Methodology
How every figure on this site is produced: where it comes from, how it is graded, which firms it is computed over, when it is published, and how it is corrected.
What this site publishes, and what it is not
FinancialAdvisor.co.uk publishes two things. The Fee Register will be one page per FCA-authorised advice firm, showing the firm details the FCA Financial Services Register holds and the adviser charges the firm has itself published, submitted or confirmed, with the source address and the capture date beside every figure. The UK Advice Fee Index is a statistical series built from those schedules, with a quarterly Release at a permanent address and a monthly refresh in place; the current benchmark page is at advice fees.
- It is not advice. Nothing here is a personal recommendation, and nothing here says whether a firm or a fee is right for you.
- It is not regulated. The operator of this site is not authorised or regulated by the Financial Conduct Authority. The firms on the register are; check any of them on the FCA Register using the reference number shown.
- It is not official statistics. This is a private statistical series. It is not produced under the Code of Practice for Statistics, although the release calendar and revisions policy below are modelled on it.
- It is not a regulated benchmark. The UK Advice Fee Index is a statistical series about advice charges published by FinancialAdvisor.co.uk; it is not used to determine any amount payable under a financial instrument or contract and is not a benchmark within the meaning of the UK Benchmarks Regulation.
Where the numbers come from
- The FCA Financial Services Register, read through the FCA's own documented API at look-up pace. Firm names, reference numbers, authorisation status and permissions, legal status, appointed-representative links, registered address and website come from there. We never scrape fca.org.uk or the register's web interface.
- Firms' own published fee pages and documents, captured by an identified, throttled crawler that honours robots.txt and any terms prohibiting automated access. The full policy, and what the crawler will never do, is in our source-capture policy. Captured pages are kept privately as evidence and are never republished; what we publish is the extracted figures with attribution, a link and the capture date.
- Firms themselves, through a claimed page: a schedule entered by the firm, and — at the verified grade — attested by a named individual on the FCA Register.
- Labelled priors from other published sources (the FCA, NextWealth and others). These are comparators shown beside our figures. They are never presented as our panel data and never rendered in the same typographic style as a panel figure.
Record grades — never merged
A figure carries exactly one grade, and the grades are never combined into a single number.
| Grade | What it means | What it may be used for |
|---|---|---|
| Declared — published | Captured from the firm's own public page or document, with the address and the date we captured it. | Published on the firm's page with its source and date; counted in Index statistics at declared grade. |
| Declared — submitted | Entered by the firm itself through its claimed page; not yet confirmed by a named individual. | As above, labelled as submitted by the firm. |
| Verified | A named adviser on the FCA Register has confirmed within the last 12 months that the schedule is the one the firm currently discloses to clients, our plausibility checks pass, and any difference from a published disclosure is annotated. | As above, plus the annual mark, the firm's own position view and the Fair Value Comparator Report. |
| Transacted | Built from clients' own fee statements (from 2027). Not yet in operation. | Aggregate only, at national and firm-type level, and only above the published thresholds. Never for a named firm in version 1. |
Cell definitions — fixed categories only
Every statistic is computed over a cell, and the cells are a fixed list. They are fixed on purpose: with ad hoc cuts, two published tables can be subtracted from one another to reveal a cell too small to publish.
| Dimension | Categories |
|---|---|
| Firm type | Independent · Restricted · Network appointed representative |
| Region | The 12 ITL1 nations and regions of the UK. City names resolve to their region page. |
| Firm size band | Advisers: 1 · 2–5 · 6–20 · 21 or more |
| Portfolio band (for total-cost figures) | £100,000 · £250,000 · £500,000 · £1,000,000 |
A cell is published unpooled only where it holds at least 10 firms. With 5 to 9 firms the figure is a model-based estimate, labelled as one. Below 5 firms nothing is published. No pound-denominated aggregate is published where the largest firm accounts for more than 50% of the cell total or the largest two for more than 75%, and a table with 60% or more of its cells suppressed is not published at all.
How a headline figure is reported
Advice charges cluster hard on round price points. The FCA's Evaluation of the Retail Distribution Review and the Financial Advice Market Review (December 2020) found more than 80% of ongoing charges at exactly 0.50%, 0.75% or 1.00%, and 38% of one-off charges at exactly 3%. A mean with a normal-distribution interval would describe a shape the data does not have.
So every headline figure is reported as the median, with an interval, the modal price points and the share of firms at each of them, the number of firms it is computed over, and the vintage — never a mean with a Gaussian interval alone. Firm-weighted is the primary weighting; an adviser-weighted cut is shown as a secondary figure where the sample allows.
Quarter-on-quarter change is reported on the share of firms at a price point, with Wilson intervals, and only when the change exceeds what random variation would produce. Otherwise the wording is "broadly stable".
