A guest post by Abdelhamid Taha, a financial regulation specialist with over a decade of operational experience in AML and compliance at HSBC, including during the bank’s US Deferred Prosecution Agreement period. He is a Member of the Secretariat Committee of the All-Party Parliamentary Group on Investment Fraud & Fairer Financial Services. He writes in a personal capacity.
Key Takeaways
MFS exposed the risks of fragmented oversight: auditors, banks, the Security Agent and regulators had separate mandates covering different parts of the structure.
More assurance did not mean end-to-end verification: multiple audits and professional gatekeepers still left gaps between entities and mandates.
Barclays had early transactional visibility, but its role and incentives were different from those of other lenders and gatekeepers.
The FCA’s remit was narrow: Annex 1 supervision covered anti-money-laundering compliance, not broader prudential or conduct risks.
The core question is structural: who, in a complex financing chain, is responsible for checking the whole?
Editor’s note: The Financial Reporting Council’s investigations relating to Market Financial Solutions remain open. The FRC has stated that opening an investigation does not indicate that it has made, or will make, any finding of breach or misconduct. This article draws on Companies House filings, administrators’ proposals, FCA statements, FRC materials, Bank of England records and FOI correspondence, and Financial Times, Telegraph, Bloomberg and other reporting. A full account of the MFS collapse and the gatekeepers surrounding it can be found in Part 1.
This essay is part of a series of 3 essays:
Part 1 — £1.8 billion and borrowed trust: the MFS structure no one was mandated to see whole
Part 2 — £1.8 billion and borrowed trust: the limits of institutional oversight
Part 2: The system around MFS
What came before
Market Financial Solutions was a Mayfair bridging lender that collapsed in February 2026. Administrators would later estimate the total group deficiency at approximately £1.8 billion in proposals filed through Companies House.
Its founder, Paresh Raja, controlled a structure fragmented across separate legal entities and professional mandates: one auditor for MFS, another for major special-purpose vehicles (SPVs), a Security Agent exempt from a standalone statutory audit, a bank with direct visibility over transactions passing through the accounts it operated, and a regulator whose supervisory remit over MFS was limited to anti-money-laundering compliance.
Raja’s spokesperson has denied the fraud and dishonesty allegations, stating that funds administrators characterise as missing were held through nominee structures for the benefit of MFS and its creditors.
The question now is not whether these were bad institutions or bad people. It is whether the institutional architecture was capable of identifying a problem that crossed the boundaries between them. The answer requires following the chain of borrowed trust across auditors, lenders, the Security Agent, banks and regulators — and ultimately to the Bank of England.
The auditor’s first duty
An auditor is not a box-ticker. The FRC’s description of an auditor’s responsibilities requires professional judgement, professional scepticism, an assessment of the risks of material misstatement and sufficient appropriate audit evidence.
The MFS collapse therefore raises a basic question about the capacity to perform those responsibilities.
Berkeley Finch Limited had eleven employees and one chartered accountant, according to Financial Times reporting. Its audit fee for MFS’s 2024 accounts was £72,375. MFS’s filed accounts reported a group loan book of approximately £2.4 billion. Elsewhere within the wider Raja-controlled structure, Zircon Bridging and Amber Bridging had loan books of £458 million and £579 million respectively.
The scale disparity does not, by itself, prove that Berkeley Finch lacked the resources to perform the audit. Firm headcount and audit fee cannot tell us how many hours were spent on an engagement, what specialists were used, or what evidence was obtained.
But they do raise an obvious question: did the firm have sufficient resources, expertise and time to perform an audit of this complexity?
The FRC’s auditing standards impose quality-management and engagement responsibilities concerning appropriate resources. Whether those requirements were met is now a matter for the FRC, which has opened a formal investigation into Berkeley Finch’s audit of MFS’s 2024 financial statements.
The same distinction matters when considering Silver Levene (UK) Limited, which audited Zircon and Amber and also audited Zircon Group’s consolidated accounts. The SPVs reported loan books of £458 million and £579 million respectively and received unqualified audit opinions.
Their publicly filed audit reports identified a fraud risk as “posting inappropriate journal entries”. The public reports do not identify servicer-level cash diversion as a specific fraud risk, although the audit working papers are not public and may contain analysis that does not appear in the audit reports.
