier
Source: TBI calculations using 2014–2016 and 2016–2018 WAS (Note: Covers FTBs aged between 25 and 44. Assumes minimum deposit of 10 per cent and maximum loan of 4.5 times income.)
This analysis tells us that while many prospective FTBs appear to lack the savings required for a deposit, this is not necessarily a reliable guide to whether they are able to access a deposit from other sources. It is likely that reducing the deposit required would enable many more families to access smaller gifts to help them get over the line and could therefore have a significant impact on home ownership.
Similarly, allowing FTBs to borrow a higher multiple of their income through greater availability of long-term fixed-rate mortgages would likely also help. When borrowers can borrow seven times their income, around 40 per cent of non-homeowners aged 25 to 44 would be able to borrow enough to purchase the average FTB property in their region, compared with only 20 per cent under the typical income-multiple restrictions in force today.
In the following two sections, we examine the constraints that prevent wider availability of both high-LTV lending and long-term fixed-rate mortgages and consider how policy reforms could address this.
We have seen that many prospective FTBs have the necessary income to qualify for a mortgage but are unable to raise the 10 per cent deposit that would typically enable them to move into home ownership at a reasonable cost. This underlines the need for accessible 95 per cent LTV mortgages, and reducing the spreads over lower LTV products.
The typical deposit requirement has fluctuated sharply over the past 15 years as lending to FTBs has waxed and waned. That volatility of FTB lending is reflected in mortgage-interest rates. Interest-rate spreads between high- and lower-LTV mortgages – the difference in the interest rates charged for each type of product – reflect the availability of high-LTV lending. When lenders are risk averse or regulations restrict the number of high-LTV loans they are able to offer, high-LTV volumes drop and their relative price increases. The result is fewer FTBs and falling home ownership.
We can see that these spreads have been very large for much of the past decade, especially compared with the pre-crisis period. This reflects the pullback in lending to families with smaller deposits that is largely responsible for the reduction in home ownership in recent years (Figure 6).
Figure 6 – Recent years have seen fewer loans at high LTVs and larger spreads
Source: TBI calculations using Bank of England and FCA data from Q1 2007 to Q4 2021 (insufficient data to calculate spreads from Q3 2008 to Q3 2013 inclusive).
The government has responded to this lending drought twice in the past decade: first in 2013 as lending to FTBs struggled to recover in the wake of the financial crisis, and again in 2021 in response to the Covid-19 pandemic. Both the Help to Buy mortgage guarantee and the current MGS have aimed to stimulate lending at higher LTV ratios during periods of risk aversion by lenders.
Under the MGS, in exchange for a one-off fee of 90 basis points, the government insures lenders’ mortgages with an LTV between 91 and 95 per cent. The insurance covers 95 per cent of any losses on the portion of the loan above 80 per cent LTV that cannot be recouped from the sale of the property. The hope is that by removing risk, lenders will be prepared to offer high-LTV mortgages – even in more turbulent times and at lower cost thanks to lower expected losses and capital requirements. The earlier Help to Buy scheme was similar in design, although it also offered insurance for lower-LTV mortgages at premiums of 46 basis points for the 85 to 90 per cent and 28 basis points for the 80 to 85 per cent LTV bands.[_]
Similar policies exist in other advanced economies. Our earlier report, exploring housing-finance models around the world, included a detailed examination of the situation in several different countries.[_] The most relevant to the UK appear to be Canada and the Netherlands, both of which have state-backed mortgage-finance institutions that provide mortgage insurance with some similarities to the UK schemes.
In Canada, mortgage insurance provided either by the Canada Mortgage Housing Corporation (CMHC) or a private provider is compulsory on loans above 80 per cent LTV. Since the CMHC is backed by a government guarantee, CMHC-insured mortgages are treated in the same way as government debt and therefore attract a zero-risk weight (that is, lenders do not have to retain capital against the risk of default). Private insurers can also obtain government backing through catastrophic insurance that would pay out in the event of the insurer’s collapse, so privately insured loans also attract a low-risk weight. As a result, there is no interest-rate spread between high- and low-LTV loans for insured mortgages in Canada, and the only difference in cost between the two products is a one-off insurance premium of up to 4 per cent of the loan amount, which can be added to the loan. Indeed, the interest rate on a lower-LTV uninsured mortgage can be higher than that for an insured mortgage at 95 per cent LTV.
In the Netherlands, the National Mortgage Guarantee Scheme (NHG, or Nationale Hypotheek Garantie) operated by the Homeownership Guarantee Fund (Waarborgfonds Eigen Woningen, or WEW) is more targeted and limited, but still effective at reducing the cost of loans to FTBs. Insurance takes over mortgage repayments in the case of unemployment, disability, relationship breakdown or widowhood and covers any losses that arise if foreclosure is necessary and proceeds from the sale of the property are insufficient to cover the outstanding loan. Using the scheme is not compulsory for the homebuyer but is an attractive option as the fee is low (a one-off fee of 0.6 per cent of the loan amount) and mortgage rates are 50 basis points lower than for equivalent uninsured loans. Eligibility is restricted, however, to purchases of €355,000 or less, somewhat lower than the average house price in the Netherlands, which is more than €400,000.
