On 30 July 2026 the First-tier Tribunal released Cogefin (Bermuda) Limited & Anor v HMRC [2026] UKFTT 1108 (TC). The company was Bermuda-incorporated, staffed by Bermudian lawyer-directors, owned by an offshore trust — and still held UK-resident for corporation tax from 1999 to 2017. Central management and control sat with a UK-resident economic settlor and beneficiary, not the board that signed the minutes.
If you sit over family-office holdcos, PE SPVs, trust-owned investment companies or any offshore board that “receives recommendations” from the UK, this is not a historical curiosity. The Tribunal walked a 20,000-page correspondence run and applied the classic CMC tests. Paper residence lost. Real decision-making won. Discovery assessments stood. Penalties were cut because the failure was careless, not deliberate — and the personal liability notice on the individual fell with that finding.
Primary sources and reads: the National Archives judgment, Devereux Chambers’ case note, Claritax’s summary, Ross Martin, HMRC’s residence review approach in INTM120180 and Statement of Practice material at INTM120200, plus the Court of Appeal framework in Wood v Holden [2006] EWCA Civ 26.
What Cogefin actually was
Cogefin was incorporated in Bermuda in February 1996 as an exempted company, ultimately owned by the Poole Family Trust. The economic settlor and beneficiary was Mr Giuseppe Ciardi, a UK-resident investment banker. The directors were Bermudian-resident lawyers from MLH Quin & Co (later Wakefield Quin), with MQ Services providing administration. Initial funding was roughly $7.7m of stocks. By 1999 the company was worth about $25m; by 2011, more than $250m.
The portfolio was not passive cash. Investments ran through major houses such as Morgan Stanley and Goldman Sachs. There were hedge-fund subscriptions, loans linked to Mr Ciardi and associated entities, European commercial property, renewable-energy project vehicles, and funding for residential property, art and jewellery held through SPVs — including the London residence at 27 Chester Square. The LDF disclosure described Mr Ciardi as an investment adviser whose recommendations the directors routinely sought and followed. HMRC opened a Code of Practice 8 enquiry and contended that CMC had been exercised from the UK since at least 1999.
The CMC test the Tribunal actually applied
There was little dispute on the law. A company is resident where its central management and control actually abides. That is a pure question of fact about who makes the high-level decisions and where. The Tribunal rehearsed the usual authorities:
- De Beers — real business is carried on where CMC abides; look at the course of business, not the by-laws.
- Untelrab — strong recommendations are not usurpation if the board still exercises discretion and would refuse an improper or unwise deal; stand back and take the whole picture.
- Wood v Holden — directors need not initiate proposals; they may rely on advice; ill-informed decisions still count if the directors actually apply their minds to whether to enter the transaction.
HMRC’s International Manual makes the same practical cut in INTM120180: ask who ought to manage the company under constitution and local law, then test whether those people actually do so — and if not, who does, and where. Board meeting location only helps if the board truly holds and exercises the controlling power.
Why the Bermudian board lost
The appellants said the directors considered advice, made decisions for themselves and understood their fiduciary duties. HMRC said Mr Ciardi’s correspondence was instruction and control: investments, loans, property purchases, bank openings, FX and fund subscriptions were directed from the UK, with the board implementing decisions already made.
The Tribunal preferred the documentary pattern over reconstructed recollection. On the overall picture for each period, decision-making was abdicated by the directors to Mr Ciardi rather than usurped by him. Direct requests were never refused. Payment instructions often went to the bank almost immediately, before any real consideration of the merits. Directors and administrative staff ensured Cogefin could execute Mr Ciardi’s proposals; they treated those proposals as decisions and did not apply their minds as directors. A handful of apparent board decisions was not enough for dual residence. Result: Cogefin was UK-resident only throughout the periods under appeal.
That is the operational distinction CFOs need. Wood v Holden still protects a real board that receives proposals, deliberates and decides. It does not protect a board that functions as a high-end company secretary for a UK principal.
