Key takeaways
- Vanguard's FTSE All-World UCITS ETF charges an ongoing charge of 0.19% a year, yet across the ten calendar years from 2016 to 2025 it never diverged from its stated benchmark by more than 0.07 percentage points, and finished 0.15 points behind in total.
- That charge excludes dealing costs by design. The methodology behind it, published in 2010 and still in use, strips out "brokerage charges, taxes and linked charges" and market impact costs.
- The iShares Core UK Gilts factsheet shows a 12-month trailing yield of 4.44% and a yield to worst of 4.63 on the same page, 0.19 points apart, answering different questions.
- That gilt fund charges 0.07% and has annualised 2.33% since its December 2006 launch against its benchmark's 2.46% — 0.13 points a year, which the factsheet never explains.
- Its securities lending split, 62.5% to the fund and 37.5% to BlackRock, appears in the Key Investor Information Document rather than the factsheet, and is excluded from the 0.07%.
Four lines change a decision. The rest is context
You've opened a fund factsheet and you want to know which lines actually matter. On the two real documents used here, four do: the charge, the index the returns are measured against, the calendar-year table, and — on a bond fund — duration. Everything else is context, or decoration.
The two documents are Vanguard's FTSE All-World UCITS ETF factsheet dated 30 June 2026, and the iShares Core UK Gilts UCITS ETF factsheet with the same data date, plus its Key Investor Information Document. Both label themselves at the top: Vanguard says "This is a marketing communication", iShares says "This document is marketing material." That isn't a technicality. The marketing document is the one everybody reads, and it's the one that leaves things out.
"Ongoing Charges Figure 0.19%" and the costs it is built to exclude
Vanguard's sheet quotes one cost. Its footnote says the figure "covers administration, audit, depository, legal, registration and regulatory expenses incurred in respect of the Funds."
What it doesn't cover is what the fund pays to trade. The calculation still follows the methodology the Committee of European Securities Regulators published on 1 July 2010, which excludes "payments to third parties to meet costs necessarily incurred in connection with the acquisition or disposal of any asset for the UCITS' portfolio, whether those costs are explicit (e.g. brokerage charges, taxes and linked charges) or implicit (e.g. costs of dealing in fixed-interest securities, market impact costs)." Performance fees and interest on borrowing are out too.
Two more properties of the number are worth knowing. It's calculated "at least once a year, on a ex-post basis" — history, not a price list — and it's quoted "as a percentage to two decimal places", so any difference finer than that never reaches the page. The layers that sit outside it are the subject of a separate piece on the total cost of owning a fund.
"Total Expense Ratio: 0.07%" is the same number wearing a different label
The gilt factsheet says "Total Expense Ratio : 0.07%". Its Key Investor Information Document says "Ongoing Charges 0.07%". Same fund, same figure, two names — and only the second document defines it. That figure, it says, "excludes portfolio trade related costs, except costs paid to the depositary and any entry/exit charge paid to an underlying collective investment scheme."
The same document carries something the factsheet never mentions. A footnote states that "the Fund will receive 62.5% of the associated revenue generated and the remaining 37.5% will be received by BlackRock as the securities lending agent", and that because this "does not increase the costs of running the Fund, this has been excluded from the ongoing charges." A revenue split that moves your return sits one document away from the page most people read. Who takes what from lending revenue is a question in its own right.
The benchmark line names an index, not which version of it
Vanguard's sheet says the fund "seeks to track the performance of the FTSE All-World Index". The performance footnote says "Basis of index performance is total return." Page one gives an index ticker: TAWNT01U.
Those are not the same statement. FTSE Russell's ground rules for the Global Equity Index Series, dated July 2026, describe two families. "Declared dividends are used to calculate the Standard Total Return Indices in the FTSE Global Equity Index Series." Then, separately: "A series of net of tax Total Return Indices are also calculated based on the maximum withholding tax rates applicable to dividends received by institutional investors who are not resident in the same country as the remitting company and who do not benefit from double taxation treaties."
A fund holding foreign shares pays withholding tax. An index that ignores it is a target a fund paying that tax cannot match. An index that assumes the harshest rate is one a fund can beat. The words "total return" don't tell you which one the bars are drawn against. The ticker is the only thing on the page that pins it down, and the page doesn't decode it.