Under the version of ISA (UK) 240 applicable to these audits, fraud-risk assessment also interacts with the requirements concerning related parties. MFS was servicing loans held by entities controlled within the same wider ownership structure.
The FRC has separately opened an investigation into Silver Levene’s audit of Zircon Group’s 2024 consolidated financial statements.
This leaves two distinct questions. If the FRC identifies breaches of professional requirements, audit execution will form part of the explanation. Regardless of the enforcement outcome, a separate policy question remains: whether the scope and standards of audit are sufficient to identify risks that cross legal-entity boundaries.
That question cannot be answered before the investigations are complete. But MFS has already exposed why it needs to be asked.
The chain of borrowed trust
The MFS structure contained multiple layers of formal assurance and institutional involvement.
There were audited financial statements. Registered security. Institutional lenders. A professional Security Agent and lawyers. A major transactional bank. FCA registration for anti-money-laundering purposes.
None, on its own, amounted to verification of the entire financing structure.
This is the borrowed-trust problem.
The financing arrangements sat alongside audited accounts and formal security structures. The public record does not establish exactly what each lender reviewed or relied upon before providing funding, or how far individual lenders independently tested the underlying assumptions behind those forms of assurance: the loan books, the operation of the servicing arrangements, the destination of cash and the exclusivity of the collateral.
The Security Agent, Intertrust Trustees Limited, held security over the Zircon and Amber assets. As Part 1 detailed, it relied on the statutory subsidiary audit exemption rather than having a standalone statutory audit. Its constitutional documents included Article 23, which allowed holders of a majority of the ordinary shares to appoint or remove directors by written notice, and in March 2023 it was briefly subject to compulsory strike-off action before that action was discontinued.
Those facts do not establish that Intertrust Trustees failed in its MFS responsibilities. Nor do regulatory findings involving other Intertrust entities in other jurisdictions establish failings by the UK Security Agent.
They do, however, illustrate the distinction between a recognised professional-services brand and the particular legal entity carrying out a particular function. The MFS case raises the question of how closely sophisticated counterparties examined the latter rather than relying on the assurance conveyed by the former.
MFS’s own governance arrangements also concentrated control. A 2019 amendment to its Articles of Association, available through its Companies House filing history, permitted a director with a personal conflict of interest to be counted in the quorum and to vote on matters in which that director had an interest.
Then there was Barclays.
Barclays was both a lender to the wider MFS network and MFS’s principal transactional bank. That combination gave it direct visibility over transactions through the accounts it operated as well as a substantial financial exposure of its own.
By November 2025, Barclays had begun restricting MFS-linked payments. The Financial Times reported that the bank began blocking transactions months before the collapse. Bloomberg later reported that a company tied to MFS borrowed approximately £143 million from Wells Fargo in late November 2025, with Barclays repaid a similar amount after Barclays had already begun restricting MFS-linked transactions.
There is no public evidence that the information prompting Barclays’ restrictions was shared across the wider lending and assurance chain before the collapse.
That does not establish that Barclays had a legal obligation to share it. Nor does the public record establish every contractual duty operating between the different lenders and facilities.
It does reveal a structural tension: the institution with some of the earliest documented visibility into the payment problems also had a direct commercial incentive to protect its own exposure.
The FCA occupied another narrow part of the chain. Its March 2026 investigation statement emphasised that MFS was registered and supervised solely for compliance with the Money Laundering Regulations and was not subject to wider FCA regulation.
Even within that limited remit, MFS had already undergone scrutiny. A 2024 skilled-person review commissioned by the FCA and carried out by DWF found MFS operating in line with UK money-laundering requirements, while recommending some potential improvements.
None of those individual facts proves that an institution simply “trusted the one before it”. Internal due diligence by lenders, banks, trustees and auditors is not fully visible from the public record.
The structural point is narrower and stronger: multiple institutions and forms of assurance were present, yet collectively they did not prevent the enormous documentation gaps, alleged diversion of funds and alleged double-pledging that administrators later reported.
That is borrowed trust.
The price of borrowed trust
The audit fee for MFS’s 2024 accounts was £72,375.
The first months of reconstructing what happened after the collapse cost millions.
Kirkland & Ellis alone accumulated about £4.4 million in legal fees during the first eight weeks after the business failed, according to the Financial Times.
The comparison is not like-for-like. An audit and an insolvency investigation involving litigation and worldwide asset recovery are entirely different professional engagements.
