From free cars and 125% mortgages to self-cert lending and shared appreciation schemes, the UK mortgage market of the early 2000s looked very different from today. This article explores how lending practices changed after the financial crisis and why tighter regulation created a safer, but less flexible, market for modern borrowers.

In 2003, you could take out a mortgage with West Bromwich Building Society and drive away in a new Škoda Fabia. Not a voucher. Not cashback. An actual car, the sensible Czech supermini of the moment, VW-engineered, unpretentious, and sitting on the driveway by the time you got your keys. It happened. It was legal. And it tells you almost everything you need to know about how dramatically the mortgage market has changed in twenty years.

The early 2000s were a genuinely strange time in UK lending. Regulation was light, competition between building societies and the banks aggressively breaking into their territory was fierce, and lenders were willing to try things that would get a compliance team fired today. Cars were the most theatrical example, but they weren’t unusual in spirit. Holiday packages, typically a two-week break in Spain or the Canaries, were a regular completion incentive from regional lenders through the same period. John Lewis and M&S vouchers worth £500 to £2,000 were handed out routinely during the remortgage boom of 2002 to 2007. Some lenders partnered with white goods suppliers and kitted out entire kitchens at new-build completions. Carpets, curtains, integrated appliances. The thinking was straightforward: you’ve just borrowed £150,000, you’re moving into an empty house, here’s something to help.

It sounds generous. The economics were less so. Every physical incentive was funded somewhere in the rate or the arrangement fee. Nobody was giving away a Fabia out of goodwill. But the regulatory environment didn’t require disclosure in any meaningful sense, and most borrowers were focused on the monthly payment rather than the total cost of the arrangement. Lenders understood that, and priced accordingly.

The structural products from the same era were stranger still. Northern Rock’s Together mortgage, a 95% residential loan combined with a 30% unsecured personal loan, enabled borrowers to secure 125% of the property’s value in a single application. It was marketed as flexible, convenient, and designed for people who needed to cover moving costs and furnishings alongside the purchase. It was also, in retrospect, a reasonable preview of what happens when credit expansion runs well ahead of any serious assessment of risk. Northern Rock was nationalised in 2008. The Together mortgage has not been reintroduced.

Cashback mortgages were a step more mainstream but structurally aggressive in the same way. Lenders would advance a lump sum, typically 3 to 5% of the loan value, on completion, in exchange for a rate above the headline market rate and an early repayment charge that could run for up to 5 years. The cashback was real money and immediately useful. The total cost of borrowing over the term was often significantly higher than that of a standard product. The tie-in was the mechanism that made the whole thing work for the lender. These were formally banned for UK lenders in April 2014, though they technically remain available from some European lenders operating in the UK market.

Self-certification mortgages occupied a different category entirely. Available to both employed and self-employed borrowers, they required no income documentation whatsoever. You stated a figure on the application form, and the lender took your word for it. The informal industry name, liar loans, was not a retrospective judgement applied with the benefit of hindsight after the crisis. People working in the market used that term at the time, with a degree of dark amusement that, in retrospect, suggests how clearly the problem was understood. By 2005, self-cert mortgages were mainstream, used by everyone from genuine freelancers with variable income to borrowers who simply wanted a larger loan than their real earnings would support. The FSA banned them in 2009 after the financial crisis made it impossible to continue ignoring their consequences.

The 100% mortgage, borrowing the full purchase price with no deposit, was also standard practice throughout this period, offered by a range of high-street lenders, including Halifax and Nationwide. Some lenders went further, combining a 95% mortgage with an unsecured top-up to cover the deposit requirement, meaning borrowers effectively acquired property with no equity and no buffer. In a rising market, this looked fine on paper. In 2008, it stopped looking fine very quickly.

There were also products that were less obviously reckless but caused serious, slower harm. Shared Appreciation Mortgages, offered by Barclays and Bank of Scotland from 1997, gave borrowers below-market interest rates in exchange for a share of future property appreciation, sometimes 25%, in some cases as much as 75% of the upside. The logic for borrowers was simple: a lower rate today in exchange for giving the bank a slice of any gain when you sell. For someone who signed in 1997 on a London property and then watched prices double in eight years, the maths became punishing. They couldn’t remortgage economically because the arrangement was transferred with the debt. They couldn’t sell without a significant portion going to the bank. They were trapped in properties they nominally owned but couldn’t extract value from, sometimes for decades. Multiple legal challenges were brought over the following twenty years. The vast majority were unsuccessful. The contracts were valid. They were simply profoundly misunderstood at the point of sale, and some of the disclosure documentation did not help borrowers understand what they were surrendering.

