UK investment-platform complaints rose 45.6% as uphold rates diverged sharply

Complaints classified by the Financial Conduct Authority as relating to investment platforms rose from 8,845 in the second half of 2025 to 12,874 in the first half of 2026. The 45.6% increase was substantially faster than the 6.3% rise in complaints across the wider financial sector. Yet the national total conceals large differences in what happened after a platform received a complaint.[1][3]

FXBrokers.ai reconstructed the investment-provision disclosures of five prominent platform operators in the FCA’s firm-level table. Among complaints closed during their reporting periods, the proportion upheld ranged from 6.7% at Trading 212 UK to 44.5% at Vanguard Asset Management. Hargreaves Lansdown reported 43.0%, AJ Bell Securities 43.7% and Freetrade 41.4%. The figures do not prove that one platform treated customers better than another: product mix, complaint type, customer profile and reporting period all matter. They do show that a customer reaching the end of a complaint process encountered very different published outcomes.[2][3]

The national increase was real, but the rate did not move in the same way

Platform complaint volumes were 6,405 in the second half of 2024, 7,980 in the first half of 2025, 8,845 in the second half and 12,874 in the first half of 2026. That is a two-year sequence of increases, culminating in a 61.3% year-on-year rise. The FCA’s aggregate contextualised rate for platforms was nevertheless displayed as 0.1 complaints per 1,000 relevant units in the latest period, compared with 0.2 in earlier periods.[1][3]

There is no necessary contradiction. A complaint count can rise while a rate falls if the customer-account denominator grows faster. The FCA also rounds category-level rates, so a movement within a displayed tenth may be invisible. In addition, the aggregate platform category and an individual firm’s investment-provision line are not interchangeable. The national figures establish volume; the firm-level records are needed to understand the customer populations and complaint outcomes behind it.[2][3]

Five firms, five different complaint profiles

Trading 212 UK reported 2,251 investment-provision complaints opened and 2,160 closed. Its contextualised rate was 2.79 opened complaints per 1,000 client accounts. Hargreaves Lansdown reported 8,328 opened and 7,147 closed, with a rate of 2.19 per 1,000. Vanguard reported 1,899 opened and 1,786 closed, at 2.04 per 1,000. AJ Bell Securities reported 1,212 opened and 1,178 closed, at 1.94 per 1,000. Freetrade reported 489 opened and 384 closed, at 1.05 per 1,000.[2]

The raw count naturally places a large incumbent such as Hargreaves Lansdown above a smaller operator. The contextualised rate reverses part of that size effect: Trading 212’s investment-provision rate was the highest of the five even though Hargreaves Lansdown received far more complaints. Freetrade’s rate was the lowest. Neither measure alone establishes service quality. Counts describe workload and customer impact in absolute terms; rates describe frequency relative to the FCA’s reported account base.[2][3]

An approximate account-base reconstruction

Dividing opened complaints by the published rate and multiplying by 1,000 produces an approximate denominator for each firm. On that basis, Trading 212’s disclosure implies about 807,000 relevant investment accounts, Freetrade about 466,000, Hargreaves Lansdown about 3.80 million, AJ Bell about 625,000 and Vanguard about 931,000. These are analytical estimates from rounded rates, not company-reported active-client totals.[2][3]

The denominator should be handled cautiously. The FCA contextualises investment-provision complaints using accounts at the reporting date, while firms can define product and legal-entity boundaries differently within the prescribed categories. Dormant and active relationships may not have the same commercial importance. A firm can also operate pension, banking or intermediation activities that sit on separate lines. The estimates are therefore useful for checking order of magnitude, not for valuing customers or ranking platform engagement.[3]

Closure speed diverged by more than 70 percentage points

Trading 212 reported closing 71.9% of investment-provision complaints within three days. At the other end, Freetrade reported 1.3%. Hargreaves Lansdown reported 23.4%, Vanguard 22.8% and AJ Bell 33.2%. The gap between Trading 212 and Freetrade was 70.6 percentage points.[2]

A fast closure is not automatically a good resolution, and a slower closure is not automatically poor handling. Some cases can be corrected immediately; others require transaction reconstruction, suitability review, fraud investigation, third-party evidence or legal analysis. The three-day statistic measures elapsed process time, not the quality of the answer. Its value lies in showing how differently complaint inventories are structured and handled.[2][3]