But in this case, the contrast shows that the early post-collapse reconstruction and litigation costs quickly exceeded the disclosed annual audit fee many times over.
Those costs are not a criticism of the administrators. They are pursuing assets, reconstructing transactions and litigating claims on behalf of insolvent estates.
Nor can the cost currently be assigned to any particular gatekeeper as a matter of culpability. Those questions remain unresolved.
The immediate economic reality is simpler: the estates incur the costs, reducing the pool ultimately available to creditors.
Borrowed trust becomes expensive when it has to be reconstructed after collapse.
The Bank of England and the collateral data question
The borrowed-trust question also reaches the infrastructure surrounding the financial system — but the Bank of England evidence needs to be understood carefully.
The Bank accepts a wide range of assets as collateral under its Sterling Monetary Framework, including Level C securities and loan collateral.
This does not mean that the Bank accepts collateral simply on the representations of counterparties.
Its published framework describes substantial due diligence. For loan collateral, the Bank’s pre-positioning guidance provides for reviews of lending practices and controls, due-diligence questionnaires, legal reviews and data audits. The Bank assesses risk and applies bespoke collateral haircuts; further data audits can also be required as collateral pools change.
The relevant point from the FOI correspondence is narrower.
On 27 July 2026, the Bank responded to a Freedom of Information request asking for the total value of eligible Level C collateral where Intertrust/CSC or Maples was recorded as Corporate Services Provider. The response, reference CAS-028671, has been reviewed by Big4News but is not currently published on the Bank’s FOI disclosure log.
The Bank said:
“Following reasonable searches, we can confirm that the Bank does not hold information within scope of your request.”
The Bank’s published eligible-collateral material provides access to its Level C securities dataset.
The public Level C dataset does not itself identify Corporate Services Providers. It therefore cannot, on its own, support a provider-by-provider mapping of Level C securities to Intertrust/CSC or Maples.
In any event, the role of a Corporate Services Provider in one securitisation should not be equated with the role performed by Intertrust Trustees Limited as Security Agent for Zircon and Amber. They are different functions and, potentially, different legal entities within wider groups.
The significance of the FOI response is therefore not that the Bank failed to conduct due diligence on its collateral. Its published procedures demonstrate otherwise.
It is a data-aggregation point.
The Bank said it did not hold information within scope of the request.
The response does not establish that the Bank lacks other means of monitoring concentrations or third-party dependencies.
That raises a narrower systemic question: to what extent can concentrations in the third-party service providers supporting financial structures be seen and assessed across a collateral portfolio, rather than transaction by transaction?
The MFS collapse does not answer that question.
It provides a reason to ask it.
The FCA’s increased scrutiny
The chain returns to the FCA’s statement of 7 August 2026.
It did not mention Market Financial Solutions by name.
It did, however, identify some of the same categories of risk that had become prominent in the discussion surrounding MFS.
The FCA said it had seen Annex 1 firms relying too heavily on the financial-crime controls of parent companies. It warned of risks from unregulated lending carried out through complex structures, including SPVs. It said each firm had to tailor its controls to its own governance, operations and risks.
And it announced two concrete actions.
The FCA would subject new Annex 1 registration applications to closer scrutiny.
It had also sent information requests to around 900 existing Annex 1 firms. That followed work with another 300 firms in late 2025 and, according to the FCA, meant it had contacted all registered Annex 1 firms.
So the FCA did not simply tighten the front door and ignore the existing population.
But the statement also illustrates the distinction between supervisory intensity and the regulatory perimeter.
The FCA announced more scrutiny of the AML risks it is responsible for supervising. It did not announce that Annex 1 lenders would become subject to prudential regulation, wider conduct supervision or an intermediate regulatory category as they increased in size.
Nor did the statement explain whether the MFS experience had prompted consideration of such a change.
The regulator has therefore responded to the Annex 1 risks it identified without yet changing the fundamental boundary that Part 1 identified.
That is the important policy question.
Not whether the FCA has done nothing. It plainly has not.
But whether more intensive supervision inside the existing perimeter is enough.
What happens now
The FCA’s enforcement investigation concerns MFS’s compliance with the Money Laundering Regulations.
That is separate from the administrators’ civil allegations concerning diversion of funds and double-pledging.