Figure 7 – High-LTV mortgages are cheaper in countries with widely used mortgage insurance, by country, Q4 2021
Source: TBI calculations. (Note: The figure shows the spread – i.e, the additional annual mortgage cost as the share of the respective loan – for 95 per cent versus 75 per cent LTV mortgage, plus mortgage insurance where appropriate. For this chart we have used Q4 2021 data. Note that given the very limited take-up of the MGS, we assume that no insurance is used for the 95 per cent LTV mortgage in the UK. Also, given widespread availability of insurance of 75 per cent LTV mortgages in the Netherlands, we assume that both 95 per cent LTV and 75 per cent LTV mortgages are insured and hence there is no additional insurance cost. Assumes UK average house price and five-year fixed-rate mortgage repaid over 20 years from a major mortgage lender. UK: Halifax; Canada: CIBC; the Netherlands: ABN Amro; Hong Kong: HSBC; Australia: Commonwealth Bank; US: Bank of America.)
Both these schemes appear to be more effective at limiting costs to FTBs than the UK’s MGS. The MGS has failed to reduce the spread of a typical 95 per cent LTV over a 75 per cent LTV one, which was over 120 basis points in Q4 2021. By contrast, the spread in borrowing costs – interest payments plus the cost of insurance – was much smaller in countries with widely used mortgage-insurance schemes. In Canada the difference in cost was 26 basis points and in Australia it stood slightly above 50 basis points, while in the Netherlands there was no cost difference between high- and lower-LTV products.
As well as being high, the additional costs for high-LTV borrowers have been remarkably volatile in recent years. Movements in the spread have tended to reflect changes in the risk environment. However, throughout the past 15 years, spreads have been much higher than the risk of high-LTV lending could plausibly justify.[_]
Why have the UK schemes not had the same effect on the cost and availability of finance to FTBs as those in Canada and the Netherlands?
To begin with, take-up of the guarantees by lenders in the UK has been low. During the core years of the Help to Buy mortgage guarantee, just over 10 per cent of FTB loans issued were backed by the scheme (Figure 8).[_] And so far, the proportion of new FTB mortgages using the MGS scheme appears to be even lower.[_]
Figure 8 – Number of FTB loans with and without the backing of the Help to Buy mortgage guarantee, 2013 to 2017
Source: HM Treasury, UK Finance, TBI calculations
It is also difficult to argue that regulations on mortgage lending introduced since the financial crisis have reduced lending at high LTVs to the extent observed. Limits on lending amounts through the 4.5 times income flow limit and affordability test affect the number of people who qualify for a mortgage and hence demand for high-LTV mortgages. But there is no obvious reason why they would affect the take-up of 95 per cent LTV mortgages but not 90 per cent or 85 per cent products, whose availability has been much less volatile.[_] Moreover, the disappearance of high-LTV products occurred in the wake of the financial crisis and well before the new regulations came into force. Higher costs cannot be explained by additional capital requirements for high-LTV mortgages introduced as part of the Basel regulations either – research by the Bank of England suggests that capital requirements add 20 basis points to the cost of a 95 per cent LTV mortgage compared with one with an LTV of 75–80 per cent. [_],[_] It appears that the supply of high LTV loans has fallen faster than demand.[_]
Rather, risk aversion among lenders and a cautious regulatory environment have been the main barriers to FTB lending. The interplay between two factors appears to matter here. Lenders wish to balance their books between high- and low-LTV lending to avoid becoming too exposed to swings in the property market. Regulators may also put pressure on lenders to avoid overexposure to the high-LTV market. As a result, competitive pressures in the high-LTV market segments are weak. Lenders limit their lending at high LTVs by setting prices at high LTVs above the cost of expected losses. Moreover, lenders’ desired balance of high- and low-LTV mortgages depends on the risk environment, leading to lower availability and higher spreads when the risk of lending is perceived to be high.
This is very different to what happens in mortgage markets in countries where mortgage insurance at high LTVs is the norm. Where these systems are in place, insured mortgages are seen by both lenders and regulators as safer than those that are not insured, even those with a much lower LTV. Recall that in Canada, interest rates on insured mortgages at 95 per cent LTV are typically lower than those on uninsured mortgages at 75 per cent LTV. Similarly, in the Netherlands, the interest rate on insured mortgages with high LTVs is around the same as for uninsured mortgages at 65 per cent LTV. The approach of lenders across the cycle is also different. Whereas UK mortgage lenders retreated from the high-LTV segments following the global financial crisis, their Dutch counterparts decided to make greater use of the NHG, thus allowing high-LTV lending to continue.[_] A total of 90 per cent of potentially eligible new mortgages in the Netherlands took advantage of the NHG in 2011, a big increase from the pre-crisis level of 41 per cent in 2007; the share has subsequently declined as the economy has recovered.[_]
In the UK, by contrast, mortgages insured under the MGS do not receive this recognition as being less risky. One symptom of this concerns the capital requirements for insured mortgages. In principle, the Prudential Regulation Authority (PRA) offers capital relief to lenders whose mortgages are backed by insurance.[_] This should reduce the cost of issuing loans to FTBs and therefore close the gap in borrowing costs for them relative to other buyers with larger deposits.
But in practice the MGS does not effectively provide capital relief. The reason is that the mortgage guarantee is classified as a securitisation and is subject to the complicated European Credit Risk Mitigation framework.[_] Lenders we spoke to while researching this paper said that the process for applying for capital relief was too onerous to be worthwhile. Consequently, at present, banks that participate in the scheme pay the cost of insuring mortgages but also have to pay the cost of capital requirements designed to protect them against losses. This makes mortgages insured under the scheme unnecessarily costly and those costs are passed on to FTBs.
Now that the UK has left the EU, it is looking to introduce different rules that should make it easier for banks to claim relief on the insured mortgages.[_] If successful, this would reduce interest-rate spreads between higher and lower LTV when the MGS is in operation. As discussed above, research by the Bank of England suggests that capital requirements add 20 basis points to the cost of a 95 per cent LTV mortgage compared with one with an LTV of 75–80 per cent, so the spread could be expected to reduce by this amount if capital requirements were calculated properly.
