A 25-year-old starting work in Stockholm and a 25-year-old starting work in London are both, in theory, doing the same thing – handing over a slice of their salary today in exchange for an income they won’t see for another 40 years. The mechanics behind that promise, and the returns actually delivered while keeping it are strikingly different, and one country has a lot more real data to show for it than the other.
In this post I’ll get into the Nordic vs UK pension systems comparison properly, starting with how each is actually built, look at what 25 years of real returns from Sweden’s default pension fund show, and explain why an honest UK comparison over the same horizon is harder to make than most explainers let on, before drawing out what that gap actually means for savers in each country.
- Nordic vs UK Pension Systems – How the Two Are Actually Built
- 25 Years of AP7 – What Sweden's Default Fund Actually Returned
- Why the UK Doesn't Actually Have 25 Years of This Data
- The Safety Net Underneath – State Pension vs Guaranteed Pension
- A Worked Example – What the Return Gap Is Actually Worth
- What This Means for Savers in Either Country
- Conclusion
Nordic vs UK Pension Systems – How the Two Are Actually Built
Sweden’s public pension runs on 18.5% of pensionable income, split into two very different mechanisms. 16% funds the income pension, a pay-as-you-go system recorded in notional individual accounts rather than actually invested, backed by four state buffer funds (AP1 to AP4). The remaining 2.5% funds the premium pension, a genuinely invested, individually owned defined contribution pot.
Savers can pick their own funds from an approved marketplace, or do nothing and default into AP7, the state-run alternative that around 5.9 million Swedes end up in. On top of all this sits ITP or SAF-LO, occupational pensions most employees also get through their employer.
The UK’s equivalent is auto-enrolment, introduced far more recently, in October 2012, requiring a minimum 8% of qualifying earnings into a workplace defined contribution pension, split as at least 3% from the employer and the rest from the employee (including tax relief).
Default funds carry a 0.75% annual charge cap. There’s no notional pay-as-you-go layer sitting alongside it, the UK’s PAYG-style safety net is the flat State Pension instead, funded through general taxation and National Insurance rather than a dedicated pension contribution.
| Sweden | UK | |
|---|---|---|
| Total mandatory contribution | 18.5% of pensionable income | 8% of qualifying earnings (auto-enrolment minimum) |
| Funded, invested component | 2.5% (premium pension, individually owned) | Effectively all of it (DC scheme) |
| PAYG/notional component | 16% (income pension) | None, replaced by flat State Pension from general taxation |
| Default fund | AP7 Såfa | Provider-specific default fund |
| Charge cap | Not directly comparable (AP7’s own fees are already very low) | 0.75% per year on default funds |
| System age | Premium pension since 1999 to 2000 | Auto-enrolment since October 2012 |
25 Years of AP7 – What Sweden’s Default Fund Actually Returned
This is where the “25 years” in the title actually comes from. AP7 recently marked its 25th anniversary, and its own reporting puts the numbers plainly, a total return of 594% since inception, an average annual return of 10.8%.
Over the 2010 to 2017 stretch specifically, AP7 Såfa outperformed Sweden’s own OMXS30 stock index, delivering 14.45% a year on average against the index’s 8.73%, a reminder that a well-run default fund with a sensible age-based glidepath can beat the headline index it’s loosely benchmarked against, not just match it.
That’s a genuinely rare thing to be able to say with a straight face, a quarter-century of real, audited, individually-owned default fund performance, not a backtest, not a projection, an actual track record covering the dot-com crash, the 2008 financial crisis, the Eurozone crisis, Covid, and the 2022 to 2023 inflation shock, all inside one number.
Why the UK Doesn’t Actually Have 25 Years of This Data
The honest structural point this comparison keeps running into, is that auto-enrolment itself is only around 14 years old. There is no UK default DC fund with a genuine 25-year track record covering the same period as AP7, because the system that would produce one didn’t exist yet for most of that window.
