Terry Smith's Fundsmith portfolio: strategy and performance

Terry Smith's Fundsmith Equity Fund is a concentrated, all-equity portfolio: at 30 September 2026 it held 28 companies, 1.9% cash and no bonds, with 82.3% of the fund listed in the US. Against MSCI World in sterling, it is partly set up for different market scenarios and has no guard against a catastrophic equity fall.

On volatility-adjusted returns and on returns after fees it is partly better: behind over the last ten years and since February 2014, ahead since its launch in November 2010 (13.0% a year against 12.4% to 30 September 2026, on Fundsmith's figures).

Key takeaways

1. Set up for different market scenarios: partly. It lost 10.3% in the Covid crash against 15.6% for MSCI World, but 13.7% in 2022 against 7.8%, and gained 22.1% in the 2023 to 2026 rally against 80.0%.
2. Guard against catastrophic equity falls: no guard. The fund holds only shares, by design; its worst month-end drawdown, 17.7%, was deeper than the index's 15.7%.
3. Better volatility-adjusted returns than MSCI World: partly. Its Sharpe ratio was 0.63 against 0.89 over ten years and 0.91 against 0.99 from February 2014, but 0.99 against 0.97 since launch, and it won 19 of 33 rolling ten-year windows.
4. Better returns net of fees than MSCI World: partly. It compounded at 9.3% a year against 12.6% over ten years and 12.7% against 13.0% from February 2014, but 13.0% against 12.4% since 2010 on Fundsmith's figures.
5. "Do nothing" changed in 2026. Turnover reached 51.8% in the first half of 2026, against 3.2% in 2024.

How does Fundsmith score on our four tests?

We judge Fundsmith Equity (I Acc) against MSCI World Net in sterling, with SONIA as cash, using monthly returns to September 2026.

Windows: 10 years, Oct 2016 to Sep 2026; full window, Feb 2014 to Sep 2026, set by our price data; since launch, Nov 2010 to Sep 2026, from the factsheet's monthly returns. There is no 20-year window and no dotcom or 2008 episode, because the fund launched in 2010.

What does Terry Smith's Fundsmith Equity Fund hold?

At 30 September 2026, the fund held 28 listed companies worth £11.1bn in total, according to Fundsmith's September 2026 factsheet. Its ten largest holdings, listed without weights, were Waters, Visa, Mastercard, Amadeus, Veeva Systems, Stryker, L'Oréal, Uber, Church & Dwight and Fortinet. By country of listing, 82.3% was in the US, 7.3% in France, 4.9% in Spain, 3.6% in the UK and 1.9% in cash. Nine months earlier, the December 2025 factsheet showed 29 holdings, £16.1bn of assets and 69.9% listed in the US.

Share of the fund at 31 December 2025, then 30 September 2026: health care 30.8% and 18.7%; information technology 10.4% and 16.7%; consumer discretionary 16.5% and 14.6%; consumer staples 19.9% and 14.3%; industrials 8.2% and 13.3%; financials 5.4% and 10.7%; communication services 8.8% and 9.8%; cash 0.0% and 1.9%. Source: Fundsmith Equity Fund factsheets for December 2025 and September 2026, page 1; chart data (CSV). Weights are shares of the whole fund by GICS sector and each column sums to 100%.

The I Acc shares charge 0.95% a year and the T Acc shares, used in Fundsmith's published performance, 1.05%. The fund uses no derivatives or borrowing and does not hedge currencies, so sterling investors carry the dollar exposure in full.

What is Terry Smith's investment strategy?

Smith summarises the strategy in three steps: "Buy good companies", "Don't overpay" and "Do nothing" (annual letter, January 2026, p. 11). Good companies, in the fund's criteria, earn high returns on capital, have advantages that are hard to copy and can reinvest cash at high rates; the letter puts the portfolio's 2025 return on capital employed at 31%, against 17% for the S&P 500 excluding financials.

The owner's manual states the goal: "We aim to run the best fund ever. By best fund, we mean the one with the highest return over the long term, adjusted for risk." It rules out market timing and adds that the fund "is not meant to provide an asset allocation tool.

We assume that if you own the Fundsmith Equity Fund you have already taken the decision to invest that part of your portfolio in equities." That is why the fund has no bond or cash allocation by design: the cash on the factsheet is a by-product of dealing, not a reserve.

Smith founded Fundsmith in 2010 and remains its chief executive and chief investment officer. For another concentrated book of listed businesses, see Warren Buffett's Berkshire Hathaway.