Ten calendar years, and a fund that finished 0.15 points behind in total
Vanguard's table runs from 2016 to 2025. Against its benchmark, the fund landed -0.02, 0.01, -0.05, 0.05, -0.02, -0.07, -0.01, 0.03, -0.01 and -0.06 percentage points. Those ten gaps are what the chart above plots. They sum to -0.15 points across the decade, roughly 1.5 basis points a year, against a charge of 0.19% a year.
That is the most useful thing the table does, and it's easy to skim past. A fund charging 19 basis points trailed its stated index by about 1.5. Something recovered the other 18 basis points a year: the index variant, lending revenue, tax treatment, or some mix of the three. The factsheet names none of them. The gap between a fund and its index has a name and a literature of its own — tracking difference is not tracking error, and the two get confused constantly.
Why the table starts in January and stops last December
The calendar framing is inherited from Article 15 of Commission Regulation 583/2010, which required a bar chart "covering the performance of the UK UCITS for the last 10 years" and barred any "record of past performance for any part of the current calendar year". Article 18 required the benchmark's bar "alongside each bar showing the UK UCITS' past performance". The regulation was revoked in the UK on 6 April 2026, but the shape it imposed is on every document quoted here.
The cost of that shape is simple. A year that started badly and ended well reads as a good year, and so does the reverse. The gilt fund's Key Investor Information Document, accurate as at 09 February 2026, shows performance to 31 December 2025 and then stands still while the year it omits runs on.
Rounding decides whether the fund beat its index
Compare the same two rows across the two documents. The factsheet reports 2024 as -3.34% for the share class and -3.32% for the benchmark. The Key Investor Information Document, obeying a one-decimal convention, reports both as -3.3.
Neither is wrong. But one page shows a fund 0.02 points behind its index and the other shows a dead heat, and a reader comparing two funds across two documents has no way to tell which convention produced which digit.
"Top 10 approximately equals 24.0% of net assets"
Vanguard's holdings block lists NVIDIA at 4.5%, Apple at 4.0% and Alphabet at 3.6%, then states that the "Top 10 approximately equals 24.0% of net assets". Market allocation puts the United States at 61.7%, and technology at 35.1% of weighted exposure. A fund holding 3,782 stocks has a quarter of your money in ten of them. That line changes a decision, and it's among the most skimmed on the page.
The same line on the gilt fund is decoration. Its top issuers read "UK CONV GILT 88.49%" and "UNITED KINGDOM OF GREAT BRITAIN AND NORTHERN IRELAND (GOVERNMENT) 11.42%", totalling 99.91% — one borrower under two labels. Concentration says nothing about a single-issuer government bond fund. It says a great deal about an equity tracker whose index has drifted. What no single factsheet can show is the overlap between the funds you hold. That's the version of the question that matters once you own more than one. A portfolio tracker such as LedgerTouch sees every holding at once; a factsheet, by construction, cannot.
One more line in that block repays a second look. The sheet reports 3,782 stocks in the fund against 4,264 in the benchmark, 482 fewer. That gap is the method: the fund tracks by "investing in a representative sample of Index constituent securities."
Two yields on one page, 0.19 points apart
The gilt factsheet gives a "12m Trailing Yield : 4.44%" and a "Yield to Worst : 4.63". Both are yields. Neither is what you'd receive.
The sheet's glossary defines one of them and not the other. Yield to maturity, it says, is "the discount rate that equates the present value of bond's cash flows with its market price", and "the measure does not include fees and expenses" — so the 4.63 is quoted before the 0.07% charge comes out. The trailing figure gets no definition at all. On the equity side Vanguard reports an "Equity yield (dividend)" of 1.5%, then detaches it from the fund in its own glossary: "The dividend yield stated on the factsheet applies to the underlying holdings of the ETF, and not the ETF itself."
Both sheets share the habit. The definition sits on a different page from the figure, in smaller type, and the figure is unreadable without it.
"Effective Duration : 7.21 years" is the only line that points forward
Nearly everything on a factsheet describes the past. Duration doesn't. The gilt fund reports an effective duration of 7.21 years against an average weighted maturity of 10.45 years. Its glossary calls duration "a measure of a fund's interest-rate sensitivity". It adds that "a fund with a duration of 10 years is twice as volatile as a fund with a five-year duration".
