A sharp fall, but on rising volume
The uphold rate for pension complaints decided by the Financial Ombudsman fell to 31.6% in the most recent 90-day window analysed by Veste, down from 47.2% in the prior 90-day window. That is a fall of 15.6 percentage points. It did not happen because fewer cases were being decided: the current window contains 285 decisions against 249 in the prior window, an increase of 36 decisions, or 14.5%. Both windows meet Veste's minimum sample threshold of 30 decisions, so this is not a comparison built on a handful of cases either side.
A fall of this size, on a rising caseload, is the kind of movement that invites a closer look at what is actually included in "pension" decisions, who the firms are, and whether the monthly figures back up the idea of a genuine shift or simply reflect ordinary month-to-month variation around a series that has been drifting downward for some time.
The numbers in context
Across the full run of data Veste holds for pension decisions, from 8 April 2013 to 15 June 2026, there have been 15,579 decisions, of which 6,599 were upheld, 1,344 partially upheld and 7,636 not upheld. That gives an overall uphold rate across the whole series of 46.7%, using Veste's definition of the uphold rate as (upheld + 0.5 × partially upheld) ÷ total. That headline figure sits well above the uphold rate for the entire Financial Ombudsman corpus Veste has analysed, which stands at 28.8% across 401,818 decisions with an outcome. Pension complaints, in other words, have historically been upheld more often than the typical Ombudsman case, even before the latest fall.
The year-on-year figures show the direction has been broadly downward for some time, not just in the last 90 days. In 2023, 2,368 pension decisions were published with an uphold rate of 60.9%. That fell to 50.5% across 1,567 decisions in 2024, then to 43.8% across 1,394 decisions in 2025, and to 38.3% so far in 2026 across 498 decisions. Each of these years meets Veste's minimum sample size. The latest 90-day uphold rate of 31.6% therefore continues a multi-year decline rather than representing an isolated event, though it is a steeper single-period movement than the year-on-year figures alone would suggest.
What the monthly series shows
Veste's monthly trend data, running from June 2024 to June 2026, gives a more granular picture and it is a noisy one. Monthly uphold rates over that period have ranged between roughly a quarter and just over a half, with no month strictly comparable to the next given fluctuating volumes. Some individual months show pronounced swings: April 2026 (72 decisions) recorded an uphold rate of 35.4%, May 2026 (92 decisions) fell to 28.8%, and June 2026 (39 decisions) recorded the lowest rate in the run at 24.4%. That June figure comes from a comparatively small monthly sample of 39, smaller than several other months in the series, which matters when judging how much weight to place on the most recent data point specifically.
By contrast, March 2025 saw 179 decisions with an uphold rate of 51.4%, and May 2025 saw 112 decisions at 53.6%, both considerably higher than the rates recorded a year later. The series does not move in a straight line. It oscillates, generally trending down since late 2024, but with individual months bucking that trend on both sides. This variability is a useful reminder that a single 90-day comparison, however large the percentage-point shift, sits within a series that has shown genuine month-to-month volatility even outside the periods being compared.
Firm mix and what it can and cannot tell us
Veste's data on individual firms handling pension complaints shows a wide spread of uphold rates, though it is important to be clear about what this reflects: outcomes on published decisions bearing a firm's recorded business name, not a verdict on that firm's overall conduct or service standards. Firms are counted by the exact business name on each decision, so subsidiaries of the same banking or insurance group are not combined, and two entities within one corporate group can appear as separate rows with different rates.
Among firms meeting Veste's minimum sample size, Aviva Life & Pensions UK Limited had 1,043 decisions with an uphold rate of 32.5%. The Prudential Assurance Company Limited had 847 decisions at 21.8%, and Scottish Widows Limited had 564 decisions at 22.3%. ReAssure Limited, which appears in several of Veste's illustrative decisions below, had 515 decisions with a higher uphold rate of 38.8%. Phoenix Life Limited recorded 480 decisions at 23.6%, Scottish Equitable Plc 407 decisions at 30.2%, and Standard Life Assurance Limited 273 decisions at 15.9%, the lowest uphold rate among the firms with sufficient sample size in this list.
