Methodology: how every number is calculated

All price history is end-of-day data from Yahoo Finance's chart API: one bar per trading day with the closing price and the adjusted close. The adjusted close is the closing price restated for dividends and splits, so a series of adjusted closes behaves like a fund in which every dividend was reinvested on the day it was paid. Every return on this site is computed from adjusted closes and is therefore a total return with dividends reinvested, quoted after the fund's own running costs — the TER is already inside the price.

Each fund is followed through one specific trading line, and that line sets the currency of every figure on its pages:

• European UCITS ETFs — the euro trading line on Xetra (for example SXR8, VWCE, EUNL). One exception: VUAA is priced through its euro line on Borsa Italiana, because Yahoo's Xetra history for it only starts in late 2024 while the Milan line covers the fund since 2019.
• UK UCITS ETFs — the pound line on the London Stock Exchange (CSP1, VWRP, SWDA and so on). Lines quoted in pence are converted to pounds.
• US-listed ETFs — the fund's US primary listing in dollars (SPY, VOO, QQQ and so on).

Fund facts are handled separately. Cost (TER), launch date, ISIN and WKN are curated by hand and were last checked on 14 July 2026 against the issuers' documents and justETF. US funds' expense ratios are additionally refreshed from Yahoo Finance each day, because fee cuts happen; for UCITS funds Yahoo's figures are unreliable, so the curated TER is used.

Prices are refreshed once a day in the early morning UTC, after the previous session's closes are final. Every data page shows the exact date its prices run to ("Prices through …") and when the series was last refreshed, and pages regenerate as soon as new prices arrive. If a refresh fails, the stamp keeps showing the older date rather than pretending — check it before quoting a number.

Yahoo's history for a few UCITS lines contains bad early data, which we detect by comparing each fund against another fund tracking the same index. Where the early bars are broken, the series is clamped to start where it becomes reliable: SPYY from January 2012, SPYI from January 2015 and VUAA from January 2021. Those funds show fewer years than they have existed; the years shown are correct.

A single-bar spike filter removes obvious bad prints: a day whose adjusted close sits more than 12% away from both its neighbours while those neighbours are within 5% of each other is dropped. Genuine crash and rally days survive, because the following day does not revert.

A daily move is the percentage change in adjusted close from the previous trading day. Days are bucketed at ±0.5%, ±1%, ±1.5% and ±2%; a big day is any move of 1% or more in either direction, and a year is called quiet, normal, bumpy or wild by the share of its days that were big days (under 15%, under 30%, under 45%, and above).

Volatility is the sample standard deviation of the year's daily moves, annualised by multiplying by the square root of 252 trading days. It is labelled calm below 13%, normal below 20%, choppy below 28% and wild above that. A streak is a run of consecutive up days or consecutive down days; an unchanged close ends the run.

A monthly return compounds the daily moves that fall inside the calendar month — equivalent to the change from the last close of the previous month to the last close of this one. The seasonality pages use the same month-end to month-end definition across every year the fund has traded, drop the still-running month and never count a partial first month.

A calendar-year return compounds the year's monthly returns. For the current year only completed months are counted and the figure is labelled (so far); the running month is excluded so the number does not swing with every session. A fund's launch year is shown as the return from its first trading day to year end, and is excluded from multi-year averages when the fund launched after January.

"10,000 invested at the start of year X became Y by the end of year Z" multiplies 10,000 by each calendar-year return in the window. The average annual return is the compound annual growth rate over that window: total growth to the power of one over the number of years, minus one.

On a head-to-head page both funds are measured over the same window: the most recent full calendar years — up to ten — in which both were trading, excluding the running year and either fund's incomplete launch year. A young fund therefore shortens the window for both; the page says how many years it counts. A year is a win for the fund with the higher return; a difference under 0.01 percentage points is a tie.

Lump sum invests the full 10,000 at the start of the year and compounds the monthly returns. DCA (dollar-, euro- or pound-cost averaging) invests one twelfth — 833.33 — at the start of every month, and each instalment rides that month's return. Because DCA still holds uninvested cash during the year, the comparison uses total wealth: invested balance plus cash not yet deployed. Gains are also shown against the money actually paid in.

The what-if pages take an amount and a start date and follow the adjusted-close path from that day to the latest price, so "10,000 five years ago" is exactly five years of history with dividends reinvested. The fixed 1, 5, 10 and 20 years ago answers and the since-launch figure use the same rule; the calculator only rescales by your amount.

The pie calculator invests the chosen monthly contribution at the start of each month, split across the selected funds by their weights, and values the portfolio at each day's adjusted close. It starts on the first date for which every fund in the pie has prices, so a young fund shortens the backtest for the whole pie.

The currency pages put the same index side by side in two currencies using two real funds: a US-listed tracker in dollars (SPY for the S&P 500, URTH for MSCI World, ACWI for MSCI ACWI) and the corresponding UCITS fund's euro or pound line (SXR8, EUNL and IUSQ on Xetra; the LSE lines for the UK). The gap is the local-currency calendar-year return minus the dollar return. It is dominated by the exchange rate, but it also contains small differences in fees, tracking and dividend tax between the two funds, which is why we compare real funds rather than quote a theoretical FX conversion.

• Taxes. No withholding, capital-gains or savings-plan taxes are modelled. Adjusted closes assume gross reinvestment of the dividends a fund distributes.
• Trading costs. No broker fees, exchange fees or bid-ask spreads. Buying on the first trading day at the close is an idealisation.
• Source risk. Yahoo Finance is an unofficial, best-effort source without a service guarantee. We clean what we can detect; report anything that looks wrong and we will check it against the issuer's data.
• The future. Everything here is history. It says what happened to real prices, not what will happen.

If a number does not match your own calculation, tell us which page, which figure and what you got: originave@atomicmail.io. Differences usually come down to one of the rules above — which trading line, which currency, whether dividends were reinvested, or whether the running year was counted — and we are happy to walk through it. See also who runs Originave and the glossary.