On 22 April 2026, Golden Pass LNG shipped its first cargo. The US Energy Information Administration’s note on the milestone contains a sentence that belongs on the wall of every capital project procurement office: the terminal “was delayed when the lead construction contractor filed for Chapter 11 bankruptcy in 2024, requiring a new lead contractor to finish the plant.”
That is a federal statistical agency attributing schedule slippage on an US$11.6 billion LNG terminal to a counterparty’s balance sheet. Zachry Holdings filed on 21 May 2024; its share of the fixed-price EPC contract was approximately US$5.8 billion of an amended contract price above US$10 billion. First cargo arrived roughly sixteen months after the original target.
Procurement functions on large capital projects are well organized around price risk, schedule risk and technical risk. Counterparty financial risk is usually delegated – to a prequalification form, a credit check at onboarding, and a bond. The evidence from the last twenty-four months suggests that is no longer proportionate to the exposure.
In the United Kingdom, the Insolvency Service’s official statistics for the twelve months ending July 2026 record construction at 3,841 cases, 17% of cases with industry captured – the largest single sector, ahead of wholesale and retail (3,422, 15%) and accommodation and food service (3,221, 14%). Construction has held that first place in every recent period. Roughly one in six UK corporate insolvencies is a construction business. BCIS notes that while the twelve-month figure is down 3% year on year, it remains 19% above the pre-pandemic 2019 level of 3,221.
The pattern repeats internationally. Dun & Bradstreet’s Global Bankruptcy Report 2026 finds that across the 45 economies it monitors, bankruptcy filings grew 7% in 2025, and that in France construction accounts for 17.3% of all bankruptcies – almost exactly the UK share. Credit reform recorded 23,900 German company insolvencies in 2025, up 8.3%, with construction up 4.7%, and estimated total creditor losses at approximately EUR 57 billion affecting some 285,000 employees. Allianz Trade’s Global Insolvency Report of 22 April 2026 forecasts a further +6% globally in 2026 with levels plateauing at a historic high in 2027, puts 2.2 million jobs directly at risk, and names “construction, retail, and services” as the sectors most affected.
Honesty about the counter-evidence matters here. Eurostat’s Q2 2026 release shows EU bankruptcy declarations up 5.7% quarter on quarter but construction bankruptcies down 1.7%. Allianz Trade’s +6% is also the highest of three credit insurer forecasts; Atradius published +3% in April 2026 and Coface +2.8%. The direction of travel is not uniform. What is uniform is that construction sits at or near the top of the sector table in almost every national dataset.
The Construction Index’s Top 100 league table for 2026 shows the UK’s hundred largest contractors turning over a combined £81 billion, up 9.1%, for a combined pre-tax profit of £2.3 billion – an average pre-tax margin of just over 3%. That is an improvement: the 2025 edition reported 2.4%, up from 1.9%.
The argument is not that margins are collapsing. It is that even after two consecutive years of improvement, the largest contractors in a mature market clear barely three pence of pre-tax profit per pound of turnover. At that margin, a single materially mispriced contract can consume an entire year of group profit. Carillion announced an £845 million hit in July 2017 and entered compulsory liquidation on 15 January 2018, owing around £2 billion to 30,000 suppliers, subcontractors and other short-term creditors.
Thin margins also mean that distress propagates fast. Petrofac demonstrates the speed. On 1 July 2025 the Court of Appeal in Saipem S.p.A. v Petrofac Limited [2025] EWCA Civ 821 allowed the appeal and set aside the order sanctioning Petrofac’s Part 26A restructuring plans. On 27 October 2025 Petrofac Limited, the group’s ultimate holding company, filed for administration, with joint administrators appointed the following day. The company’s own statement identified the trigger as “Tenet’s decision to terminate Petrofac’s scope of work on the 2GW programme in the Netherlands” – a contract reported to represent over 80% of the engineering and construction division’s revenue. A listed EPC contractor went from a contested restructuring to holding-company administration in under four months, and the proximate cause was client concentration.
When a counterparty fails, what does an unsecured creditor actually get back?
The Insolvency Service’s own research, published 17 December 2024 and covering 2,717 completed creditors’ voluntary liquidations, is unambiguous: in 90% of cases there was no distribution to unsecured creditors at all. Its verbatim finding — “The median was 0% for all cases with unsecured creditors. In cases where a distribution was made, the median was 9%.”