Freetrade’s next line reinforces that point: 96.6% of its closed investment complaints were resolved between three days and eight weeks. Hargreaves Lansdown reported 75.5% in that band, Vanguard 68.3%, AJ Bell 66.7% and Trading 212 28.1%. Across all five, only a small residual sat beyond eight weeks on the published percentages. The operational contrast is therefore primarily between immediate closures and cases worked through during the statutory eight-week window.[2]

The uphold gap is harder to dismiss as timing alone

Uphold rates measure the proportion of closed complaints where the firm accepted the complaint, at least in part, or offered redress or remedial action under the FCA’s reporting definitions. Trading 212’s 6.7% investment-provision rate was far below the other four. Freetrade, Hargreaves Lansdown, AJ Bell and Vanguard all sat in a relatively narrow 41.4% to 44.5% band.[2][3]

Applying those rates to closed complaint counts gives an approximate scale. Trading 212 upheld about 145 investment complaints, Freetrade about 159, Hargreaves Lansdown about 3,074, AJ Bell about 515 and Vanguard about 795. The calculations multiply rounded percentages by closed volumes, so they should not be treated as exact case counts. They nevertheless show that the divergence is not generated solely by a tiny sample.[2]

Several innocent explanations remain available. Trading 212 may receive a different mix of complaints, including a greater share that falls outside the firm’s responsibility or concerns market movements rather than execution. Its immediate-closure process may reject or resolve simple cases quickly. Other firms may use a broader threshold for partial uphold or remediation. The FCA table does not publish complaint narratives, issue codes at firm level or redress amounts by product, so those hypotheses cannot be tested from this dataset alone.[2][3]

Opened and closed complaints are not the same cohort

Another common analytical error is to divide upheld complaints by the number opened in the same six months. FCA percentages relate to complaints closed in the period. Some were opened earlier; some opened near the period end remained unresolved. The five firms collectively opened more investment-provision complaints than they closed. That does not by itself indicate a backlog problem, because the table does not provide the opening inventory or age profile.[2][3]

At Hargreaves Lansdown, 8,328 investment complaints were opened and 7,147 closed, a difference of 1,181. Trading 212’s difference was 91, Vanguard’s 113, AJ Bell’s 34 and Freetrade’s 105. These gaps are useful prompts for the next reporting period. If they persist, the stock of open cases may be growing. A single period cannot establish that conclusion.[2]

AJ Bell’s reporting dates require a separate warning

The table lists AJ Bell Securities for the six months from 1 October 2025 to 31 March 2026, whereas several other firms report the calendar half-year to 30 June. Its figures therefore overlap a different market and operational period. The comparison remains informative because the FCA uses the same definitions, but it is not a simultaneous snapshot. Any event concentrated in April, May or June would affect the other firms and not AJ Bell’s disclosure.[2][3]

Corporate structures also matter. A brand can operate through more than one authorised legal entity, and a financial group can submit several complaint returns. This analysis uses the named entities and the investment-provision category shown in the table. It does not combine affiliates, pension administrators, banking entities or overseas operations unless they are included in that return.[2][3]

What the platform increase means for customers

At national level, 12,874 platform complaints represent thousands of customers who believed something had gone wrong badly enough to enter a formal process. The increase can reflect service problems, but it can also reflect account growth, market volatility, fraud attempts, transfers, tax documentation, corporate actions or greater awareness of complaint rights. Without issue-level data, the aggregate does not identify a single industry failure.[1][3]

The firm-level differences matter because complaint handling is part of a platform’s infrastructure. A platform can offer low commissions and a polished interface while still imposing costs through delayed transfers, unresolved records or poor communication. Conversely, a high complaint rate may coexist with rapid remediation and a willingness to uphold legitimate cases. The data should be read as an operational risk signal, not a verdict.[1][2]

A better disclosure would connect cause, outcome and redress

The present system separates several pieces of the story. The FCA publishes aggregate product causes and outcomes, and firm tables publish rates, speed and uphold percentages. It does not publicly connect a specific firm’s issue type with its outcome and redress amount. That prevents outsiders from answering whether a platform’s upheld cases arise mainly from transfers, administration, execution, fraud controls or another mechanism.[1][2][3]

A more revealing return would disclose, for each product line, complaint causes, median age, reopened cases and redress distribution, while protecting customer identities. That would allow customers and regulators to distinguish a platform that promptly fixes administrative errors from one whose unresolved cases cluster around a recurring control weakness. The current figures identify where to ask; they rarely explain why.[1][3]