What does exist is recent performance data – the Department for Work and Pensions’ Pension Provider Survey 2024/25 put gross annualised growth for savers with 30 years to retirement at 8.6% over the prior five years, and separate industry data (CAPAdata) put young savers’ five-year annualised returns at 9.3%, with pre-retirement pots at a more conservative 6.1%. Average default fund returns for 2024 alone fell just below 5%, a reminder that any single year can look quite different from the five-year trend around it.
For long-run planning, the Government’s own DC modelling assumes a blended 88% equities, 12% bonds allocation, with average assumed returns of 6.15% and 4.43% respectively, working out to a long-run blended assumption closer to 5.7 to 6%, noticeably more conservative than the recent five-year run.
None of this is a fair like-for-like substitute for AP7’s actual 25-year number. It’s the best available proxy, and the honest conclusion is that the UK’s current pension system hasn’t been running long enough, in its current form, to produce the equivalent evidence. That’s a genuinely different kind of gap to the return-rate gap itself, and arguably a more interesting one.
The Safety Net Underneath – State Pension vs Guaranteed Pension
Both countries sit a state-funded floor underneath the invested pot, but the shape is different. The UK’s full new State Pension is £241.30 a week for 2026/27 (£12,547.60 a year), rising each year under the triple lock, the higher of average earnings growth, inflation, or 2.5%, provided you have 35 qualifying National Insurance years.
Sweden’s equivalent is the guaranteed pension, a means-tested minimum for those over 65 with low or no income and at least 40 years of Swedish residency, layered underneath the income pension and premium pension rather than sitting alongside them as a universal flat payment. It’s a genuinely different design philosophy, the UK’s floor is largely universal and contribution-record-based, Sweden’s is means-tested and sits beneath an already much larger mandatory contribution.
A Worked Example – What the Return Gap Is Actually Worth
Purely as an illustration of how much a sustained return differential compounds over 25 years, not a real currency comparison, take two savers each contributing £200 a month for 25 years.
- At AP7’s realised 10.8% average annual return, that grows to roughly £267,000.
- At the UK’s conservative long-run planning assumption of around 6%, the same contributions grow to roughly £132,000.
That’s not quite double, from a return gap of under 5 percentage points a year. This is exactly the same lesson as compounding always teaches, small differences in assumed annual return produce enormous differences in outcome over a 25-year horizon, but it’s worth being honest about what’s actually being compared here, a realised historical number from an unusually strong 25 years for global equities against a deliberately conservative forward-looking planning assumption.
If you instead extrapolate the UK’s recent five-year figures (8.6 to 9.3%) forward, the gap narrows considerably. Which comparison is “fairer” says as much about how you choose to model future returns as it does about which country’s pension system is better designed.
What This Means for Savers in Either Country
A few things worth taking from this rather than treating either system as simply “better” –
- Sweden’s mandatory contribution rate is more than double the UK’s: 18.5% versus 8% is the single biggest structural difference here, and it explains a large part of any outcome gap before returns even enter the picture,
- A 25-year live track record is worth more than it sounds: AP7’s number has actually lived through several genuine crises. Any UK equivalent is still, in a sense, a forecast dressed up as a comparison,
- Assumption choice matters enormously in any pension projection: The same UK data can support a 6% or a 9% long-run assumption depending on which window you pick, worth remembering next time a pension projection tool spits out a single confident number,
- A universal, inflation-linked state pension and a means-tested guaranteed minimum solve different problems: One protects everyone equally regardless of contribution record, the other protects specifically those who’d otherwise fall through the gaps.
Conclusion
Twenty-five years is long enough to be genuinely instructive, and Sweden’s premium pension system happens to have exactly that much real history behind its default fund, while the UK’s auto-enrolment system is still, relatively speaking, finding its feet.
That’s not a criticism of the UK’s design, a lower mandatory contribution rate paired with a more generous universal state pension is a coherent choice, just a reminder that some pension comparisons are really comparisons of contribution rates and design philosophy dressed up as comparisons of investment performance.
The returns are the easy part to compare – the structure underneath them is where the real difference lives.