The construction-specific illustration is ISG. Eight ISG entities entered administration on 20 September 2024. EY’s public administration page states plainly: “No payments are expected to be made to unsecured creditors of the Companies.” The administrators’ progress report for the period to 19 March 2026 records £885 million of unsecured claims against £38.5 million realized in total, of which £12.2 million was collected in the reporting period from £241.6 millions of book debts, retentions and work-in-progress balances – a recovery of roughly five pence in the pound on that category alone. HMRC is owed £91 million and will not be fully repaid.
Retention, the traditional hedge, is both smaller and less reliable than its reputation. The most recent official UK estimates — BEIS’s 2017 impact assessment, at 2015 prices – put total retentions held at between £3.2 billion and £5.9 billion a year, with a central value of £4.5 billion, of which £229 million a year was lost to upstream insolvency. The underlying Pye Tait research found average retention at 4.8% of contract value and that 44% of contractors had experienced non-payment of retention through upstream insolvency in three years – though on only about 1% of their contracts.
And retention is now on the way out. The UK Government’s response to the 2025 Late Payment consultation, Time to Pay Up, published March 2026, states: “The Government proposes to take forward a legislative measure to prohibit the deduction and withholding of retention payments under the terms of a construction contract.” Implementation is subject to further consultation, with respondents favoring a 12–24-month transition. If it lands, the weight shifts decisively onto bonds, parent company guarantees and – the point of this article – upstream financial monitoring.
The bond market is already repricing. AM Best reported in February 2025 that the US surety direct loss ratio for the first nine months of 2024 reached 25.0%, the highest in five years, with Robert Valenta noting that “Rising losses have led carriers to tighten underwriting standards for certain bond classes… Sureties are also monitoring contractor performance closely, particularly those that have had performance issues or are highly leveraged.”
Here the evidence is genuinely counter-intuitive.
Payment behaviour is a poor predictor at the top of the supply chain. CreditRiskMonitor’s research on what it calls the “cloaking effect” finds that trade payment scores do not reliably predict public company bankruptcy, because “Some companies pay their credit obligations in a discount or prompt manner right up to the actual filing/closing date.” Carillion was a Prompt Payment Code member while imposing 120-day terms on suppliers. A large contractor with a clean payment record can be weeks from collapse.
Statutory debt is a far sharper signal. Research published in the Journal of Risk and Financial Management on 29 June 2026, analyzing ASIC insolvency data for 2021–2024, found that non-payment of statutory debts – tax, payroll, superannuation – “were observed in more than 80% of reported cases,” that inadequate cash flow was the most frequently cited cause (16.6%–20.2%), and that construction represented nearly a quarter of all Australian company insolvencies in 2024. Critically, more than 92% of failures in every year were firms employing fewer than 20 people 0- that is, the risk concentrates in tier two and tier three, exactly where buyers have least visibility.
Statistical models give lead time, not certainty. A 2025 meta-review in the same journal reports that Altman Z-score studies across more than thirty countries show “average one-year accuracy rates of approximately 75%, which rise above 90% when the coefficients are tuned to local datasets.” Altman’s original model was 72% accurate two years out. A screen, not a decision.
Credit insurers often move first, and publicly. In August 2026 it was reported that Allianz Trade had cut credit cover for Vistry’s suppliers by as much as 70% – on new trading agreements rather than existing cover – against a backdrop of £470 million net debt on 30 June 2026 and guidance of an approximate £30 million first-half loss before tax after cash-generation actions. An underwriter with its own money at stake repriced a listed contractor’s supply chain exposure, in public, with no insolvency event in sight.
Supplier finance usage is now disclosable – and therefore readable. After Carillion concealed an estimated £498 million of debt through its Early Payment Facility, and Greensill’s March 2021 collapse froze US$10 billions of Credit Suisse funds, both standard setters acted: FASB’s ASU 2022-04 and the IAS 7 / IFRS 7 amendments effective 1 January 2024 require disclosure of supplier finance programme obligations. In the first quarter of reporting, around 80 S&P 500 companies disclosed at least US$64.1 billion of such obligations. In September 2025, Moody’s warned that payment terms under these programmes are being pushed beyond the 90 days it treats as a reasonable upper limit for trade payable classification, and that “Should a programme’s financial sponsor withdraw or curtail it, the programme can unwind rapidly, putting stress on liquidity.”