Why has Fundsmith underperformed MSCI World since 2022?

Fundsmith's own figures show the fund behind MSCI World in sterling in each calendar year from 2022:

Source: Fundsmith factsheet, September 2026, page 1.

Smith's letters explain the gap in his own terms, not as a tested attribution. For 2022, he wrote that rising interest rates hit highly rated shares hardest (January 2023 letter, pp. 4-5). For 2025, he named index concentration, with the ten largest stocks at 39% of the S&P 500; flows into index funds, which he calls a momentum strategy; and a weaker dollar (January 2026 letter, pp. 2-8). Its own picks hurt too: Novo Nordisk cost it 3.0 percentage points in 2025.

What changed at Fundsmith in 2026?

In his July 2026 letter, Smith kept the three-step mantra but wrote that "a buy and hold strategy can only work if you are not subject to flows, and we are" (p. 6). He added that investors "should therefore expect that we will be more active in future", taking more account of momentum in business results and share prices.

Turnover was 51.8% in the first half of 2026 (p. 10), against 3.2% in 2024 and 12.7% in 2025 (January 2025 letter, p. 7; January 2026 letter, p. 14). It began buying 12 companies, including TSMC and Mastercard, and exited or began exiting 13, including Novo Nordisk and Unilever. Morningstar's analyst, in a note dated 6 August 2026 that Fundsmith republishes, called the turnover "a stark departure" but judged the underlying philosophy "broadly unchanged".

What does Terry Smith advise individual investors, and what do imitators miss?

Smith's advice to individuals is mostly about what the fund is not: not an asset-allocation tool, not a market timer, and not an index fund, since "If you want an index fund you can buy one with much lower costs than we or any other active investment manager apply" (January 2026 letter, p. 9). In January 2025 he described the aim as "a high likelihood of a satisfactory return rather than the chance of a spectacular return" (p. 11).

Popular "Terry Smith portfolio" lists copy the factsheet's top ten names, which carry no weights, omit positions still being built and changed within months in 2026. "Do nothing" never meant never selling.

Is Fundsmith set up for different market scenarios?

Partly: its quality tilt helped in a sudden crash and hurt in a rate shock and a narrow rally. In the Covid crash of February and March 2020 the fund lost 10.3% against 15.6% for MSCI World. In 2022, when interest rates rose, it lost 13.7% against 7.8%. In the rally from January 2023 to September 2026 it gained 22.1% against 80.0%. Our data start in 2014, so they cover neither the dotcom bust nor the 2008 crisis, both of which predate the fund.

Fundsmith against MSCI World, in GBP: Covid crash (February to March 2020) -10.3% and -15.6%; inflation and rate shock (2022) -13.7% and -7.8%; rally (January 2023 to September 2026) +22.1% and +80.0%. Source: our calculation from Yahoo Finance prices for the I Acc shares and MSCI World Net Total Return levels in GBP; chart data (CSV); monthly returns (CSV). The fund is net of its charges and the index has none.

In the 10 months when MSCI World rose 5% or more, the fund gained 6.1% on average against 7.0%; in the 10 months when it fell 5% or more, the fund lost 4.9% against 6.2%, and rose only once. Its monthly returns had a correlation of 0.81 with the index and a beta of 0.87, so about two thirds of their variation was shared.

By count Fundsmith looks spread, with 28 companies in seven sectors; by behaviour it is a different flavour of world equities, not a separate driver (how correlation drives portfolio volatility).

Does Fundsmith guard against catastrophic equity falls at all?

No guard: nothing in the fund is designed to hold value or pay out in a broad equity crash. It holds no bonds, gold, hedges or cash reserve, and it does not short; its only cushion is a lower sensitivity to the index, which helped in 2020 and failed in 2022.

Its worst month-end drawdown was 17.7%, from December 2021 to June 2022, and it regained that peak only in February 2024; MSCI World's worst, 15.7% in early 2020, was recovered by June 2020. At September 2026 the fund was still 6.5% below its January 2025 peak while the index stood at a new high (how recovery time works). It also had falls of its own: 9.5% in January 2022 and 9.3% in March 2026, when the index fell 4.4% and 4.5%.

Smith's owner's manual (p. 11) is candid about concentration: "research has shown that you can achieve close to optimal diversity with 20 stocks. But this may not be true in our case, as our investment criteria typically lead to a concentration in certain sectors." On the manual's logic, any defence against a crash is the investor's decision, made in the rest of the portfolio.

