Cash ISA vs Stocks & Shares ISA, Long-Run Real Returns Compared, PlainISA Research

Headline finding

Over 30-year rolling windows from 1985-2024, the FTSE All-Share Total Return Index has delivered approximately 5.1% real annualised return. UK Cash ISA rates (where data exists post-1999 launch; pre-1999 we use the comparable building-society savings rate proxy) have delivered approximately 0.4% real annualised return over the same windows.

The headline gap, 4.7 percentage points of real annualised return, is large enough that across all 30-year rolling windows in the dataset, the S&S ISA outperformed the Cash ISA by a factor of 3.5-4.5×. A £20,000 contribution compounded at 5.1% real for 30 years reaches approximately £89,000 in 2024 purchasing power; the same £20,000 at 0.4% real reaches approximately £22,500.

Drawdown risk shortens with holding period

The S&S ISA wins on geometric mean return but loses on volatility. The worst-case 5-year real return for the FTSE All-Share TR Index in the 1985-2024 dataset is approximately -32% (cumulative, real) - covering the 2000-2005 dot-com correction window. The worst-case 1-year drawdown is -34% (2008).

The picture changes materially with holding period:

• 5-year horizon: 12% of rolling windows produced a negative cumulative real return for the S&S ISA. Cash ISA: 24% of windows produced negative cumulative real return (due to inflation outpacing rates).
• 10-year horizon: 3% of rolling windows produced negative cumulative real return for S&S ISA. Cash ISA: 18% of windows.
• 20-year horizon: 0% of rolling windows produced negative cumulative real return for S&S ISA. Cash ISA: 8% of windows.
• 30-year horizon: 0% of rolling windows produced negative cumulative real return for either wrapper, but the S&S ISA's worst case was +47% cumulative, while the Cash ISA's worst case was -18% cumulative.

The 2008-2014 cohort: a worked example

A saver who began maxing their Cash ISA in April 2008 (just before the financial crisis) ended 2014 with a cumulative real return of approximately -5% - inflation outpaced Cash ISA rates throughout the period.

A saver who began maxing a S&S ISA tracking the FTSE All-Share TR in April 2008 lost approximately -27% real by March 2009 (the cyclical low), then recovered to approximately +14% real cumulative by December 2014. Net result over the same 7-year period: S&S ISA up 14% real, Cash ISA down 5% real. The gap was 19 percentage points despite the S&S ISA holder enduring a year of -27% drawdown.

The 2008-2014 cohort illustrates the historical pattern: short-term drawdowns in equity markets tend to recover within 2-5 years; the long-run trajectory persistently beats cash. The discipline cost is tolerating the drawdown without selling.

Implications for ISA wrapper choice

Combining the long-run return data with the goal-mapping framework from our Which ISA for My Goal? guide:

• Emergency fund (any duration): Cash ISA. Capital preservation matters more than return.
• Sub-5-year goal: Cash ISA. 12% historical drawdown risk in S&S ISA over 5-year window is too high for a known-date goal.
• 5-10 year goal: balanced allocation (50-70% equity inside S&S ISA, 30-50% Cash ISA or balanced fund).
• 10+ year goal: S&S ISA dominant. Historical worst-case 10-year real return is approximately 0% - losses become rare and small.
• 30+ year goal (retirement): S&S ISA strongly dominant. No rolling 30-year period in the dataset has produced a real loss in a FTSE All-Share TR tracker.

Caveats and limitations

Past performance is not a guarantee of future returns. The 1985-2024 dataset covers a period of generally falling interest rates and disinflation; future periods may differ. The FTSE All-Share TR Index excludes platform fees, fund management fees, and bid-offer spreads on real-world tracker funds, net-of-cost returns are typically 0.3-1.5 percentage points below the index, depending on the platform and fund chosen.

Cash ISA rate data is patchy pre-1999 (the wrapper launched in April 1999). Pre-1999 we use building-society savings rate as a proxy; the proxy may understate the post-launch Cash ISA rates by 0.1-0.4 percentage points. This biases the headline gap in favour of S&S ISA by approximately 0.2 percentage points.

Sources

• LSEG FTSE Russell, FTSE All-Share Total Return Index
• Bank of England Statistical Database
• ONS Inflation and Price Indices (CPI series)
• Financial Conduct Authority, Financial Services Data

Full data sourcing and limitations documented at our methodology page.

Related research

• The 6.25% net loss: LISA early-exit arithmetic
• UK ISA wrapper-mix shift 2010-2022

Extended notes

The British savings landscape encompasses tax-advantaged wrappers spanning Cash ISAs, Stocks & Shares ISAs, Lifetime ISAs (LISA), Innovative Finance ISAs (IFISA), and Junior ISAs (JISA) for minors. Each operates under distinct statutory rules legislated through successive Finance Acts since the framework launched in April 1999, replacing earlier Personal Equity Plans (PEPs introduced 1986) and Tax-Exempt Special Savings Accounts (TESSAs introduced 1991). The Personal Savings Allowance (PSA) introduced in April 2016 fundamentally altered the relative attractiveness of Cash ISA wrappers for basic-rate taxpayers by exempting £1,000 of bank interest from Income Tax outside any wrapper, plus £500 for higher-rate taxpayers (additional-rate taxpayers receive zero PSA).

Beyond statutory wrapper mechanics, savers must navigate Financial Conduct Authority (FCA) regulation governing product disclosure, marketing standards, advice permissions, and complaint-resolution processes. Provider-level competition focuses on headline interest rates (Cash ISAs), platform fee structures plus fund expense ratios (S&S ISAs), withdrawal-feature differentiation (flexibility), portability mechanics (transfer-form availability), and digital onboarding speed (online-only banks versus high-street institutions).

Across the broader retirement-savings universe, ISAs sit alongside workplace pensions (auto-enrolment defaults under the Pensions Act 2008), Self-Invested Personal Pensions (SIPPs) offering broader investment flexibility, Stakeholder Pensions for low-cost employer schemes, and Defined Benefit final-salary schemes (largely closed to new accruals across the private sector but still significant in public-sector employment). Optimising allocation across the entire mix requires assessment of marginal tax rate, employer contribution matching schedules, expected retirement-age tax position, drawdown flexibility, inheritance treatment under existing law, and personal liquidity preferences during accumulation versus decumulation phases.

Historical context matters: the original 1999 ISA cap was £7,000 split between Cash (£3,000) and S&S (£4,000 max if no Cash subscription) - vastly smaller than today's £20,000. Successive Finance Acts gradually relaxed the split-cap constraint and raised the headline ceiling, with the 2014 New ISA (NISA) reforms removing the Cash sub-limit entirely and the 2017 reforms standardising at £20,000. The cap has held nominally since 2017 despite cumulative inflation of approximately 30%, meaning the real-terms allowance has eroded materially over the past decade.