In the UK, a free public register makes much of this screenable before award. Under the Reporting on Payment Practices and Performance Regulations, large companies publish average days to pay, payment bands by number and value, and disputed-payment percentages twice a year. The Amendment Regulations 2025 (SI 2025/75), in force 1 March 2025 and applying to financial years beginning on or after 1 April 2025, added retention practices in construction contracts — including standard retention rates and the ratio of retention withheld from suppliers to retention withheld against the reporting company. Build UK publishes the construction sector results contractor by contractor.
The gap between how well procurement functions believe they monitor counterparty risk and how well they actually do is the most striking finding in the current survey evidence.
Sphera’s Supply Chain Risk Report 2026, published 22 January 2026 and drawn from 800 CPOs and chief supply chain officers across the US, UK, Germany and Canada, reports that 98–100% of respondents say they are confident in the completeness and timeliness of supplier risk data – while 73% report financial or operational losses due to supply chain disruptions in the past 12 months, averaging 3.48 material disruptions each. In the same dataset, “supplier viability risk remains the largest category by volume for the third consecutive year and continues to rise, increasing by approximately 10.3% in 2025 versus 2024.”
Achilles’ Annual Risk and Sustainability Report 2026, published 13 March 2026 from 2,805 organizations across construction, energy, manufacturing, transport and the public sector – a panel that maps closely onto capital project supply chains – found only 6% have full visibility into tier-two and tier-three suppliers, with nearly half having limited or no visibility beyond their immediate supplier base. Achilles also reports that supplier financial failure or distress, and quality or performance failures, were each cited as the most common form of disruption by approximately 31% of respondents – financial distress is joint-first, not a distant second to operational causes. Nearly 58% manage supplier risk without any dedicated software platform, and only 19% use an established third-party platform.
KPMG’s Global Construction Survey 2025/2026, published in early 2026 from 375 industry leaders (fieldwork January–March 2025), frames the same gap as a widening “risk delta”: 75% report increased risk aversion compared with twelve months earlier, while only 43% have adopted risk monitoring tools.
Meanwhile the priority has already shifted. Reporting on The Hackett Group’s 2026 Procurement Agenda, ISM’s Inside Supply Management placed supply continuity first among procurement priorities for 2026, ahead of spend cost reduction – a reversal of Hackett’s 2024 finding that cost reduction had returned to the top.
Concentration makes the exposure worse than the averages suggest. CAPS Research benchmarks reported by ISM show that across industries, companies place 31% of sourceable spend with their top ten suppliers and 54.3% with their top fifty. On a capital project the concentration is typically higher still, because the long-lead packages are single-sourced by design. No public benchmark for EPC-specific supplier concentration exists -which is itself telling: procurement heads are managing concentration risk without a reference point.
Monitoring counterparty financial health at project scale is a data problem before it is a credit problem. The signals that matter are scattered: statutory filings and their timeliness; credit and failure scores; published payment practices and retention behaviour; supplier finance disclosures; order book and client concentration; trade credit insurer appetite; director and ownership changes; adjudication and litigation records. Each is available. None of them is useful as a one-off check at prequalification, because the whole point of the evidence above is that the deterioration happens after award.
What turns those signals into something usable is a supplier record that persists – one place where a counterparty’s ownership, financial indicators, certification status, spend concentration and performance history sit together and stay current across the life of the project, rather than being reassembled from a prequalification pack, a credit report bought in 2024, and the memory of whoever ran the tender. That persistent supplier record is what supplier relationship management platforms such as Dharas are built to hold, and it is the difference between learning that a fabrication yard is in difficulty from a news alert and knowing it from your own vendor master three months earlier.
Golden Pass lost sixteen months. ISG’s unsecured creditors are recovering nothing against £885 million. Petrofac went from contested restructuring to administration in under four months. In each case the financial information that mattered existed in public, before the failure. The gap was not availability. It was attention.
Local content law now binds subcontractors directly and penalizes the operator. What DRC, Cameroon, Senegal...
READ ARTICLE →
Industrial organizations authorize an average of 77 third parties into their OT environment; a quarter...
READ ARTICLE →
A steel importer settled at $19 million in May 2026 for declaring Chinese and Turkish...
READ ARTICLE →