But while marginally lowering the cost of such products would help, the bigger barrier is a lack of availability. Changing that will require a broader shift in attitudes towards the riskiness of insured mortgages in the UK mortgage market. As we have seen, spreads have long exceeded the additional riskiness of high-LTV mortgages even with the MGS in place. Only by entrenching an insurance model at the heart of the FTB mortgage market will lenders and regulators begin to recognise insured mortgages as less risky and treat them accordingly. This should then boost the availability of high-LTV products and reduce their associated spreads, as observed in other countries with greater experience of these products.
The remainder of this section discusses five design questions that, if addressed, would allow a reformed MGS to be more effective in supporting home ownership. These include:
Should the mortgage-insurance scheme be permanent or temporary?
Should it be voluntary or compulsory?
Should insurance be publicly or privately provided?
How extensively should insurance be used within the market?
How much of a loan should be insured?
A Permanent or Temporary Insurance Scheme?
On the two occasions when a mortgage-insurance scheme has been used in the UK, 95 per cent LTV loans had essentially disappeared from the market and, even at slightly lower LTVs, spreads had increased due to higher perceived risk.
It is reasonable to deploy the MGS as a tool to shore up home ownership by easing lending only during the tightest points in the credit cycle, as HM Treasury has been doing since 2013. But there are practical difficulties with this approach that go to the heart of the scheme’s limited impact.
It is not straightforward to activate an MGS quickly. Even once a decision has been made to start the scheme, banks must apply to join and then their applications need to be processed – and these procedures take time. Often, by the time banks’ applications have gone through the system, the worst of the credit crunch is over, with the damage to home-ownership levels already done, as was the case by 2013.
The case for a permanent MGS becomes even stronger if we consider that spreads have never got close to the real cost of higher-LTV lending even with the MGS in place. Making insurance permanent would help shift the MGS from a peripheral part of the market that is only in place when risks are perceived to be high to an integral part of the environment. Once lenders and regulators realise that insured mortgages are less risky, fluctuations in spreads over the cycle should disappear, as is the case in countries like Canada and the Netherlands.
There is therefore a strong case for a permanent MGS: once firmly established, it should not only attenuate swings in interest-rate spreads, but reduce the cost and increase the availability of high-LTV mortgages across the cycle. It should also be cheaper than a scheme that is only available during crises when risk is higher.
Voluntary or Compulsory Insurance?
Lenders do not have to participate in the MGS at present. Indeed, Nationwide (the second-largest mortgage lender in the UK) has declined to participate in the current scheme and was at one point able to undercut other lenders and (presumably) achieve larger margins on its high-LTV lending as a result.[_] Mortgage insurance is compulsory above 80 per cent LTV in Canada, but this is rare elsewhere. In other advanced economies, insurance is attractive either by virtue of reducing the mortgage-interest rate (the Netherlands), because major lenders insist upon it (Australia) or because it makes mortgages eligible for securitisation by a government-sponsored enterprise (US).
If, as described above, a permanent MGS led to greater willingness to lend at high LTVs among participating lenders in the same way as occurs in Canada and the Netherlands, other lenders would also likely find it attractive to participate. They would not be able to compete with participating lenders offering much lower interest rates.
However, the risk of a voluntary scheme is that take-up will remain low and little will change. A compulsory scheme is much more likely to shift the burden of risk management away from the current reliance on blunt and intangible lender attitudes to risk and towards an insurance mechanism.
Given the UK’s history, making mortgage insurance compulsory might give pause for thought. Mortgage-indemnity guarantees existed in the UK for many years but the market collapsed following the property crash of the early 1990s.[_] However, the shortcomings of the UK’s old mortgage-indemnity market are well documented, and both the experience of the recent government guarantee schemes and their long-established counterparts in other countries show that it is eminently possible to create a stable and reliable mortgage-insurance scheme.
A further case for compulsory insurance is that it avoids the problem of banks only choosing to participate when they perceive the risk of high-LTV lending as being high. Such adverse selection would lead to insurance premiums being higher and therefore reduce participation further. Mortgage insurance that is both permanent and compulsory, by contrast, avoids the risk of adverse selection, both in terms of which lenders choose to enter the scheme and when they choose to participate.
Private or State Insurance?
A further question that arises is whether insurance from private providers should be permitted as an alternative to the state-backed mortgage guarantee. There are arguments on both sides. Bringing private providers into the market could help price discovery, potentially lowering costs if the government is overestimating the riskiness of high-LTV lending.
Allowing private providers to insure lenders would not be without risk: it is likely that they would vary their insurance premiums over the cycle in a way that contributes to the cyclicality in lending that these reforms should seek to prevent. And the experience of the UK and other advanced nations has shown that private mortgage insurance must be heavily regulated: only monoline insurers (that is, those who specialise in mortgage insurance and do not offer any other insurance products) should be able to enter this area, and they should have to retain a certain proportion of premiums in reserve for at least ten years in case there is an economic downturn that significantly increases claims.[_] Nonetheless, it is still possible that private insurers could fail if several loans default simultaneously as their business model is based on risk pooling, a situation that is not possible with a state-backed insurer. Finally, privately insured mortgages would incur a higher (non-zero) risk weight as such products will no longer be considered equivalent to government debt. This would make it harder for private companies to compete with a state-backed scheme.
How Extensively Should Insurance Be Made Available?
The fourth design question for mortgage insurance relates to how widely insurance should be offered. For example, there is little need to insure low-LTV loans at all since the owner has substantial equity to absorb any losses, raising the question of what the appropriate LTV threshold should be. Similarly, should there be a house-price limit above which it is not seen as a public-policy priority to offer or mandate insurance?
