A piece of adviser research published this week is, on the face of it, about platform software. Financial Software Limited asked 178 advice firms what they want from the tax packs their platforms produce, and 94.1% said they want capital gains tax reports included in the same routine way that dividend and interest figures already are. Money Marketing, reporting the same research, added the number that actually matters: around 60% of advisers say they are not confident the platforms they use accurately reflect their clients’ CGT positions.
Read that from an accountant’s chair and it stops being a software story. If the regulated firm closest to the investments does not trust the capital gains data coming out of the platform, then the person who eventually has to put a number on a self-assessment return is relying on figures that the professional nearest to them has already flagged as unreliable — and is doing so eight months later, with less context and no access to the underlying trade history. The gain is computed by the accountant. The evidence for it lives somewhere else entirely.
What the research actually found
Three findings are worth separating out, because they point in slightly different directions.
- Demand for the data is close to unanimous. 94.1% of the 178 advice firms surveyed want CGT reporting in platform tax packs (Financial Software Limited, reported by Professional Adviser, 9 September 2026).
- Confidence in the current data is not. Roughly 60% of advisers say they are not confident their platforms accurately reflect clients’ CGT positions (Financial Software Limited, reported by Money Marketing, 8 September 2026).
- It is now a buying criterion. CGT calculators are the fourth most-selected “must-have” feature in advisers’ platform due diligence, up two places year on year, behind only the SIPP and whole-of-market fund range features.
The context sits alongside it. The lang cat’s latest State of the Advice Nation research found 39% of advice firms now cite tax planning as a priority, with the same proportion reporting a sharp rise in clients seeking reassurance about tax and legislative uncertainty. Michael Edwards, managing director of Financial Software Limited, put the cause plainly: firms are contacting the company directly because they are “frustrated with CGT reporting across platforms, viewing it as either lacking or inconsistent”.
Greg Moss, a chartered financial planner and founder of Eleven.2 Financial Planning, described the practical effect in the same coverage: the information needed to report accurately “already sits with the platforms”, and piecing it together takes hours of manual reconciliation.
Why this lands on the accountant’s desk
Advisers are complaining about reconciliation work. Accountants inherit the consequence of it.
The division of labour around a client’s investments is rarely designed; it just settles. The planner decides what to sell and when, for reasons that are about the portfolio rather than the return: rebalancing after a run in one holding, raising cash for a withdrawal, using an annual exempt amount before 5 April. The accountant meets that decision later, as a line on a statement, and has to turn it into a computation with a date, a base cost and a matching rule attached.
The gap between those two moments is where the data degrades. By the time a January deadline is close, the client has forgotten which units were sold, the adviser has moved on to the current year, and the platform report — the one 60% of advisers do not fully trust — becomes the default source of truth by exhaustion rather than by choice.
This is the same structural point we made about the CGT signal in a client’s file, seen from the other end. There the question was who notices the exposure. Here it is who can evidence it.
The parts of a gain a platform report often will not carry
Platform CGT reporting is not wrong so much as partial. A platform can see what happened on its own system with complete accuracy. What it frequently cannot see includes:
- Reinvested income on accumulation units. The income is retained in the fund and taxed as income when it arises, so it must be added to allowable base cost. A report showing only the original cash invested overstates the gain.
- Equalisation on the first distribution after a purchase. That element is a return of capital and reduces base cost, not income.
- Holdings of the same fund on another platform. Share matching rules operate across the client’s whole holding, not per provider, so a same-day or 30-day repurchase elsewhere changes what the disposal is matched against.
- In-specie transfers that arrived without acquisition history. Units moved in from a previous provider commonly land with a transfer date and no original cost.
- Share class switches and fund mergers, some of which are disposals and some of which are not.
None of these are edge cases in a portfolio of any age. Each of them moves the figure on the return.
Putting real numbers on it
Consider a client with a general investment account holding accumulation units bought eight years ago for £40,000, sold this year for £49,400. The platform tax pack reports a gain of £9,400 — proceeds less the cash originally invested, which is exactly what the platform can see.
Over those eight years the fund retained and reinvested £5,200 of income, on which the client was taxed as it arose. Added to base cost, the allowable cost becomes £45,200 and the real gain is £4,200.
For 2026 to 2027 the annual exempt amount is £3,000. On the platform figure, £6,400 is taxable; on the corrected figure, £1,200 is. Assume a higher-rate taxpayer, so the 24% rate applies:
- Platform figure: £6,400 × 24% = £1,536
- Corrected figure: £1,200 × 24% = £288
A difference of £1,248 on one holding, arising entirely from base cost the platform had no way of knowing. Nothing here is aggressive, and nothing is a planning technique. It is the arithmetic being done with the whole of the information rather than the part of it that was easiest to obtain.
The figures used above are the published ones: the annual exempt amount is £3,000 (£1,500 for trusts), and gains from 6 April 2026 are taxed at 18% within the basic rate band and 24% above it, with the basic rate band at £37,700 for 2026 to 2027 (GOV.UK, Capital Gains Tax rates and allowances).
The platform knows what it sold. The adviser knows why. The accountant has to defend the number. Those three facts sit in three different places, and nobody owns the join.
What a joined-up conversation looks like here
This is a good example of a problem that neither profession can fix alone, and that both are currently absorbing as unbilled time.
