Civmec Limited

Australia · Heavy engineering, construction & shipbuilding

ASX: CVL · SGX: P9D · Information cut-off 28 August 2026

What the company is

Civmec earns money one way: it wins fixed-price and schedule-of-rates construction contracts and recognises revenue over time as it executes them. What distinguishes it from most contractors is that it owns its facilities. The balance sheet carried A$619.6m of property, plant and equipment at 30 June 2026 against A$903.0m of revenue — a ratio no asset-light contractor would show. The principal asset is the Henderson yard in Western Australia — a 70,000 m² undercover fabrication and assembly hall, including a covered shipbuilding hall, on a 200,000 m² site, which the issuer rates at 80,000 tonnes of throughput a year. The Henderson buildings stand on leasehold land, held under three land leases dating from 2010, 2014 and 2016, each with a renewal option. A second facility at Tomago, Newcastle adds 227,000 m² and a further 25,000 tonnes, with newer sites at Port Hedland and Gladstone.

Revenue by customer location in FY2026 was Australia A$903.0m — 100%, as in every year since the change of domicile. There is no offshore revenue to analyse.

The company reports three segments: Energy; Resources; and Infrastructure, Marine & Defence. The issuer states it aggregates Infrastructure with Marine & Defence “as they have similar long-term average gross margins”. As the segment section below shows, that aggregation stopped being descriptive in FY2025 — and became descriptive again in 1H FY2026.

Three structural breaks the series must respect

Anyone plotting Civmec’s revenue from listing to today on one line is plotting a fiction. There are three breaks, and two of them land on the same year.

The FY2018 bridge is the issuer’s own, from note 32 of the FY2019 annual report, and it ties exactly: A$712,850k of translated revenue less a A$10,435k SFRS(I) 15 adjustment equals A$702,415k restated.

As-filed history

Two eras, held apart. First the Australian-dollar era on the current framework:

EBITDA is not a statutory measure. It is derived here as profit before tax plus total finance costs plus depreciation. Civmec books lease interest inside cost of sales (A$3,797k in FY2025, A$3,993k in FY2026), so the profit-and-loss finance-cost line is not the group total; on the total basis FY2025 EBITDA is A$91.7m at an 11.3% margin and FY2026 is A$107.3m at 11.9%, each exactly the issuer’s own published figure. Two consecutive years tying to the cent is the reason this derivation is used rather than the P&L line. Net debt is borrowings plus lease liabilities less cash, on ending balances.

And the Singapore-dollar era, exactly as originally published — which is where the company’s only completed downcycle sits:

Revenue fell 30.7% from FY2015 to FY2017 and profit to owners fell 76.0% over the same two years, through the end of the Australian mining-capex boom. The company came out the other side without a rescue equity raising. That is the most useful thing the old series records, and it is invisible in any chart that starts at FY2019.

Segment economics — and the disclosure gap

State the gap first. Note 32 of the FY2025 financial statements says the Group “does not identify nor segregate its assets and liabilities in operating segments as these are managed on a ‘group basis’.” A labelled “Segment assets” block does appear, but the only asset allocated to any segment is A$10k of intangibles (to Resources; the same A$10k in FY2024). FY2026 changed nothing: note 4.1 of the FY2026 statements again allocates that same A$10k and presents A$1,210.1m of assets and the whole A$628.7m of liabilities as unallocated, with capital expenditure given for the group only. At 30 June 2025 the comparable unallocated figures were A$906.8m and A$381.4m. Return on segment assets, segment net capital employed, and the profit-share-versus-asset-share comparison are therefore not computable from the public record. Nothing on this page estimates them.

Correction, 18 August 2026. An earlier version of this paragraph said there were “no segment assets and no segment liabilities disclosed in any year”, and quoted the note as “as they are managed”. Both were wrong: the blocks exist but are almost entirely unallocated, and the filing reads “as these are managed”. The substantive point — that no meaningful capital is allocated by segment — is unchanged. Verified at printed p112–113 of the FY2025 report of directors and financial statements.

What the issuer does disclose is revenue and a “segment result” per segment. That segment result equals group gross profit — it ties to the reported gross profit line exactly in every year, which is worth stating because the label does not say so.

FY2021 Infrastructure, Marine & Defence gross profit is derived by difference from the group total; every other cell is as reported. Segment revenue and segment results both sum exactly to the group figures in all five years.

The FY2026 full year settles both FY2025 anomalies

Two things were unresolved in the FY2025 segment note: an Infrastructure, Marine & Defence gross margin of 34.5% against an 11–14% history, and a Resources margin of 7.6% against 11–13%. The full year has now answered both, and in the direction the half-year note pointed. Infrastructure, Marine & Defence came in at 11.0% and Resources at 11.5%. Neither was a new level; both were timing.