The MGS is currently available only on mortgages for owner-occupation with an LTV ratio between 91 per cent and 95 per cent where the purchase price is less than £600,000. Its forerunner, the Help to Buy mortgage guarantee, offered insurance on mortgages between 80 and 95 per cent LTV up to the same house-price threshold.
Other countries’ schemes are more expansive, both in terms of the range of LTVs or property values covered. In Canada, the CMHC offers insurance on mortgages up to 95 per cent LTV for owner-occupiers and 80 per cent LTV for buy-to-let mortgages with a maximum property value of CA$1 million in both cases. Insurance is compulsory above 80 per cent LTV. The NHG in the Netherlands is more like the UK’s MGS in that it is available only to owner-occupiers and on property values below €355,000. It is available at any LTV though never compulsory. In practice, there is little difference in interest rates offered with and without insurance if the LTV is below 65 per cent, so it is only worth purchasing for those with smaller deposits.
Should compulsory insurance apply only above a specific LTV threshold? It is notable that it has not just been 95 per cent LTV mortgages that have disappeared or significantly risen in price at times of increased uncertainty over recent years: in early 2009, there were some 90 per cent LTV products available and the cost spread over 75 per cent LTV has similarly fluctuated over the cycle. This is also true even at 85 per cent (Figure 9).
Figure 9 – The availability and spreads on mortgages above 80 per cent LTV has also varied over the cycle
Source: Bank of England
The arguments in favour of a compulsory mortgage guarantee above 90 per cent LTV would therefore appear to apply above 80 per cent. The cost of the guarantee would of course be lower at lower LTVs. Increasing availability to a wider range of owner-occupiers might also help establish mortgage insurance as a mainstream product rather than as a niche option required only for high-risk borrowers.
Whether there should be a maximum property value or mortgage amount is a closer call. The basis for using a house-price threshold, such as the one currently in place under the MGS, is unclear. The government may not want to be liable for large mortgages taken out by those with higher incomes. On the other hand, if insurance is priced such that the scheme pays for itself then it should not be necessary to limit its coverage, and doing so may create unnecessary distortions in the market. Moreover, if – as we have argued in this paper – such insurance would make the market function more effectively, dampening unjustified swings from pessimism to optimism and back again, it is unclear why insurance should be limited only to less expensive housing.
Nevertheless, there would seem to be little justification for extending state-backed insurance to buy-to-let mortgages. Doing so would weaken, even if it would not eliminate, the redistribution of credit towards FTBs.
How Much of a Loan Should Be Insured?
Offering mortgage insurance could run the risk that banks may choose to lend more recklessly, knowing that any loan that is not repaid by the borrower will be repaid by the insurer. This is known as “moral hazard”, and eliminating it is the final design challenge. To counter it, some countries introduce a “vertical slice” so that lenders cannot fully insure the loan and have to retain some exposure to any losses: they have “skin in the game”. For example, the MGS covers only 95 per cent of losses above 80 per cent LTV. Similarly, in the Netherlands, banks have to bear 10 per cent of losses on an insured mortgage. Keeping some risk on banks’ books ensures that they are less likely to take unnecessary risks.
But this approach is not universal. In Canada and Australia, insurance policies typically cover the entire loan. In other countries, there is a “horizontal slice”; in other words, only the top slice of the mortgage (between 10 per cent and 30 per cent LTV) is covered. The UK’s MGS has this feature, as do private insurance policies in the United States. Of course, in practice it would be unlikely that the lender would face any losses from such an arrangement as this would require a catastrophic reduction in property values for them to be unable to recover the remainder of the loan from the sale of the property. It does, however, limit the maximum potential liability to the insurer.
Figure 10 – Different countries make different choices about how much of the loan to insure
Source: TBI (Note: Lines show the part of a £95,000 loan on a £100,000 property purchase under the mortgage-insurance system in place in each country.)
There are risks both in making insurance too generous and not generous enough. If it is too generous, there is a greater risk of moral hazard which may ultimately lead to losses for the taxpayer or increases in the cost of mortgage insurance, undermining any cost savings to borrowers. But if insurance does not significantly reduce the risk faced by banks, financial regulators will still apply a high-risk weight to high-LTV mortgages and so it will not significantly reduce the cost to borrowers.
Other countries find a way around this dilemma through strict underwriting conditions rather than offering only partial insurance to ensure responsible lending. In Canada, to obtain insurance, borrowers must have a certain credit score and the mortgage repayments must represent less than 39 per cent of gross income with total debt-servicing costs less than 44 per cent.[_] There are also stringent documentation requirements. Similarly, in the Netherlands, general mortgage regulations are relatively strict in terms of maximum debt-servicing cost-to-income ratios and credit checks, although there are no particular conditions for using the NHG.
The UK’s existing “stress test” for mortgage lending forces lenders to ensure that mortgage repayments are affordable for the borrower and would remain affordable even under a “stressed rate”, currently three percentage points above the reversionary rate. Regulations also require documentation of borrowers’ income. It is therefore unclear why a state-backed insurer should decline to insure any mortgage that was issued by a regulated entity.
That is not to say that a reformed MGS should simply insure any mortgage without question: with mortgage insurance available, there would be a risk that banks would relax their stress tests as far as possible to maximise lending. There would still be a role for the insurer to act as a “second set of eyes” on decision-making and to reject loans that they consider too risky. This role might need to be expanded if insurance were compulsory. And it is important to bear in mind that insurance does not completely eliminate risks for lenders: insurers can and do refuse to pay out claims if the lender has not performed adequate checks on the borrower or made an independent property valuation.
