FCA disclosures: delivery is not understanding
Only 6% of 132 investment disclosures met the FCA’s plain-English readability assessment. Here is a practical way to test what a reader has understood.
The Financial Conduct Authority’s September speech on investment culture returned to a finding that deserves careful reading: only 6% of the pre-sale disclosures it assessed were written in plain English. Read the FCA’s 18 September speech.
The underlying review, published on 2 July, assessed 132 documents for readability using the Flesch–Kincaid method. A separate intelligibility assessment covered 172 documents; all fell below the benchmark used by that assessment. These were assessments of text, not observations of 172 investors making decisions. The FCA’s review explains the samples and methods.
The distinction is useful for communications teams. A document can be delivered successfully while leaving its recipient unable to explain a restriction that matters to their decision.
Separate three questions
Our analysis distinguishes three kinds of evidence. Each can help a firm understand its process, but each supports a different conclusion.
- Delivery: was the intended information made available to the intended recipient through the chosen route?
- Acknowledgement: did the recipient perform an action such as opening the document or confirming receipt?
- Comprehension: can the recipient explain and apply the information relevant to the decision?
A click does not contain an explanation. Treating it as one makes the record appear more informative than it is. Equally, a successful comprehension exercise is evidence about the questions tested, not proof that a person understands every feature or that an investment is suitable for them.
Test a decision rather than asking whether the text is clear
We propose a small exercise using a fictional product and scenario, with no live investment decision attached. Choose one consequential question, such as what happens when an investor asks to withdraw money.
Give the reader the document as it would normally be presented. Then ask them to explain, in their own words, what they could request, when they might receive money, which conditions could alter that timing and where they found the answer. Do not supply the correct explanation inside the question.
Ask a follow-up that changes the circumstances. For example: “If the request is made after the stated cutoff, what changes?” The scenario should use only terms actually present in the fictional document. The aim is to see whether the reader can apply those terms, not to test unstated knowledge.
Record the answer before a member of staff helps. Keep any prompted answer separate from the initial response. Otherwise, the explanation supplied by the tester can be mistaken for understanding produced by the document.
Use misunderstandings to make a specific edit
Group the results by the question misunderstood. If readers locate the correct passage but interpret it differently, review the wording. If they do not find it, review navigation and placement. If they can repeat the sentence but cannot apply it to the scenario, examine the explanation and example.
Record the document version, test question, response and resulting edit. Repeat the same decision exercise with new readers after the change. A small exercise should be reported as a small exercise: record who participated and avoid presenting it as representative research.
This is an editorial proposal for improving communication. It does not establish compliance, suitability or a regulatory safe harbour.
Make the record describe what was actually established
For a communications lead, the immediate task is to inspect one existing confirmation screen. Does its language say “received”, “read” or “understood”? Compare that wording with the action the person actually took and the evidence retained.
That is a concrete formal communication problem. The record should make clear what was sent, which version was used, what the recipient did and what remains unknown. Precision here helps both the reader and the colleague who later needs to explain the exchange.
The same discipline applies when a client brings an outside explanation into the conversation. Our analysis of Swiss investor trust in AI proposes a way to check the claim and preserve the adviser’s response.
We checked the FCA’s 18 September speech against its 2 July review, accessed on 29 September 2026. We distinguish the 132-document readability sample from the 172-document intelligibility sample. We have not re-run either assessment or tested investors. The comprehension exercise is original editorial analysis; this record does not interpret firms’ legal duties.