What that leaves is a result whose shape is not the one the headline suggests. Group profit to owners rose 22.5%, and the segment that produced the increase is Resources, whose revenue fell: A$587.1m against A$641.2m, down 8.4%, with gross profit up 37.5% to A$67.3m on the margin recovery alone. Infrastructure, Marine & Defence did the opposite — revenue doubled to A$210.2m as Civmec Defence Industries consolidated for a full year, while its gross profit fell 35.8% to A$23.1m. Energy grew on both lines, from a small base.

One issuer statement does not survive its own segment note. Item 2.6 of the Appendix 4E attributes the revenue growth to the Luerssen acquisition “and the increase in activity within the Resources segment which saw the commencement of several large projects”. On a full-year basis Resources activity did not increase — it fell 8.4%, and the same document set says so twelve pages later (“reflecting lower activity levels during the year”). The increase is real but it is a second-half event: Resources revenue rose 55.6% half-on-half, from A$234.3m in 2H FY2025 to A$364.5m in 2H FY2026. Read as a full-year driver the sentence is wrong; read as a second-half one it is right, and the filing does not say which it means.

The half-year note that got there first

Note 26 of the 1H FY2026 interim financial statements, published 12 February 2026, gives the segment note for the half. It answers the question the full-year figures raise:

Two things followed. First, the Infrastructure, Marine & Defence spike was concentrated in the first half of FY2025 — 37.9% against 28.1% in the second — which is what a completion-driven recognition release looks like. Second, both anomalies were already unwinding: the segment was back to 12.5%, and Resources had recovered from 6.9% to 10.9%. This page concluded on 17 August 2026 that “the FY2025 full-year Resources margin of 7.6% was a trough, not a new level”. The FY2026 full year settled it at 11.5%. The half-year segment note is the disclosure that carried the information, six months before the full year confirmed it.

The Luerssen acquisition, as disclosed

Note 12 of the same statements sets out the provisional purchase-price allocation for Luerssen Australia, now Civmec Defence Industries:

Identifiable net assets equal the net consideration exactly, so no goodwill arises. Cash acquired of A$36.5m exceeded the A$20m paid, making the transaction a net cash inflow of roughly A$16.5m. The issuer states that contracted confidentiality obligations prevent it from disclosing Civmec Defence Industries’ separate financial contribution; the FY2026 accounts repeat that refusal word for word. The FY2026 note carries the same acquisition-date amounts as the interim note, so no measurement-period adjustment was recognised.

A A$92.8m payable to Naval Vessels Lürssen for pre-acquisition contract milestones sat in trade and other payables at 31 December 2025. That payable is the visible half of a structure worth understanding: the acquisition brought in A$128.315m of trade and other receivables and A$128.315m of trade and other payables — the same figure to the dollar. The interim statements name both sides: the receivable is “outstanding pre-acquisition contract milestone payments recoverable from the Commonwealth of Australia”, the payable is the matching amount owed “through to NVL Australia GmbH in accordance with the Share Sale and Purchase Deed”. Civmec collects from the Commonwealth and pays it on. It is economically neutral and it does not touch profit, but it grosses up both sides of the balance sheet and it runs through the working-capital lines of the cash flow statement, which is why the FY2026 balance sheet and cash flow below cannot be read without it.

The issuer’s own explanation for the Infrastructure, Marine & Defence move is one clause, in the FY2025 review of performance: the increase was “mainly due to the timing of profit recognition from the completed Boorloo Bridge project (known as the Causeway Pedestrian and Cyclist Bridges)”. The issuer does not quantify it anywhere. Applying the segment’s own FY2021–FY2024 aggregate margin — A$53.5m of segment result on A$459.6m of segment revenue, or 11.6% — to its FY2025 revenue would give roughly A$12.1m of gross profit rather than A$35.9m, an indicative A$23.8m above that benchmark. That figure is a benchmark-based computation shown here for scale, not an issuer disclosure, and it should not be read as an estimate of an “abnormal” amount: recognition timing moves profit between periods, it does not create or destroy it.

FY2026 has now tested that benchmark. The 11.6% figure was struck from FY2021–FY2024 before the segment’s post-acquisition shape was known; the segment reported 11.0% in FY2026, on revenue two-thirds larger than any year in the benchmark window. The FY2026 review of performance names the same cause for the fall that it named for the rise — “primarily due to the timing of profit recognition from the completed Boorloo Bridge project in FY2025”. The issuer still does not quantify it in either direction. What can now be said, which could not be said a week ago, is that the segment’s underlying margin is a low-teens number and the FY2025 print was not it.