Retaining the “horizontal slice” would also be sensible, but if insurance were compulsory on all lending above 80 per cent LTV, it would make sense to extend insurance down to 75 per cent so that some meaningful cover would be given to loans at the lower end of this range. This effectively turns a 95 per cent LTV mortgage into an 75 per cent LTV loan with some government-backed debt, which Bank of England research suggests should reduce the risk weighting by around 20 percentage points, as detailed earlier in this paper.[_] It would take a very large fall in house prices for it to be impossible to recoup a 75 per cent LTV mortgage from the sale of the property.
One argument in favour of insuring the whole loan is that under current regulations it would be simpler to claim capital relief with such an arrangement. This is because a guarantee would no longer be classified as a securitisation. If it proves impossible to simplify the Credit Risk Mitigation framework so that banks are able to claim capital relief on their insured mortgages, this would be an alternative way to achieve the same objective. It could also potentially further reduce the cost to borrowers: 75 per cent LTV mortgages still attract a risk weight of 20 per cent compared with none for government debt, implying a potential saving of 20 basis points.[_]
But insuring the entire loan would also increase the cost of the guarantee (since the government would have to set aside more capital to cover the increased risk) as well as the government’s contingent liabilities, even if it is unlikely that insurance payouts would significantly increase in practice. Reforming the Credit Risk Mitigation framework would, on balance, seem to be a much more straightforward way of achieving this objective.
Insufficient savings for a deposit remain a significant constraint preventing prospective FTBs from making their first property purchase. Action to raise and stabilise the availability of high-LTV lending across the economic cycle is therefore key to sustainably increasing home-ownership levels.
The temporary MGSs implemented by the Treasury over the past decade may have had, at best, a marginal effect in achieving this goal, and would need to be radically reformed to achieve better outcomes.
First, making the MGS a permanent intervention should in time lead to banks and regulators treating insured mortgages as less risky. This would in turn increase the willingness of lenders to lend at high LTVs and bring spreads down considerably as lenders compete for this business. This would significantly enhance the scheme’s ability to increase lending at high LTVs and therefore home ownership as well as militating against excessively cheap and risky lending late in the credit cycle.
Second, participation in the scheme should be compulsory for eligible mortgages. This will encourage the attitudinal change required for our proposals to have the desired effect. It could also make the scheme cheaper by addressing adverse selection problems associated with a voluntary scheme.
Third, eligibility should be extended to all mortgages above 80 per cent LTV. There is evidence that the availability and price of all high-LTV mortgages, relative to lower-LTV ones, fluctuated sharply in the wake of the financial crisis, suggesting a useful role for insurance below the very highest LTV levels.
Fourth, by simplifying the Credit Risk Mitigation framework so that the appropriate capital relief is given to participating banks; the MGS could reduce the cost of capital to banks to issue these mortgages in a way that does not currently happen. It is likely that applying the appropriate level of capital relief would lower the cost of mortgages with an LTV of above 90 per cent by 20 basis points.
Figure 11 – Spreads have been higher than under our proposal for most of the past decade
Source: Bank of England, TBI calculations
Together, these changes should ensure that high-LTV mortgages are available across the credit cycle and reduce the spreads relative to lower-LTV loans, while also tempering excessively risky lending during boom times. Observed spreads have been over 100 basis points for most of the period since the financial crisis (Figure 11). By contrast, a pessimistic estimate of the additional cost under our proposal is just 40 basis points (including the cost of insurance as well as an estimate of the additional administrative costs associated with high-LTV lending). This would increase the number of prospective FTBs able to enter home ownership and reverse the declines in home ownership seen since 2008.
A potential criticism of using mortgage insurance to boost home ownership is that it will lower borrowing costs for everyone and therefore simply boost leverage and house prices, leaving FTBs no better off. This is misconceived for a few reasons. First, outside financial bubbles, house prices are determined by economic fundamentals such as interest rates and rent levels. Lending follows these market signals rather than affecting prices without reference to any underlying economic value. Second, the mortgage-insurance premium should have a calming effect on the market at key moments, preventing risky lending from becoming excessive, or excessively cheap, as happened in the run up to 2008.
Overwhelmingly, the effect of mortgage insurance would be to lower the cost of high-LTV loans compared with lower-LTV mortgages, because this is where credit risk is most acute. Consequently, the policy would redistribute credit towards FTBs, rather than increasing the volume of lending. This tilting of lending towards FTBs and away from other borrowers could be assisted through complementary fiscal measures. Further steps could be taken to reduce incentives for buy-to-let, second homes and overseas buyers if concerns about any inflationary effect from the policy remain.
A more comprehensive system of mortgage insurance would tackle the most significant barrier to home ownership: the high deposit requirement. But, recalling our analysis, it is not just the size of the required deposit that makes it challenging for non-homeowners to get on to the property ladder.
More than three-quarters of non-homeowners aged between 25 and 44 do not have sufficient earnings to qualify for a mortgage under current rules, which oblige them to show that they would not only be able to afford the initial monthly mortgage repayments but still be able to afford repayments if interest rates increased. This “stress test” is imposed despite the fact that many renters are paying a significantly higher rental yield than prevailing mortgage interest rates.[_] This leads to the curious situation where an FTB looking to buy the house that they currently rent may be deemed unable to afford the mortgage repayments, despite these repayments being significantly lower than the rent they are currently paying.
Sensibly, this “stress test” does not apply to mortgages with an initial fixed rate of more than five years: borrowers need only show that they can afford the actual payments if there is no prospect of them changing in the short to medium term. If mortgages with interest rates that were fixed for more than five years were more widely available, it would be possible for FTBs to safely borrow more.