Reconciliation with the published priors
Two published figures are quoted more often than any other when people ask what UK advice costs, and neither measures the same thing this site measures. Both are shown on Index pages as labelled priors, with their own definitions beside them, and the differences are set out here rather than smoothed away. Weights and decay for each prior are published in the prior register.
| Prior | What it measures | How it differs from a FinancialAdvisor.co.uk figure |
|---|---|---|
| FCA, Evaluation of the RDR and FAMR (December 2020): 2.4% initial, 0.8% ongoing, 1.9% all-in | Charges observed across the market at that date, including the clustering finding on round price points. | Six years old at our launch and not refreshed; drawn from regulatory returns and firm samples rather than from firms' own published schedules; "all-in" includes product and platform costs that our adviser-charge figures exclude. It is a level check on a different definition, not a trend line we extend. |
| FCA retail intermediary market data (RMAR 2025, Tables 25 and 26): initial median range 1.0–3.0%, ongoing 0.5–1.0% | What firms report to the regulator, for firms charging a percentage of investment. | Reported by firms to the regulator as ranges rather than as a schedule a client would be shown, and covering all retail-investment adviser firms rather than the firms that publish a schedule. We use its firm-type, size-band and charging-structure margins to calibrate our panel, and we publish the raw and the calibrated figure side by side. |
| NextWealth, Financial Advice Business Benchmarks and the Fee Benchmarking Report (2026 edition: ongoing 83 basis points, total 180 basis points, around 545 advisers) | An adviser-supplied survey of what firms charge, asset-weighted, with a large-firm data-request component. | Adviser-weighted rather than firm-weighted; client-paid basis points rather than the adviser charge alone; a survey panel rather than firms' own published schedules. Its own series moves 63.7 → 77 → 83 basis points across editions on changing definitions, so it is modelled as three edition-specific figures with wide variance, never as one trend. NextWealth is invited to comment on this methodology before launch. |
The release calendar
The calendar is published in advance and does not move. Each quarterly Release is published at 09:30 UK time on the first Monday of the quarter's last month, from an input snapshot taken at the end of the quarter's second month. Monthly refreshes update figures in place on the first Monday of the month; they do not create new pages.
| Release | Input snapshot | Published at 09:30 UK |
|---|---|---|
| 2026-Q4 (the first) | End of November 2026 | Monday 7 December 2026 |
| 2027-Q1 | End of February 2027 | Monday 1 March 2027 |
| 2027-Q2 | End of May 2027 | Monday 7 June 2027 |
Revisions policy
- A published Release is permanent and is not regenerated. Its address never changes.
- A revision is added to the Release as a dated note. The original figure stays visible, struck through, beside the revised one.
- Nothing is silently overwritten, and no figure is removed from a published Release. Every change is a dated entry in the machine-readable changelog.
- Monthly refreshes change figures in place on the statistic, firm-type, region, scenario and firm pages, and the visible "last material change" date moves with them.
- A figure under dispute is flagged "figure under review" on every surface where it appears — page, feed and download — while we check it, and is excluded from the next statistics build.
Corrections and right of reply
Any firm can correct or dispute its own figures without an account and without accepting any agreement. Consumers can report an error too. Use the complaints and disputes form, or email [email protected].
- The figure is marked "figure under review" on every surface as soon as the dispute is logged.
- We acknowledge within 5 working days (the statutory limit for a data-protection complaint is 30 days).
- We resolve within 20 working days, inside the one-month limit for a rectification request.
- Outcomes are: corrected; annotated with the firm's own statement of up to 1,000 characters shown beside the figure; recorded as an accurate capture of what the firm published on the stated date; or, where a firm tells us its charges have changed, the captured schedule moves to history and the page says so.
- History is annotated, never deleted. We publish a monthly count of sourced facts removed under pressure, and the target for it is zero.
Who writes this, and what software does
Every editorial and statistical page on this site carries a named author. This page, and the methodology it describes, are written by Nathan Hill-Haimes, who is an independent researcher and not a financial adviser.
Software captures pages, extracts figures, checks arithmetic and lints wording against the rules on this page; it never writes copy about a named firm, never characterises a firm, and never authors the promotional text on a claimed page — that text is written and approved by the firm itself and stored with an approval record.
Each quarterly Release is reviewed by a named external reviewer before publication. Where no reviewer has been secured for a Release, the Release says "not independently reviewed" rather than implying a review took place.
Status of the statistical methods
Eight parts of this methodology are held back until a dated review with the named author has taken place. Until each is signed, no figure computed under it is published anywhere on this site. The target date for the review is Friday 25 September 2026; the state below is as at the version date of this page, and this page is reissued with a new version number whenever an item is signed.
| Method | Reviewed and signed | Consequence while unsigned |
|---|---|---|
| The estimator, how thin cells are pooled, and the weight given to priors | Not yet | No figure computed under it is published anywhere |
| How total cost of ownership is normalised so two firms can be compared | Not yet | No figure computed under it is published anywhere |
| How a captured page is matched to a firm on the FCA Register | Not yet | No figure computed under it is published anywhere |
| The provenance, lawful-basis and consent record kept against every fact | Not yet | No figure computed under it is published anywhere |
| The minimum sample sizes and suppression rules | Not yet | No figure computed under it is published anywhere |
| What may be published about a named firm at each record grade | Not yet | No figure computed under it is published anywhere |
| The measured accuracy of the software that reads a firm's fee page | Not yet | No figure computed under it is published anywhere |
| The content of the Fair Value Comparator Report | Not yet | No figure computed under it is published anywhere |
About this page: information only, from a publisher that is not authorised or regulated by the FCA, sells no advice and endorses no firm. Nothing here is a personal recommendation. See how we make money, how we rank and our source-capture policy. To correct or dispute anything on this site, use the complaints and disputes form.