Combining frequency and outcome changes the comparison

Complaint frequency and uphold rate answer different questions. Multiplying the published complaint rate by the share of opened cases closed and then by the uphold percentage produces a rough upheld-complaint rate per 1,000 relevant accounts. On that basis, Trading 212 was about 0.18, Freetrade 0.34, Hargreaves Lansdown 0.81, AJ Bell 0.82 and Vanguard 0.85.[2][3]

This composite is an investigative estimate, not an FCA metric. It combines rounded fields and assumes the opened and closed populations are sufficiently comparable for a directional calculation. They are not identical cohorts. Its purpose is to prevent a low raw complaint rate or low uphold rate being read in isolation. Hargreaves Lansdown, AJ Bell and Vanguard produced similar reconstructed upheld rates despite different complaint frequencies and volumes.[2][3]

Trading 212 remained the outlier after that adjustment. Its higher complaint frequency was offset by very fast closure and a low uphold percentage. The table cannot show whether that pattern reflects more non-meritorious complaints, a narrower interpretation of responsibility, different customer behaviour or a materially different product mix. Those explanations require case-level evidence.[2]

The FCA publication threshold shapes what the public sees

Firm-specific disclosure is not a census of every authorised company. The FCA publishes data for firms that report at least 500 complaints in a six-month period or 1,000 over a year, alongside certain firms that meet related criteria. The regulator says the disclosed firms account for about 98% of complaints. Smaller platforms can therefore be absent even when their rate is high, simply because their absolute volume remains below the threshold.[2][3]

That selection rule favours meaningful national coverage but creates survivorship and scale effects. The published list is dominated by firms large enough to cross the volume threshold. It should not be used to claim that an unlisted platform had no complaints or performed better. Nor should a listed firm be penalised merely for being large enough to appear.[2][3]

Group structures can also fragment disclosure. A customer may know a consumer brand while the complaint return belongs to an authorised subsidiary. Acquisitions can move a platform into a different group without immediately making historic periods comparable. The legal entity, reporting dates and product line must be preserved in any trend analysis.[2][3]

Redress is the missing materiality measure

The national H1 data reported about £240 million of redress across financial complaints, but the sortable firm table does not provide a platform-specific redress amount for each investment-provision line. An upheld complaint can involve a corrected statement or apology, while another can involve substantial financial loss. Treating every uphold as economically equal would obscure the customer consequence.[1][2]

A low uphold rate paired with high redress per upheld case could be more material than a high uphold rate dominated by administrative corrections. The current public fields cannot test that possibility. They also do not show how many customers accepted the firm’s outcome, escalated to the Financial Ombudsman Service or later received a different decision.[1][2][3]

For customers choosing a platform, the most useful next disclosure would connect complaint cause, elapsed time, uphold outcome and redress amount. Until then, the present figures are best used to identify operational patterns and questions, not to award a definitive service ranking.[1][2][3]

The adversarial check rules out a service-quality league table

The data cannot establish that the firm with the lowest uphold rate provides the best service or that the firm with the highest rate provides the worst. A business that recognises and remedies errors readily can report a high uphold percentage. A business receiving many complaints outside its responsibility can report a low one. Case mix is the missing variable.[2][3]

Nor does closure speed settle the issue. Rapid rejection, rapid correction and rapid explanation all appear in the same timing field. The comparison is publishable because it identifies a large, documented divergence and the questions it creates—not because the table supplies a final causal answer.[2][3]

A cohort view would turn the signal into an explanation

The current table follows firms, not complaint cohorts. A cohort disclosure could track cases opened in a particular month through closure, uphold, redress and ombudsman escalation. That would eliminate the mismatch between opened and closed populations and reveal whether slow cases eventually produce different outcomes.[2][3]

Without that evidence, the strongest conclusion remains comparative rather than causal. The platform sector’s complaint workload rose sharply, and the five firms reported materially different processing and outcome patterns. The reasons remain unproven.[1][2]

The next six months will test whether the jump was temporary

Three indicators deserve attention. The first is the national platform count: another large increase would make the H1 jump harder to attribute to timing. The second is the gap between opened and closed complaints at individual firms. The third is the persistence of the uphold divergence, especially Trading 212’s low rate relative to peers.[1][2]

For now, the defensible conclusion is limited but important. Platform complaints increased 45.6% in six months. Among five large operators, complaint frequency ranged from 1.05 to 2.79 per 1,000 investment accounts, immediate closure ranged from 1.3% to 71.9%, and uphold rates ranged from 6.7% to 44.5%. Those differences do not prove misconduct or poor service. They reveal that “a platform complaint” can lead to markedly different processes and outcomes depending on the firm receiving it.[1][2][3]

Report an error or submit an update