Greater availability of long-term fixed-rate mortgages may be desirable for other reasons too. They would be a popular choice for more risk-averse consumers even if they did not want to borrow more. Furthermore, the Miles Review published in 2004 argued that they would improve macroeconomic stability as house prices would become less sensitive to interest rates, and the effect of interest-rate rises on the economy would occur less through the impact of house prices and consumers’ spending power and more through the impact on business lending and the exchange rate.[_]
However, much has changed since 2004. Not only have interest rates been close to zero for more than a decade, there has already been a remarkably rapid shift from variable to fixed rates, albeit with initial terms of no more than five years in almost all cases. In 2003, only a quarter of the mortgage stock was fixed rate, but that has increased to more than 80 per cent according to the latest available data from Q3 2021. Nearly 95 per cent of all mortgages issued in Q3 2021 had an initial fixed-rate period.[_] Gone are the days when a change in the Bank of England’s base rate would have an immediate impact on the finances of most working-age households.
Even so, few products offering fixed rates for more than a five-year period are available. This is largely because banks’ funding comes from deposits, on which they pay variable interest rates. Covered-bond markets – which finance long-term fixed-rate mortgages in other advanced economies, passing interest-rate risk to investors – are poorly developed in the UK: only £8 billion was raised during 2020.[_]
It is often thought that demand for long-term fixed rates is low because borrowers prefer products with a lower initial monthly payment – typically, variable-rate mortgages with an initial discount or two-year fixed-rate mortgages offer the lowest initial monthly payments. But, as Miles pointed out in 2004, it is important to bear in mind that borrowers could achieve the objective of having mortgage payments that rise in nominal terms over time without taking on substantial interest-rate risk. Long-term fixed-rate mortgages that began as interest-only, at least in part, with increasing monthly repayments over time would also achieve this objective while still avoiding negative amortisation in the early years.
The UK is unusual in not having long-term fixed-rate mortgages as a mainstream option (Figure 12). Fixed rates of more than five years are commonplace in almost all other advanced economies, and rates that are fixed for the full length of the mortgage are the norm in many, most notably the United States (data from the Urban Institute shows that nearly 90 per cent of mortgages initiated in November 2021 had a fixed rate for the full term).[_]
Figure 12 – The UK is unusual in not offering the option of long-term fixed-rate mortgages
Source: European Mortgage Federation data
Mortgage markets where long-term fixed rates are the norm look very different to the UK’s. The mortgage market in Denmark, for example, is dominated by mortgage banks that operate on the “match funding” principle where every mortgage is sold as part of a bond with equivalent characteristics. And in the United States, the funding of long-term fixed-rate mortgages is facilitated by the government-sponsored enterprises that purchase them from originators for securitisation. In all these cases, banks simply originate the loans and then sell the debt on to avoid the challenge posed by maturity mismatch – very long-term assets (mortgages) funded by very short-term liabilities (deposits). Should the UK opt for a similar model?
There are some prospective FTBs for whom long-term fixed-rate mortgages would be an attractive proposition. Those with steady employment whose earnings are insufficient to qualify them for a mortgage to purchase a property in their local area would likely welcome an opportunity to borrow more on a long-term fixed-rate mortgage, even if they incurred a higher interest rate. People who fall into this group are, with some justification, confused and frustrated that the stress test means that they are told by lenders they are unable to afford a mortgage repayment less than they are paying in rent. Reforms to make these long-term, or at least longer-term, fixed-rate mortgages more widely available would therefore be a positive step.
However, it is far from clear that wholesale change in the structure of the UK mortgage market and the role of banks, to move to a system where long-term fixed-rate mortgages were the norm, would be desirable. The introduction of institutions like those in Denmark and the United States would significantly disrupt banks’ business models, leaving them seeking an alternative use for deposit funds. At the same time, as our analysis implies, even in the best-case scenario such a reform would probably only have a limited payoff in terms of increasing home ownership. Furthermore, the existing system works well for many borrowers who prefer to fix their mortgage-interest rate for five years or less at a lower rate, rather than paying a higher rate to insure against interest-rate risk more fully. Moving wholesale to a different mortgage-market model would not only be very disruptive but also not be in the interests of many consumers.
In this section, we argue that we should instead aim for a more diverse market, offering products to suit the different needs of FTBs. Achieving this requires a rethink of several factors that have increasingly shaped the market to reinforce the dominant model of short-term lending. Regulations, consumer attitudes, data availability and the business models of brokers have all developed around the dominant lending model, which in turn inadvertently squeezes out the scope for a mainstream market in long-term fixed-rate mortgages to develop.
There are both demand- and supply-side barriers to the greater availability of long-term fixed-rate mortgages in the UK. These could be addressed by making three changes to break out of the one-size-fits-all approach and stimulate this segment of the market. First, the government should finesse the blunt affordability regulations that put longer-term products at an unjustified disadvantage. Second, there is a need to address perceptions of mis-selling that are often perverse. Finally, we propose ways to remove the "prepayment risk" barriers (that is, the risk that borrowers will pay back the mortgage early rather than holding it for the full term) that make it hard for life insurers to become more involved in funding mortgages.
Affordability Regulations and Long-Term Fixed-Rate Mortgages
One of the clearest examples of the way in which rules have been designed around the short-term lending model to the detriment of long-term fixed-rate availability is in affordability regulations. The Financial Policy Committee limits the proportion of loans that lenders can issue that represent more than 4.5 times the borrower’s income to no more than 15 per cent of their overall loan book. This constraint will become more binding if the Financial Policy Committee of the Bank of England withdraws its recommendation that affordability calculations use an interest rate of the reversion rate plus 3 per cent as the “stress rate”.[_] The minimum stress rate will revert to the reversion rate plus 1 per cent, in line with the FCA’s guidelines.
