Key takeaways
- Legal & General's own published table puts £50,000 at 7% at £98,358 after 10 years and £193,484 after 20. Run the same arithmetic one decade further and it's £380,613.
- At 7% a rolled-up balance doubles every 10.2 years. Nationwide's UK house price series grew 2.66% a year over the 20 years to the second quarter of 2026.
- What compounds is the ratio. At 7% against 2.66% house price growth, the debt's share of the house multiplies 2.29 times over 20 years and 3.47 times over 30.
- The Equity Release Council put the average new lump sum at £113,779 in the second quarter of 2026. At 7% that reaches £440,289 after 20 years and £866,115 after 30.
- The PRA has used a minimum deferment rate of 4.5% a year in its Effective Value Test since March 2026, alongside a house price volatility parameter of 13%.
Equity release compound interest: the number you came for
You want to know what the debt comes to. Legal & General publishes the answer for its own lifetime mortgages, and it isn't buried. On an initial loan of £50,000 at an example rate of 7%, the lender's table shows £98,358 owed after 10 years and £193,484 after 20. That's their arithmetic, printed on their rates page, not a model built by someone else.
The table stops at 20 years. Extend it one more decade on exactly the same basis and the balance is £380,613. At 6% the 30-year figure is £287,175. At 8% it's £503,133.
None of this is hidden or unusual. It's simply what a fixed rate does when nobody pays it down. A lifetime mortgage is a loan with no contractual monthly payment: interest is charged on the loan plus the interest already added, and the whole thing is settled when the last borrower dies or moves permanently into care. The Equity Release Council, the sector's trade body, records that lifetime mortgages account for more than 99% of the market.
The size of the number isn't the interesting part. The interesting part is that the number is racing something. It's secured on a house, and the house has its own growth rate. Over the last two decades the house has been slower than every rate in that table.
The lender's table, and what a third decade does to it
Here's the full shape, with Legal & General's published figures and the 30-year extension of them. The loan is £50,000 and no repayments are made.
| Rate | 10 years | 15 years | 20 years | 30 years |
|---|---|---|---|---|
| 6% | £89,542 | £119,828 | £160,357 | £287,175 |
| 7% | £98,358 | £137,952 | £193,484 | £380,613 |
| 8% | £107,946 | £158,608 | £233,048 | £503,133 |
The 10, 15 and 20-year columns are Legal & General's. The 30-year column applies the same annual compounding for ten more years, because their table ends where a lot of borrowing lives don't.
Read across any row and you'll see the shape. Between year 10 and year 20 at 7%, the balance grows by £95,126. Between year 20 and year 30 it grows by £187,129. The second decade of roll-up interest costs roughly twice what the first one did, because the interest is being charged on a much larger number.
The clean way to hold this is doubling time. At 6% the balance doubles every 11.9 years, at 7% every 10.2 years, and at 8% every 9.0 years. A single percentage point on the rate moves the doubling point by about a year and a half, which compounds into a very large gap by year 30.
What the house did over the same windows
The debt is settled out of a sale price. So the series that matters on the other side of the ledger is house prices, not housing total returns. Rent doesn't pay down a lifetime mortgage, and an owner-occupier receives their rent as shelter rather than cash. If you want that distinction argued properly, it's the subject of our piece on your house as an investment.
Nationwide's UK series, which runs back to 1952, put the average UK house at £278,784 in the second quarter of 2026. In the second quarter of 2016 it was £204,238. In the second quarter of 2006, £165,035. In the second quarter of 1996, £53,032. Those are compound growth rates of 3.16%, 2.66% and 5.69% a year.
Here's the same exercise across three price series, all measured from June 1996, June 2006 and June 2016 to June 2026.
| Annualised price growth to mid-2026 | 10 years | 20 years | 30 years |
|---|---|---|---|
| UK house prices (Nationwide) | 3.16% | 2.66% | 5.69% |
| UK share prices, excluding dividends (OECD) | 5.36% | 3.08% | 3.48% |
| US house prices (Case-Shiller) | 6.35% | 3.05% | 4.78% |
Two things fall out of that table. The first is that the only cell above 6% is US house prices over the last decade, and a UK loan isn't secured on a US house. The second is that the UK's strong-looking 30-year number is an artefact of where the window starts. The second quarter of 1996 sat just after the early-1990s crash. Start at the previous peak instead, the third quarter of 1989, and UK house price growth to mid-2026 is 4.14% a year. Start at the 2007 peak and it's 2.24%.
Every one of those figures is below 6%, which is the cheapest rate in Legal & General's own example table.
The ratio is what actually compounds
Most explanations of equity release compound interest stop at the balance. That's the wrong object, because the balance on its own tells you nothing without the house beside it. What matters is the debt as a share of the property, and that ratio has its own growth rate: roughly the loan rate minus the house price growth rate.
At 7% against 2.66% growth, the ratio grows about 4.2% a year. So it multiplies by 1.51 over 10 years, 2.29 over 20 and 3.47 over 30. That's true whatever you borrowed. A loan starting at 20% of the house is at 46% by year 20 and 69% by year 30. One starting at 30% reaches the whole value of the house in year 30.
Put the two published averages side by side and you get the scale. The Equity Release Council's average new lump sum in the second quarter of 2026 was £113,779. Nationwide's average UK house in the same quarter was £278,784. At 7%, that loan is £223,821 after 10 years, £440,289 after 20 and £866,115 after 30. Against a house growing at the 20-year rate, the debt passes the value of the property in year 22.