Discovery assessments still stuck
Nineteen discovery assessments covered 1999–2017. Later years were in time without controversy. For 1999–2014 HMRC needed the extended gateways under Schedule 18 FA 1998. The Tribunal held the assessments were validly raised: there was a discovery of a loss of tax, no reasonable excuse for failure to notify chargeability, and carelessness (or the predecessor negligence standard where relevant) that opened the longer window. The LDF report had asserted Bermuda residence; HMRC’s discovery crystallised after further information in November 2018, after the ordinary enquiry window closed. “We told you in the LDF that we thought it was non-resident” is not a free pass when the facts show UK CMC and no UK CT notification.
Penalties cut; PLN off — careless, not deliberate
Failure-to-notify penalties for 1999–2013 did not survive at HMRC’s deliberate-behaviour level. The Tribunal found the behaviour careless, not deliberate. Substituting the decision HMRC could have made, it fixed the overall penalty percentage at 25% of potential lost revenue for the periods in question, keeping HMRC’s mitigation judgment broadly intact. Because deliberate behaviour was not made out, Mr Ciardi’s personal liability notice under paragraph 22(1) Schedule 41 FA 2008 could not stand. Company tax and reduced penalties remain; personal penalty transfer does not follow automatically from influence or even shadow-director allegations when deliberate failure is not proved.
For boards and principals, that split matters. Residence risk is a tax and disclosure control. Deliberate-behaviour risk is a separate evidence trail — advice taken, positions documented, disclosures made, and how far anyone knew the CMC picture was wrong.
What every CFO must lock now
- Inventory offshore decision rights. List every non-UK company in the group, family office or PE stack where a UK person originates investments, treasury moves, related-party loans or SPV acquisitions. Map constitutional directors vs actual proposers vs bank mandate holders.
- Stress-test CMC with documents, not org charts. Sample a year of emails, board packs, payment instructions and bank auth trails. If “recommendations” are executed the same day with no board paper, challenge or minutes that show minds applied, you have a Cogefin fact pattern — not a Wood v Holden fact pattern.
- Rewrite the language and the process. UK principals and family offices should send recommendations for board consideration, with pack, time to consider, and a recorded decision (including reasons to refuse). Offshore directors need standing, expertise and a demonstrated willingness to say no. Rubber-stamp minutes are exhibits for HMRC, not a defence.
- Separate “advice” from mandates. If a UK individual can move bank balances, instruct brokers or commit the company without a prior board decision, fix mandates before the next audit committee cycle. The 2008 crisis instruction to move funds into government bonds was exactly the kind of beyond-authority act that lit HMRC’s fuse in the LDF narrative.
- Revisit CT notification and historic open years. If CMC analysis has never been written down, commission one. Where UK residence is the honest answer, quantify unpaid CT, interest and failure-to-notify exposure under Schedule 18 / Schedule 41. Do not wait for a disclosure facility to do your taxonomy for you.
- PE and M&A due diligence. Buy-side tax DD on trust-owned or founder-advised offshore holdcos should demand the correspondence run, not only incorporation certificates and director CVs. Warranties limited to “validly incorporated in Bermuda/Cayman/Jersey” are worthless on residence. Ask for CMC memos, board calendars, who proposed each material deal, and any LDF/COP8/nudge history.
- Audit committee one-pager. Three facts (entities with UK influencers, last CMC review date, open HMRC touchpoints), three controls (board process, bank mandates, notification policy), one decision (any entity that needs re-domicile, board rebuild or voluntary disclosure).
The CFO takeaway
Cogefin does not invent a new residence test. It shows how a modern Tribunal will try one: long hearing, full chronological bundle, honest witnesses with unreliable memory, and a standing-back conclusion that administrators are not decision-makers. Bermuda brass plates, trusted local counsel and a trust above the company are not enough if the UK principal’s inbox is where the company actually lives.
Lock the governance so the board really decides. If it does not, lock the UK tax compliance as if the company were onshore — because for CMC purposes, it already is. For the judgment and practitioner notes, start with the FTT decision, Devereux, Claritax, INTM120180 and Wood v Holden. Then open your own offshore board packs and ask the only question that matters: who actually said yes?