What a given rate move does to a given duration is worked through separately. For reading the sheet, the point is that duration is the one number answering "what happens next". The 2022 row of the same table shows the answer when it arrives: the share class returned -23.83% and the benchmark -23.83%.
Three figures that aren't on either sheet, and where they live
Guidelines issued by the European Securities and Markets Authority on 1 August 2014 put two of them in the annual report. The annual report "should also disclose and explain the annual tracking difference between the performance of the UCITS and the performance of the index tracked". It should also contain "the revenues arising from efficient portfolio management techniques for the entire reporting period together with the direct and indirect operational costs and fees incurred."
The third, dealing costs, was written out of the charge figure in 2010 and restored nowhere on the sheet. So the numbers that would explain a tracking gap sit in the umbrella company's annual report, a document covering every sub-fund at once, while the promotional one-pager covers a single share class. That's the disclosure architecture, not an accident of layout.
A single month moved the gilt fund's five-year number by 0.82 points
The same fund publishes a sheet every month. Between the 30 June 2026 edition and the 31 July 2026 one, effective duration went from 7.21 to 7.18 years, the trailing yield from 4.44% to 4.51%, and yield to worst from 4.63 to 4.90. The five-year annualised return went from -4.20% to -5.02%.
Nothing was restated. The window rolled forward one month and a five-year number moved 0.82 percentage points. Any figure ending "5 years" or "since inception" is as much a function of the print date as of the fund.
The objection: for a plain tracker, two lines would have done
There's a serious case against reading any of this, and the Vanguard evidence is what makes it serious. Over ten calendar years that fund never diverged from its index by more than 0.07 percentage points. Its ten-year annualised return was 12.73% against the index's 12.74%. Read two lines — the index name and the 0.19% charge — and you'd have known almost everything the rest of the page tells you.
For a large, liquid, developed-market tracker, that's right, and the line-by-line read earns nothing.
It stops being right where the numbers stop agreeing. The gilt fund charges 0.07% and has annualised 2.33% since launching on 1 December 2006, against 2.46% for its benchmark. That's 0.13 points a year, off the same row of the same table, and more than the whole fee. The factsheet offers no explanation, and none of the three candidate explanations is printed on it. The read is cheap. The case where it's needed is the case you can't identify in advance.
The limitations of a sample of two
This is two funds, two months and one document family each. Both are cheap index trackers, which is where a factsheet is least likely to mislead. An active fund with a performance fee and a soft comparator is a harder document, and none is examined here.
Every return quoted is on a net asset value basis. Vanguard states that its performance "does not take account of the commissions and costs incurred in the issue and redemption of shares", and the iShares sheet warns that "the market price at which the Shares are traded on the secondary market may differ from the Net Asset Value per Share". Neither is what landed in your account after platform charges and a spread.
The rules cited are UK and European rules; a US fund's factsheet answers to different ones. And the tables are past performance, which both documents say plainly on the page: iShares that "past performance is not a reliable indicator of current or future performance", Vanguard that it "is not a reliable indicator of future results".
What would change this
The disclosure regime is mid-replacement. The Financial Conduct Authority published final rules for Consumer Composite Investments on 8 December 2025. The legislation commenced on 6 April 2026, and the rules "will come into effect on 8 June 2027". Three of the changes land directly on the problems above.
The charge gains a second form. Firms must "disclose the OCF as the headline figure, both as a percentage, and as a pounds-and-pence number". That turns 0.19% into a sum of money. The bar chart becomes a line graph of "the product's past 10 years of performance". It starts from an investment of "£10,000", on monthly data points rather than quarterly. And explicit dealing costs get a number of their own, "presented separately to ongoing costs" — though the same rules confirm that "transaction costs are not included in the OCF", and that a bid-ask spread counts as implicit and stays out.
Two of the gaps here close if that lands as written. The one that doesn't is the index variant: nothing in the new rules decodes TAWNT01U. Until a sheet says in words which version of its benchmark those bars belong to, the comparison it invites you to make is the one it hasn't defined. That, rather than the charge, is where the next fund document is worth testing.