One outlier deserves specific mention because of the scale of the difference. Options UK Personal Pensions LLP recorded 468 decisions with an uphold rate of 99.7%, with 466 of those decisions upheld and only one not upheld. That is a markedly different pattern to every other firm in the top-ten list and, given the size of the sample, is not attributable to statistical noise in the way that smaller-sample swings might be. Veste's data does not explain why this firm's published decisions differ so sharply from the rest of the sector, and no inference beyond the recorded figures should be drawn.
St. James's Place Wealth Management Plc had 260 decisions with an uphold rate of 27.1%, and The Royal London Mutual Insurance Society Limited had 330 decisions at 23.6%. None of the top-ten firms individually explains the latest 90-day fall on their own, since the period comparison is drawn from the whole pension category rather than any single firm, but the wide dispersion across firms, from 15.9% to 99.7%, shows that "pension" as a category masks considerable variation underneath.
What might be driving the movement
Veste's category breakdown for related complaint types gives some further context, though it should be read as description rather than explanation of the pension-specific movement. Pension transfer advice complaints, a distinct but related category, had 7,497 decisions with an uphold rate of 54%, notably higher than the pension category's own long-run rate. Investment mis-selling complaints, at 2,874 decisions, had an uphold rate of 50.2%. Account administration errors, by contrast, had 1,519 decisions at 37.6%, general financial advice complaints had 1,301 decisions at 35.8%, and other regulated complaints had 635 decisions at just 14.5%. If the current mix of pension decisions contains proportionally more administrative or service-related cases and proportionally fewer transfer-advice or mis-selling cases, that would tend to pull the blended uphold rate down, since those categories carry structurally lower uphold rates. Veste's data does not break down the latest 90-day window by these sub-categories, so this can only be noted as a plausible contributing factor rather than demonstrated directly.
Individual decisions as illustration
Four decisions published on 15 June 2026, all within Veste's dataset, illustrate the range of issues being decided rather than any statistical trend. A complaint against ReAssure Limited over delays and administrative errors in paying death benefits following a bereavement was not upheld by the Ombudsman, who found the firm's offer of £400 compensation plus £29.58 interest fair despite acknowledging service failings. A separate complaint against ReAssure over failure to locate original annuity policy documents was upheld, with £100 compensation ordered for the frustration caused by repeated administrative errors, even though the original documents themselves could not be produced. A complaint against Mattioli Woods Limited concerning a missed trading embargo notification during a pension transfer was not upheld, the Ombudsman finding that the information had reached the complainant's independent financial adviser even though it had not reached the trustee firm directly. A complaint against Scottish Equitable plc (Aegon) over inconsistent communication about pension terms was partially upheld, with £850 compensation awarded for distress caused by administrative shortcomings, though the Ombudsman found the complainant's delay in accessing benefits stemmed from an early surrender charge rather than confusion caused by the firm.
These four cases show the mix of outcomes that sit behind any single monthly or quarterly uphold rate: some upheld in full, some not upheld, one partially upheld, each turning on its own facts. They illustrate the kind of dispute typically raised, but none of them proves a wider trend on its own.
What the data does and does not show
The supplied data shows a genuine fall in the pension uphold rate between two 90-day windows, on rising volume, sitting within a broader multi-year decline visible in the year-on-year figures since 2023. It does not, on its own, establish why that fall occurred. The monthly series shows enough volatility, including the smaller sample size of 39 decisions in the most recent month, that some caution is warranted before treating the 90-day figure as a permanent new baseline rather than the latest point in a fluctuating series. The firm-level data shows wide variation in uphold rates across the sector, which is a feature of the published-decisions population generally and not evidence of a change in any one firm's practices during the period in question.
It is also worth restating that these figures describe published Ombudsman decisions, which are a small and specific subset of all complaints made to pension providers, not a full record of complaint volumes or resolution rates industry-wide.
The practical implication
For anyone tracking pension complaint outcomes, the useful takeaway is not that pension providers are suddenly winning more disputes, but that the category-wide uphold rate has been on a downward path for at least three full years, from 60.9% in 2023 to 38.3% so far in 2026, and the latest quarter continues that direction more steeply than the annual figures alone would predict. Whether the steeper recent movement persists, moderates, or reverses will only become clear as further decisions are published and added to the series.