Customer concentration

The FY2026 segment note discloses that the Group supplied three major external customers, now in the Resources and the Infrastructure, Marine and Defence segments, accounting for approximately 50.09% of external revenue (FY2025: three customers, all in Resources, 51.4%; FY2024: two customers, 53.5%). Because only A$10k of assets is allocated to any segment, that exposure cannot be sized against the capital standing behind it. It is worth setting against the listing-era position: the 2012 IPO prospectus discloses that five customers were 88.3% of FY2011 revenue, and that at 1Q2012 the top two alone — Leighton Contractors at 40.9% and BHP Billiton Iron Ore at 35.1% — were about 76%. The bases differ (top five then, top three now), but even comparing the top two of 2012 with the top three of 2025 runs conservatively and still shows a large reduction. Concentration here is high in absolute terms and materially lower than at any earlier point in the listed history. Named FY2026 work in the issuer’s own updates includes BHP’s Port Debottlenecking Project 2 at Nelson Point, Iluka Resources’ Eneabba rare-earths refinery and Chevron Gorgon modules — a large-cap counterparty set.

The order book

Civmec publishes a forward order book at most quarterly updates. It is an issuer-defined, unaudited disclosure, and it has been stated on at least three different bases: point values, “in excess of” floors, and — on 4 June 2026 — a figure including only Civmec’s participating interest in an alliance contract. That inconsistency is recorded here rather than smoothed over.

A fourth measurement basis appeared with this result. The FY2026 book is given “as at 31 July 2026” — a month after the balance-sheet date — and no 30 June 2026 point value is published anywhere in the results set. The FY2025 book was stated the same way, at 31 July 2025. That is now three inconsistencies in one series: point values against “in excess of” floors, participating-interest-only treatment of alliance work from 4 June 2026, and a measurement date that moves between 30 June and 31 July. Five observable pairs let the relationship between the book and the following year’s revenue be computed rather than asserted:

FY2023 is excluded because no 30 June 2023 point value was published — the issuer said only “in excess of A$1 billion”. The FY2025 row uses the 31 July 2025 figure because no 30 June 2025 value was published either; it is one month of awards further on than the other rows and is labelled rather than adjusted. On that basis conversion has ranged 0.72x to 0.95x over five pairs. The issuer has published a conversion statement once, with the FY2020 results: of the A$900m book at 30 June 2020, “over half” was expected to be realised in FY2021. No conversion statement accompanies the A$1.4bn book; the company says only that it provides “visibility into FY27 and FY28”.

The Q3 FY2026 business update of 15 May 2026 reported quarterly revenue of A$244.2m, nine-month revenue of A$624.7m, nine-month EBITDA of A$73.8m (11.8% margin) and nine-month net profit after tax of A$34.9m (5.6% margin), with earnings per share of 6.86 Australian cents for the nine months. The full year now brackets that update: subtracting the nine months from the reported year gives a fourth quarter of A$278.3m of revenue and A$17.2m of profit — the largest quarter on record and a 6.2% net margin, above the 5.6% of the preceding nine months. Both Q4 figures are derived by difference; the issuer does not report quarterly figures at the full year.

The programmes the company has named, and how their dates have moved

Named initiatives, with the dates as first announced and as most recently restated. Built from the issuer's own filings; where a figure has been published more than one way, every version is shown rather than one being chosen. Execution risk is the author's read and carries its basis in the cell.

What the balance sheet is made of

Civmec carries freehold land and buildings under the revaluation model, with gains taken to other comprehensive income and never through profit. The consequence compounds:

This matters for anyone computing a return on equity or a price-to-book from screening data, because the denominator contains a large and growing amount of capital the operating business never earned or deployed:

Funding is modest but not disclosure-light, and FY2026 moved it. At 30 June 2026: cash was A$54.6m against A$102.9m a year earlier, borrowings A$60.0m unchanged, and lease liabilities A$66.4m against A$60.5m — giving lease-adjusted net debt of A$71.7m against A$17.6m, a A$54.2m increase in one year. On derived EBITDA that is 0.67x against 0.19x, with 11.4x EBITDA coverage of total finance costs against 9.5x. Leverage is still low in absolute terms; the direction is the fact.