Endowment mortgages operated on a similar logic of deferred damage, just on a longer timeline. Interest-only loans are attached to investment policies that were supposed to grow over 25 years and clear the capital balance at the end of the term. The policies were sold alongside the mortgages, often by the same adviser earning commission on both, with projected returns that assumed sustained market performance at rates that looked reasonable in the 1980s and increasingly optimistic thereafter. When the projections failed to materialise, and for a substantial proportion of endowment holders, they didn’t come close, borrowers arrived at the end of their mortgage terms with the capital still outstanding and policies worth a fraction of what they had been told to expect. Some received shortfall letters years before maturity, giving them time to adjust. Many didn’t, or didn’t understand what the letters meant. The FCA spent years processing mis-selling complaints. Compensation paid out ran into billions of pounds.

The interest-only mortgage without a credible repayment vehicle was a related problem, though quieter. Throughout the 1990s and 2000s, interest-only was routinely offered without meaningful scrutiny of how the capital would be repaid. Borrowers were told to rely on property price growth, which in rising markets felt like a reasonable answer. Regulators eventually required lenders to verify repayment strategies, and the MMR effectively killed speculative interest-only for residential owner-occupiers. The product itself was sound. The way it was sold was not.

Go back further, and the creativity was of a different kind. In 1965, a brewery in England raised capital for expansion by offering investors a mortgage on the building, with a lifetime supply of beer as the return on their investment. The brewery presumably still had to repay the principal at some point. The beer was incidental to the economics, if not to the investors. It is an outlier by some distance from the mainstream UK mortgage market, but it captures something real about what secured lending looks like when regulatory frameworks are either absent or so permissive as to be effectively absent.

What ended most modern versions of this was the 2008 financial crisis and the Mortgage Market Review that followed. The MMR, which came into full force in April 2014, required lenders to properly verify income, assess affordability on a stressed basis, and take responsibility for the suitability of their recommendations. Self-cert went in 2009. The 125% LTV products vanished with Northern Rock. Cashback mortgages were banned. Material completion gifts now require disclosure on the UK Finance Disclosure of Incentives Form, and lenders cap total incentive packages at 5% of the purchase price before reducing their loan calculations accordingly. The Škoda Fabia on the driveway is, in the most literal sense, a compliance issue now.

The market is structurally safer for all of it. A borrower in 2026 is substantially less likely to take on debt they cannot service, secured against a property valued above its real worth, with a repayment vehicle that doesn’t perform. That matters, particularly when you look at the damage the pre-MMR era caused, not dramatically, in the way that Northern Rock’s collapse was dramatic, but quietly, in the form of negative equity, interest-only shortfalls, and shared appreciation traps that took twenty years to fully resolve.

The cost is some genuine innovation that never got the distribution it deserved. Offset mortgages, which reduce the interest charged by setting liquid savings and current account balances against the outstanding loan, work exceptionally well for people with lumpy or seasonal income, entrepreneurs, freelancers, and anyone whose cash position varies month to month. They were never properly marketed to the clients who would benefit most, and they remain a fraction of the overall market. Current account mortgages, where salary flows into a single facility, and interest is calculated daily on the net balance, were quietly excellent products pioneered by First Direct and Virgin in the early 2000s. The maths works well for disciplined borrowers. They remain available and remain chronically underused. Flexible drawdown structures, borrow, overpay, and redraw against the paid-down capital, are commonplace in Australia and exist in limited forms in the UK through some specialist lenders, but have never broken into mainstream distribution.

These are products that solve real problems for a significant number of borrowers. Their obscurity is partly a marketing failure and partly a consequence of a regulatory environment that rewards simplicity and standardisation. An offset mortgage is harder to explain than a two-year fixed rate. A current account mortgage requires discipline that not every borrower has. Lenders find it easier to sell straightforward products, and regulators find them easier to oversee.

The Fabia is not coming back as a completion gift. Neither is the 125% LTV, the liar loan, the shared appreciation contract, or the endowment policy sold without a meaningful stress test. That is the correct outcome across the board. The question worth sitting with is whether the line was drawn in the right place, whether the regulatory response was precise enough to remove genuinely harmful products while preserving flexibility for complex borrowers, or whether it defaulted to a framework designed around the median case, leaving everyone outside it significantly underserved.

The evidence on self-employed borrowers, entrepreneurs, and high-net-worth individuals with non-standard income structures suggests the answer is not straightforward. The mainstream mortgage market remains poorly designed for anyone whose finances don’t look like a payslip. That is a structural problem the MMR didn’t solve, and one the market is still working around rather than through.

What is certain is that the market of 2003, for all its creativity and occasional absurdity, was producing outcomes that damaged a significant number of borrowers in ways that took years to become fully visible. The Fabia on the driveway was real. The total cost of the arrangement was less often discussed, and rarely in the same conversation.