From the perspective of short-term fixed-rate products, the logic behind this figure makes some sense: it corresponds to a household’s debt-servicing ratio (DSR) – that is, the share of income taken up by mortgage payments, rising to between 35 per cent and 40 per cent if the borrower’s future interest rate were to jump to 7 per cent, which is the level used for the stress test.[_],[_] DSRs above 40 per cent are associated with greater risk of default.
But the risk of a higher interest rate is not relevant for long-term fixed-rate mortgages by definition: there is no risk that the mortgage will become unaffordable to the borrower because of increasing monthly repayments. There is consequently no rationale for the 4.5-times-income cap when the mortgage rate is fixed for the whole term. All regulators should care about is whether the mortgage is affordable for the borrower at the prevailing interest rate.
Excluding long-term fixed-rate mortgages from the LTI flow cap would therefore be a sensible step. Alternatively, the cap could be recast in more flexible terms as a DSR cap of, say, 40 per cent. This would allow the “stressed rate” to be used in the assessment for those with variable or short-term fixed rates, and the actual rate to be applied for long-term fixed-rate mortgages.
Such an approach would be in line with rules in other advanced economies. In the United States, “qualified mortgages” under the consumer-protection requirements of the Dodd-Frank Act must have a DSR of less than 43 per cent,[_] and in Canada mortgage repayments must be less than 39 per cent of gross income for the mortgage to be insured.
Mis-Selling Risks and the Brokerage Market
Other regulatory reforms could focus on the advice consumers are given when taking out a mortgage. A commonly raised point in the interviews we conducted for this report was that, culturally, long-term fixed-rate products are somewhat alien to UK borrowers. But it is also clear that there are some softer institutional constraints that prevent consumers being offered them.
Critically, lenders and mortgage brokers worry that they will be accused of mis-selling if they advise consumers to take out a long-term fixed-rate mortgage. There are two reasons for this. The first concern is that, if interest rates subsequently fall and borrowers are faced with the unpalatable choice of either paying a large early repayment charge or continuing to pay a higher interest rate than they could obtain if they refinanced their mortgage, providers may be accused of mis-selling. With long-term interest rates still close to historic lows, this scenario may seem unlikely, but regulation needs to be designed to take possible future scenarios into account. Second, it has historically always been cheaper for borrowers to choose a short-term fixed-rate mortgage at a high LTV and then refinance once the LTV ratio is lower.
But these are not the only relevant considerations. For some prospective FTBs, affordability constraints may mean that they are only able to borrow as much as they require with a long-term fixed-rate mortgage. Taking on a long-term fixed-rate loan is not necessarily bad advice for someone whose only route into home ownership is by borrowing more relative to their income. Moreover, it would still be possible to refinance when the LTV fell, though an early repayment charge could be payable.[_] Further, even if interest rates subsequently fall, this does not mean that having taken out insurance against them rising, in the form of a long-term fixed-rate mortgage, was a bad decision. The value of home insurance is not seen as being conditional on whether one’s home subsequently burned down.
Therefore, to encourage development of this market, the FCA and the Financial Ombudsman Service should clarify that they are neutral between short- and long-term fixed-rate mortgages. They should set out that they would not regard an advisor recommending a long-term fixed-rate mortgage as mis-selling, even if interest rates subsequently fell, as long as this possibility had been made clear to the borrower.
A related institutional constraint on the emergence of long-term fixed-rate products is the functioning of the brokerage market. Here too, the dominance of the short-term lending model fosters a brokerage market dependent on repeat custom from short-term borrowers, returning for advice each time they need to refinance. This business model reinforces that dominance.
One solution would be to extend the rules that advisors face on such “churning” in the investment market to the mortgage market. In the investment market, advisors are warned that recommending switching products at a frequency that is not in the client’s best interest would be considered mis-selling. Brokers’ advice could also be encouraged to take account of the borrower’s broader context, for example whether a consumer should rent or buy a property. This would enable them to advise a prospective FTB to purchase a property using a long-term fixed-rate mortgage if the only alternative were renting, given the advantages of home ownership in the longer run.
Freeing up Sources of Long-Term Finance
The third area for action relates to freeing up funding from longer-term sources. As already discussed, deposits are the main source of mortgage funding for the major lenders. It is therefore difficult for them to offer long-term fixed-rate mortgages because these would worsen the maturity mismatch between their short-term liabilities and long-term assets. Increasing the supply of long-term fixed-rate mortgages will therefore require other actors with longer-term liabilities to enter the mortgage market.
In those countries where long-term fixed-rate mortgages are more commonplace, life insurers are a major buyer of the bonds that back these products. UK life insurers might be interested in purchasing such securities too. Indeed, they could enable significantly cheaper long-term fixed-rate mortgage products to come on to the market.
However, there is a snag. The matching adjustment under Solvency II regulations allows life insurers’ capital requirements to be calculated in an attractive way if the duration of the assets matches those of the liabilities. But the risk that borrowers will repay their mortgage earlier than planned would upset the otherwise close maturity of assets and liabilities. This prepayment risk then makes the life insurer ineligible for the lower capital requirements under the matching adjustment.
How could this be remedied? The easiest solution would be to separate out mortgage bonds into one component that provide a predictable income stream and a second “companion bond” that bears the prepayment risk. The first of these would be eligible for the matching adjustment. The insurance company Rothesay has entered this market by financing Kensington Mortgages’ long-term fixed-rate product, so it does not seem impossible for life insurers to provide this financing.
