A Guardian-commissioned study of more than 1,700 administrations from 2019 has produced the clearest picture yet of who wins and who loses in the UK's pre-pack insolvency system. The answer is uncomfortable: the same directors who presided over failure are buying back their companies' assets — often on deferred payment plans — and failing again at nearly double the rate of outside buyers. Unsecured creditors, including HMRC, are left holding the bag. The mechanics are straightforward. A company enters administration. Before the process formally begins, a deal is pre-arranged for the assets to be sold — frequently back to the same directors or shareholders who ran the business into the ground. The new entity emerges debt-free. Creditors who supplied goods, extended credit, or are owed tax receive almost nothing. The study by Turnstone Intelligence found that in connected-party pre-packs where payments were deferred, unsecured creditors received just 1.58p in the pound. When an outside buyer was involved, the return was 7.86p — nearly five times as much. The failure rate gap is equally stark. Some 44% of connected-party pre-packs had entered a second insolvency by July 2026, compared with 26% when an unconnected buyer took over. This is not a marginal difference. It suggests that in a large share of cases, the "rescue" is simply a mechanism for shedding liabilities while continuing a non-viable business. As UCL honorary law professor Riz Mokal put it, these are "zombie businesses" that harm not only their creditors but competitors who continue paying their way. The deferred-payment angle is particularly damaging. Turnstone found that 65% of connected-party pre-packs allowed former owners to pay in future instalments, compared with 28% for unconnected buyers. Deferred payments reduce the immediate cash returned to creditors and increase the risk that the buying entity itself will fail before completing payment. The correlation between deferred payments and poor creditor outcomes is consistent with findings from the 2014 Graham Review and a 2018 EU-funded study — the same pattern, identified three times over twelve years. HMRC estimates that phoenixism costs taxpayers hundreds of millions of pounds annually. This is not an abstract accounting loss. It represents VAT, PAYE, and National Insurance contributions that were owed by the original company, wiped out in administration, and never recovered from the successor entity. Every phoenix deal that fails a second time doubles the loss. The taxpayer funds both the original shortfall and the subsequent one. The insolvency industry's defence rests on two pillars: that the practice is legal and regulated, and that 2021 rule changes requiring independent evaluators to approve connected-party sales have improved the system. R3, the industry body, correctly notes that the study's data predates those reforms. But the structural incentives remain unchanged. Directors still have an information advantage over outside bidders, administrators still have an incentive to close deals quickly, and deferred-payment arrangements still allow connected parties to acquire assets with minimal upfront capital. The study's methodology — AI-assisted reading of official filings, verified by humans and checked by secondary models — represents a scalable approach to monitoring insolvency outcomes. The question is whether regulators will use it. Three separate studies over twelve years have now reached the same conclusion: connected-party pre-packs produce worse outcomes for creditors and higher failure rates. The 2021 reforms added scrutiny but did not change the fundamental economics. Until the cost of phoenixing is borne by those who benefit from it rather than those who are harmed by it, the pattern will persist.