Treat 41% as an illustration of the arithmetic rather than a typical loan. The two averages describe different populations: the Council's figure covers equity release borrowers, and Nationwide's covers every house it lends on. Dividing one by the other is a ratio of two unrelated samples, not a loan-to-value anyone was offered. The ratio multipliers above don't depend on any of it, which is why they're the more useful number.
Choose a faster house: at 7% against the 30-year Nationwide rate of 5.69%, the ratio still multiplies by 1.45 over 30 years. The gap narrows a great deal. It doesn't close, and it never runs the other way at any rate in the lender's table.
Why the rate is 6% to 8% when Bank Rate is 3.75%
Bank Rate was 3.75% in mid-September 2026. The example rates on that lender page are between 6% and 8%. The spread isn't only funding cost and margin. A large part of it is an option the borrower is being sold.
Equity Release Council product standards require a no negative equity guarantee: "the borrower or estate will never owe more than the property is worth, after deduction of reasonable sales costs". They also require the rate to be fixed, or capped and fixed for the life of the loan, and they give the customer the right to live in the property for the rest of their life. Those are real protections, and somebody has to price them.
The regulator has put a number on the pricing. In supervisory statement SS3/17, the PRA describes the guarantee as "an important source of risk", and warns that "assuming future house price growth in excess of risk-free rates should not lead to a lower valuation of the NNEG". Its Effective Value Test has run on a minimum deferment rate of 4.5% a year since March 2026, and a house price volatility parameter of 13% a year, unchanged at that review.
A deferment rate of 4.5% means the regulator requires insurers to value the right to own a house at a future date at a discount to its price today, compounding at 4.5% a year. That's a supervisory parameter, not a forecast. But it's the closest thing to an official statement that you can't fund a long-dated guarantee out of assumed house price growth, and it's set to be conservative for exactly that reason.
The strongest case against reading this as a warning
Equity release compound interest is the weakest part of the case against these products, and the counter-argument deserves stating properly.
Start with the guarantee. Because of the no negative equity guarantee, the year-22 crossover above is largely notional. Past that point the estate owes the sale proceeds and not a penny more. The lender absorbs the shortfall. What the arithmetic actually destroys is the inheritance, not the borrower's solvency, and those are different problems with different weights attached to them. A smaller estate isn't only a smaller gift either, since the debt reduces what's measured against the UK inheritance tax thresholds.
Then the term. Thirty years of roll-up assumes the loan runs for 30 years, and from the Council's minimum age of 55 that carries the balance to age 85. Every year a borrower waits shortens the horizon, and shortens the table with it. The Council's Q2 2026 data shows a market designed around this: returning drawdown customers numbered 6,978 against 5,307 new customers, with average initial drawdown of £63,642 against £113,779 for a lump sum. Drawing less, later, starts the clock later on most of the balance.
Then the repayments. Council standards require that customers can make repayments without penalty, subject to the provider's lending criteria. A lifetime mortgage that's serviced isn't a compounding instrument at all; it's an interest-only loan with no term. The 7% column describes the case where nothing is paid, which is the design, not an obligation.
And finally the comparison. The alternative to a lifetime mortgage isn't usually nothing. It's a sale, and selling has its own arithmetic, which we set out in downsizing in retirement. A person who can't or won't move has a narrower menu than the compounding table alone suggests.
What these numbers can't tell you
The biggest limit on any model of equity release compound interest is the rate itself, and the rate in the table isn't a market average. Legal & General labels 6%, 7% and 8% as examples, and says the rate depends on individual circumstances. Your own arithmetic runs on your own fixed rate, and that rate is the single largest input here.
The house price series are national indices. Nationwide's is mix adjusted, which it describes as tracking "a representative house over time", and it is built from the society's own lending for owner-occupier houses. The company notes that the methodology has changed several times since 1952, most recently in 2024. One house in one street doesn't track it, and the dispersion around a national index is wide.
The historical growth rates are not forecasts. They're a record of three arbitrary windows, and the section above shows how much the answer moves when you shift the starting quarter by a few years. A 30-year figure from a sample that contains one boom and two crashes is a description of that sample and nothing more.
The debt side, by contrast, is not a projection at all. A fixed rate compounding on a known balance is arithmetic. That asymmetry is the whole point: one side of this comparison is certain and the other isn't, and the certain side is the one that grows fastest.
Finally, this piece doesn't price the costs around the loan. Advice and the mandatory independent legal advice both sit outside it, and early repayment charges can apply where a loan is settled other than on death or a permanent move into care.
What would change the conclusion
Three things, in order of how much they'd move it.
A sustained fall in lifetime mortgage rates would do most of the work. The gap between a 6% loan and 5.69% house price growth is almost nothing over 30 years: the ratio multiplies by 1.09, and the inheritance survives more or less intact. If rates on these products settled nearer Bank Rate plus two points than the 7% and 8% examples, the compounding argument would lose most of its force.
A return to 1996-to-2026 house price growth would do the rest. UK prices have run at 2.66% a year over the last 20 years and 2.24% from the 2007 peak. A repeat of the 5.69% thirty-year rate would change every ratio on this page. There's no mechanism in the data that makes that more or less likely, which is precisely why it belongs here as a condition rather than a prediction.
And the PRA's deferment rate is worth watching for its own sake. It sat at 0.5% from the September 2019 review through March 2022, then rose in steps, reaching 4.5% at the September 2025 review and holding there in March 2026. That parameter is a regulatory judgement about how much a house in 30 years' time is worth today. When it moves, the cost of writing the no negative equity guarantee moves with it, and lifetime mortgage rates follow. It's the number that sits upstream of everything else in this article.