Where an accountant and a client’s financial planner have a working relationship, the fix is unremarkable: the disposal history is requested once, in the spring, at transaction level, and the two firms agree which of them holds the base cost record. The planner already has the trade rationale and the platform access. The accountant knows what a computation has to survive if HMRC asks. Between them the reconciliation happens once rather than twice, and in April rather than January.
Where there is no such relationship — and for most owner-managed clients there is not — the same work still gets done, but twice, badly, by two firms who never speak, on either side of a client who is guessing. That is the ordinary cost of the advice gap, and it is worth being clear that it is a cost borne by professionals as much as by clients. The lang cat’s State of Advice Report 2025 found just 9% of UK adults had paid for financial advice in the preceding two years, while 91% of those who did take advice found it valuable. Where the client has no adviser at all, the acquisition history has no custodian either.
The boundary between the two roles is worth keeping firm while the data flows across it: the accountant computes and files, and the regulated firm advises on what to hold and when to sell. We set that division out in who does what. Sharing a disposal history does not blur it.
Two things worth doing this week
First, pick three clients with general investment accounts and check whether you hold acquisition costs at all. Not the current valuation — the original cost, the dates, and whether the units accumulate or distribute. If the answer for any of them is “we take the platform’s gain figure”, that is the exposure in miniature, and it is quicker to resolve now than in the second week of January.
Second, put one line in your standard spring information request: a transaction-level disposal history for the tax year, including acquisition dates and costs, share class switches, and confirmation of whether reinvested income has been included in base cost. Where the client has an adviser, send it to them as well as to the client. It takes the adviser minutes while the year is fresh, and it makes your firm the one that noticed.
For the wider set of things a client file already shows you, the signals sitting in the accounts covers the same discipline applied to the year-end pack.
What is still uncertain
Three things are genuinely unresolved, and it is worth being honest about which.
The FSL research measures what advisers want, not what platforms will build. There is no industry standard for CGT reporting in tax packs and no regulatory requirement to produce one, so improvement will arrive platform by platform, at whatever pace commercial pressure sets. The finding that CGT calculators have risen to fourth in platform due diligence is the mechanism by which that pressure gets applied, but it is a slow one.
The direction of CGT policy itself is not settled either. The Budget has not been held, and speculation about capital gains is running well ahead of anything announced. Nothing in this article depends on a rate change, and no client decision should be taken on the assumption of one — the point stands whatever the rate turns out to be, because it is about the base cost rather than the percentage applied to it.
Finally, the split of responsibility between platform, adviser and accountant for the accuracy of a capital gains computation has never been formally allocated. In practice it falls to whoever signs the return. Until that changes, the practical answer is to gather the data early and from the party that holds it.
Common questions
Is the platform’s CGT report good enough to file a return from?
Treat it as a starting point rather than a computation. A platform sees the transactions that happened on its own system, which means it can generally report proceeds and purchase costs accurately. What it often cannot see is everything that changes the base cost or the matching: units of the same fund held elsewhere, an in-specie transfer that arrived without its original acquisition history, reinvested income on accumulation units, equalisation on the first distribution after a purchase. None of those are exotic. All of them move the number you put on the return. The safe working assumption is that the report tells you what happened on that platform, and you remain responsible for whether that is the whole disposal.
Why does reinvested income on accumulation units change the gain?
Because the client has already been taxed on it. With accumulation units the fund’s income is not paid out; it is retained and reflected in the unit price. That notional distribution is still taxable income in the year it arises, so if it were not added to the acquisition cost the same money would be taxed twice, once as income and again as a gain. Over a long holding period the accumulated total can be substantial. A platform report that shows only the cash originally invested will therefore overstate the gain, sometimes considerably. The correction is not a judgement call or a planning choice — it is simply the right base cost.
Whose job is the CGT computation, the accountant’s or the planner’s?
The computation on the tax return is the accountant’s. The disposal decision that created it is usually the planner’s, taken months earlier and for reasons that have nothing to do with the filing deadline: rebalancing a portfolio, funding a withdrawal, using an annual exempt amount before the tax year closes. That split is exactly why the data goes missing. The planner records why the trade happened; the accountant needs what it cost and when it was bought. Neither party is doing anything wrong. The information simply has to survive a journey between two firms that often have no routine reason to speak to each other.
What should we ask a client’s adviser for, and when?
Ask in the spring, not in January. What you want is a transaction-level disposal history for the tax year just ended: dates, units, proceeds, original acquisition dates and costs, any switches between share classes, and any holdings of the same fund on other platforms. Ask specifically whether accumulation units are involved and whether reinvested income has been added to base cost, because that is the single most common gap. Getting this in April or May, while the adviser still has the year fresh, is far quicker than reconstructing it eight months later from statements. It also tells the adviser their own reporting has a second reader.
Does this only matter for wealthy clients?
No, and that is the shift. The annual exempt amount is £3,000 for 2026 to 2027, and £1,500 for trusts. At that level a modest general investment account can produce a reportable gain from an ordinary rebalance, so clients who never previously appeared on a capital gains page now do. The people affected are frequently not investment-sophisticated: a business owner with a legacy portfolio, someone who inherited units, a client who moved cash off deposit. They are the least likely to have the acquisition records, and the most likely to assume the platform statement is the answer.