But for the market to flourish at scale, much would depend on the scale of prepayment risk: if it were cheap and easy for borrowers to refinance long-term fixed-rate mortgages when interest rates fell, most would do so, making prepayment a major headache for investors looking for long-term assets.
There are broadly three options to solve this problem. First, prepayment risk could be eliminated by setting early repayment charges on long-term fixed-rate mortgages on a “make whole” or mark-to-market basis so that the holder of the debt was fully compensated for the loss of income resulting from refinancing.[_] That would significantly lower the interest rate on such mortgages but leave borrowers vulnerable to charges for unforeseen life events like divorce. If interest rates were to fall significantly, this could amount to tens of thousands of pounds on a typical mortgage.
Alternatively, life insurers could purchase insurance against prepayment risk. Purchasers of mortgage bonds could purchase call options on non-callable bonds with similar characteristics to the mortgage bonds they are purchasing. The government could write call options on long-dated government debt to help meet this demand.
Finally, one could leave the prepayment risk with the holders of the debt but change Solvency II regulations so that mortgage bonds could still qualify for the matching adjustment. The insurance company Willis Towers Watson recently suggested reforms in their response to a consultation on the future of the Solvency II rules.[_] They proposed that instead of an eligibility “cliff edge” that prevents assets with prepayment risk being eligible for the matching adjustment, it should be possible for these assets to qualify if capital is provided to insure against the prepayment risk. This would allow life insurers to hold mortgage bonds within a “match adjustment portfolio”, albeit at some additional cost.
While the main barrier to life insurers funding long-term fixed-rate mortgages relates to the prepayment risk itself, that risk is magnified by uncertainty about its scale. Since there is no track record with long-term fixed-rate products in the UK, there are also limited data on the likely prepayment rates for these loans. This makes it difficult for regulators to assess the extent of prepayment risk, creating yet another barrier to the competitive pricing of such products. Here too, then, the dominance of the short-term lending model makes diversification of mortgage products more difficult.
However, these bonds, and the long-term fixed-rate mortgages they fund, have a long track record in other countries. Data from overseas could and should be used to assess the likely extent of early repayments, as can data emerging from the new long-term fixed-rate products available in the UK from providers like Kensington and Habito.
Interaction with Mortgage-Insurance Proposal
Earlier in this report we argued that mortgage insurance could significantly raise home ownership by better managing the risk associated with FTBs. Should such a mortgage guarantee treat long-term fixed-rate mortgages differently? Since long-term fixed-rate mortgages do not carry any interest-rate risk, they should in principle be less risky and cheaper to insure.
However, it is not clear that they are significantly less risky than, say, a five-year fixed-rate mortgage as interest-rate risk would not materialise until five years into the mortgage term. At that point the borrower will have significantly more equity in the property, increasing the likelihood that the lender would be able to recoup the amount owed from the sale of the property if the borrower did default.
As a concrete example, the MGS covers the mortgage for only the first seven years of the loan – by this point the LTV ratio can be expected to be around 80 per cent. It is therefore unclear that a long-term fixed-rate mortgage is significantly less risky than one with a five-year fixed rate for the duration of the guarantee. To the extent that there is a case for a lower fee for mortgages with an initial term of more than five years, this should also be applied to those with an initial term of exactly five years.
Long-term fixed-rate mortgages are uncommon in the UK, unlike in other countries. There are good reasons to encourage greater availability of these products in the mortgage market. Long-term fixed-rate mortgages would enable FTBs to borrow more, allowing more people to access home ownership. They effectively offer consumers an insurance product to protect them against the risk of rising interest rates. By strengthening households’ financial resilience, they may also have macroeconomic stability benefits.
But long-term fixed-rate mortgages will certainly not be the best choice for everyone and there are benefits to having a diverse market of different products to suit the varied needs of consumers. Consequently, moving wholesale to a different mortgage-market model would not be worth the significant disruption and risk this would cause.
Where policymakers should act is to lean against the regulations, guidance, business models and information barriers that have built up around the dominant short-term lending model and which unnecessarily frustrate the emergence of long-term fixed-rate products. Taking action around affordability rules, mis-selling guidance and approaches to managing prepayment risk could unlock greater diversity in the mortgage market and boost home ownership as a result.
Many factors, not least house prices, affect home-ownership levels. But the cost and availability of mortgage finance to FTBs is its single most important determinant. Wild swings in lenders’ risk appetites – from excessively risky lending before the financial crisis to an FTB-lending drought for many years afterwards – combined with blunt changes to financial regulations since the financial crisis, have had the unfortunate side effect of reducing home ownership.
In this paper, we have argued that the key to reversing these trends and insulating the prospect of home ownership from the vicissitudes of the credit cycle is better management of the risks of lending to FTBs.
We propose wide-ranging changes to mortgage-market institutions to introduce comprehensive mortgage insurance and foster the emergence of long-term fixed-rate products. With these changes to the operation of the mortgage market, policymakers could expand and stabilise the availability of mortgage finance for FTBs, and potentially see home-ownership rates return to the levels seen before the financial crisis – but without returning to the same level of financial risk.
We are grateful to Lloyds Banking Group for their support towards the research presented in this report, as part of Lloyds Banking Group’s support for broadening access to housing.
The Wealth and Assets Survey was collected by the Office for National Statistics and distributed by the UK Data Service (UKDS). Crown Copyright Material is reproduced with the permission of the Controller of HMSO and The Queen’s Printer for Scotland.
The authors would also like to thank Franklin Allen, Andrew Asaam, Andy Gray, Andy Mason, Rob Thomas, Peter Williams and officials at HM Treasury and the Department for Levelling Up, Housing & Communities for their input during the course of the research. All errors remain those of the authors.
Lead Image: Getty Images