The FRC’s four MFS-related investigations concern the conduct of individual accountants and firms and compliance with applicable professional requirements. They are not a policy review of whether the audit framework itself should be redesigned.
Meanwhile, the administrators continue to pursue Paresh Raja and Barclays.
MFS Limited has sued Barclays for access to frozen bank accounts. Barclays held around £160 million across accounts tied to various MFS entities at the time of the collapse, although the amount belonging to MFS Limited itself is unclear. Barclays has asserted that it is entitled to set funds off against losses and is defending the claim.
There is also an earlier related outcome.
Trident Funding Limited, another MFS-linked funding vehicle, obtained a court order on 23 April requiring amounts held in collection accounts to be transferred by 8 May. Its administrators reported approximately £21.1 million held in those accounts. The company’s administrators’ proposals are available through Companies House.
That earlier order is relevant context for the current Barclays litigation. It should not, however, be treated as a binding legal precedent for the MFS Limited claim without knowing the legal and factual basis on which the earlier order was made.
The creditor base also illustrates the sophistication of the capital involved.
Financial Times reporting has identified Elliott Management with roughly £200 million of MFS-linked exposure. Bloomberg Law has reported exposure involving Sumitomo Mitsui Banking Corporation and Macquarie as well.
These were sophisticated institutional creditors. The wider financing structure also involved lawyers, auditors, a Security Agent, registered security and professional counterparties.
That makes the failure of the structure more — not less — important to understand.
The recovery effort continues. In May, the High Court allowed administrators to sell part of Raja’s collection of luxury cars for £1.625 million, according to Financial Times reporting. Raja continues to deny fraud and dishonesty.
But the policy question remains.
If a lender with MFS’s scale and structure applied for Annex 1 registration today, its application would face greater FCA scrutiny. Existing firms have been contacted and the FCA has gathered more information about their business models and risks.
What has not yet been announced is a change that would automatically bring a large Annex 1 lender into wider prudential supervision simply because of its size and interconnectedness.
The public measures discussed here have not created a single end-to-end audit responsibility across the separate legal entities in an MFS-like structure.
And the Bank of England’s July FOI response states that, following reasonable searches, it did not hold information within scope of the request for aggregate Level C collateral where the specified Corporate Services Providers were recorded, even though the underlying collateral itself is subject to extensive due diligence.
MFS therefore does not prove that every institution failed to perform its assigned role.
It demonstrates something more difficult.
A financial system can contain multiple auditors, lawyers, banks, trustees, institutional investors and regulators — each performing a bounded function — without those functions necessarily adding up to end-to-end verification of the economic reality.
That is borrowed trust.
Whatever Paresh Raja’s intentions when the structure was built, it ultimately operated across the spaces between those mandates: separate audit perimeters, information silos, contractual boundaries and divided institutional responsibilities.
Some of those gaps are now receiving greater scrutiny.
It is not yet clear that they have been closed.
And that is the question left by an estimated £1.8 billion group deficiency: not simply who failed to check, but who, in a structure like this, was responsible for checking the whole thing in the first place?
Scope and limitations
This article is based on publicly available information, including Companies House filings, FCA statements, FRC materials, administrators’ proposals, Bank of England publications, a Bank of England FOI response provided to Big4News, and Financial Times, Telegraph, Bloomberg and other media reporting.
It does not draw on internal bank or audit working papers, privileged communications or confidential regulatory submissions. The public record therefore cannot establish the full extent of the due diligence, audit procedures, internal communications or risk assessments undertaken by every institution discussed.
No finding of breach or misconduct has been made in the MFS-related FRC investigations, which remain open, and nothing in this article should be read as a conclusion about the legal liability of any party in relation to the MFS collapse.
Paresh Raja’s spokesperson has denied the fraud and dishonesty allegations and has said that funds administrators allege were received into personal accounts were held through nominee structures for the benefit of MFS and its creditors.
The purpose of this analysis is to examine what the public record reveals about the distribution of responsibility across the MFS structure, and what that distribution may tell us about audit scope, regulatory design and the risks created when institutional assurance is fragmented.
This article is part of a three-part series:
Part 1 — £1.8 billion and borrowed trust: the MFS structure no one was mandated to see whole
Part 2 — £1.8 billion and borrowed trust: the limits of institutional oversight
This article is part of the Big4News Investigations & Analysis series, which examines the structural forces shaping Deloitte, PwC, EY and KPMG.