Are Fundsmith's volatility-adjusted returns better than MSCI World?

Partly: behind over ten years and over our full window, narrowly ahead since launch, and ahead in 19 of 33 rolling windows, all ending by July 2025. The Sharpe ratio divides the average monthly return above cash by the volatility of that excess return, annualised. The Sortino ratio divides the same excess return by downside deviation, so only months that fell short of cash enter the denominator.

Of the 33 rolling ten-year windows in our data, the fund had the higher Sharpe ratio in 19 and the higher Sortino ratio in 19, all ending by July 2025; it lost every window since.

Split at 2022, the Sharpe ratio was 1.63 against 1.19 from February 2014 to December 2021, then -0.15 against 0.66, when the fund returned less than cash (what a negative Sharpe ratio means). To compare at equal volatility, we diluted the fund with 5.7% cash to match the index's 11.5%: that mix compounded at 12.1% a year from February 2014, against 13.0%.

Fundsmith's published ratios point the other way. Its January 2022 letter gave a Sharpe ratio of 1.31 against 0.78 from launch to 2021, with a 1.5% risk-free rate, calculated by FE Analytics; later letters give only FE's Sortino ratio, 0.75 against 0.48 from launch to 2025, with a 3.5% rate.

The window, the cash rate (fixed, against SONIA; see which cash rate belongs in a Sharpe ratio) and the definition differ, and FE does not state its data frequency, so only the ranking is comparable; over our 10-year and full windows it is reversed. Smith himself writes: "We don't agree that true volatility is measured by ratios such as the Sharpe or Sortino ratio" (January 2025 letter, p. 10).

Are Fundsmith's returns net of fees better than MSCI World?

Partly: no over the last ten years, narrowly no over our full window, and yes since launch on Fundsmith's figures.

Fund returns are net of fees; MSCI World Net is an index and carries no costs, while a fund that tracks it does (the iShares Core MSCI World UCITS ETF charges 0.20% a year), so a shortfall smaller than such a fee counts as close rather than a clear loss.

Over the ten years to September 2026 the I Acc shares compounded at 9.3% a year after fees, turning £10,000 into £24,424, against 12.6% and £32,644 for MSCI World. From February 2014 they compounded at 12.7% against 13.0%: a 0.3-point gap, a little more than that tracker's fee.

Since launch, Fundsmith reports 13.0% a year for its T Acc shares against 12.4% to 30 September 2026; the first 39 months, before our data, contributed a gain of 56.6% against 36.4% (our calculation from the factsheet). The fund won 20 of 33 rolling ten-year windows on compound growth, every one ending between January 2024 and August 2025, and has lost all 13 since.

Fees are part of the gap: Fundsmith's total cost of investment, including dealing, was 1.06% for T Acc shares in 2025 and 1.12% in the first half of 2026. A compound growth rate in one window is an outcome, not proof that a portfolio maximises long-term growth (Sharpe ratio versus CAGR).

Weighing the four results

Fundsmith gave up any built-in defence against a broad equity fall, and accepted long stretches of trailing the index, in exchange for a portfolio of high-return companies that compounded faster than the market in its first decade. Since 2022 that trade has cost it, both before and after adjusting for volatility, and in 2026 the manager changed how he runs it. Whether the earlier record or the recent one is the better guide is a judgement each reader has to make.

What can an individual investor learn from Terry Smith?

Smith's record shows a strict quality rule beating an index over almost 16 years and trailing it badly for four. Questions to examine in your own portfolio:

• What is your active equity fund's correlation with the world index you already own, and what does it add beyond it?
• Was your equity allocation set deliberately, as Smith's manual assumes, or by the funds you happened to buy?
• How would you respond to four years of trailing the market?
• What is your funds' total cost, including dealing?
• How much rests on one sector, one currency or one market?

See how Bill Ackman runs concentrated positions and how conservative allocations compare.

How can copio help you examine your own allocation?

Copio is in private beta across Europe, with early access by application. Its fund look-through breaks funds into asset classes, and its return-correlation map places each as a main driver, true diversifier, dead weight or shock absorber, so a quality fund's shares are counted with the world equities they overlap.

A concentration view models a fall in your largest position, a correlation matrix shows what moves together, a crash test shows how a mix like yours would have fared in past crises, and an efficient-frontier chart places your portfolio against a modelled frontier, labelled not a forecast. Copio provides educational analysis, not personalised advice, and does not place trades; it shows trade-offs, not an optimal allocation. The free portfolio checker runs a first reading.