The facility architecture is now one shared limit rather than the FY2025 structure this page previously described. The A$86m corporate-loan limit and its A$2m quarterly step-down were superseded during FY2026: the existing facilities were renegotiated and consolidated into a single revolving multi-option facility of A$160m (FY2025: A$156m) covering corporate market loans, letters of credit and bank guarantees together, so the quarterly limit reduction is no longer in effect. It was 39.0% utilised at 30 June 2026 (FY2025: 40.5%), leaving roughly A$97.6m of unused limit; interest ranged 4.49%–5.34% (FY2025: 4.98%–5.55%). The facility is secured by a general security deed over certain property, plant and equipment and investment properties. The issuer states all financial covenants — leverage, tangible net worth and debt-service cover — were met at 30 June 2026, and still does not publish their definitions, thresholds or numerical headroom. The facility expires on 31 January 2028, with a possible bank extension; draw conditions, minimum operating cash and the accessibility of secured cash remain incomplete in the public record. Because one limit now serves cash borrowing and bonding at once, unused limit is not the same thing as spare cash capacity.

Issued shares went from 500,985,000 at FY2020 to 509,625,000 at FY2026, a 1.7% increase over six years, entirely from the conversion of performance rights — 1,097,000 of them in FY2026. There have been no placements or rights issues in the Australian-dollar era.

Cash conversion

Cumulatively over FY2019–FY2026 the picture is still different from any one year: operating cash flow of A$438.5m against profit to owners of A$325.9m — a ratio of 1.35x, down from 1.69x on the FY2019–FY2025 window. Cash capex over the same period was A$222.1m, of which A$138.3m fell in FY2019 and FY2020 during the facility build-out. FY2026 cash capex of A$5.4m was 0.23x that year’s depreciation of A$23.7m — a second consecutive year at roughly a fifth of the depreciation charge.

What the FY2026 outflow is made of

The media release reports “operating cash flow before working capital movements” of A$107.2m, up 20.0%, and does not print the statutory figure. Both numbers are true and they are A$130m apart. The reconciliation is the issuer’s own, from the FY2026 statement of cash flows:

As reported. The statutory outflow is disclosed by the issuer — in the Listing Rule Appendix 7.2 commentary of the same document (“Operating cash flow was an outflow of A$23.1 million for FY2026”) — but not in the media release or the highlights page of the investor presentation, both of which carry the A$107.2m pre-working-capital figure instead.

Three separate things are inside the A$115.3m, and only one of them is a trading deterioration:

• The pass-through. The Commonwealth receivable and the matching NVL payable acquired with Civmec Defence Industries move together and are the same size. They inflate the gross movements on both the receivable and the payable lines without being a net use of cash.
• Genuine growth funding. Second-half revenue was A$522.5m against A$307.7m a year earlier — up 69.8%. A contractor that grows a half by that much funds the receivables and unbilled work that come with it, and the payables only partly offset the timing.
• Unbilled work. Contract assets rose A$122.7m to A$277.7m, or 112 days of FY2026 revenue against 70 days a year earlier. The issuer attributes this to “the timing of progress claim billings and the measurement of project progress”. This is the line that carries recognition judgement, and it is the one the auditors singled out.

The dividend was not covered by the year’s cash. Free cash flow before dividends was negative A$28.0m (statutory operating cash flow less A$5.4m of capex plus A$0.6m of disposal proceeds), against A$30.6m of dividends paid. The gap was met from the opening cash balance and from the A$16.5m net cash that arrived with the acquisition — which is a one-off, not recurring free cash flow. Cash fell A$48.3m over the year and borrowings were unchanged at A$60.0m; the company drew A$132.8m and repaid A$132.3m on the revolver during the year, so the flat closing balance conceals heavy intra-year use.

What this does and does not establish. One year of negative operating cash flow in a contractor growing its second half by 70% is what growth funding looks like, and this business has produced A$438m of operating cash over eight years. It is not, on this evidence, a cash-quality failure. What it does establish is that the FY2026 profit increase was not converted to cash in FY2026, that the conversion now depends on collecting a receivable and contract-asset book that grew A$307.6m in twelve months, and that the next print — 1H FY2027, due around February 2027 — is the first place that collection becomes observable.

Where the cash went, FY2021–FY2026

Capital allocation, as filed. The point of the table is the third and sixth rows read together: cash capex has now run at 0.6% of revenue for two consecutive years while property, plant and equipment rises every year. The property base is carried at valuation, not at cost. Note 10 of the FY2026 condensed statements states that had the revalued freehold land and buildings been carried at historical cost their net book value would be A$201.0m against a revalued carrying amount of A$453.8m — an excess of A$252.8m, or 55.7% of the carrying amount (30 June 2025, from the same note: A$205.9m against A$414.0m, an excess of A$208.1m or 50.3%). FY2026 alone booked a A$54.8m gross revaluation gain against A$5.4m of cash spent. Committed forward capex is the issuer's only forward-looking capex disclosure and remains near zero; it is not disclosed before FY2024. Correction, 27 August 2026: an earlier version of this caption gave the 30 June 2025 excess as A$254.9m and 55.3% of the class. That figure could not be reproduced from the note's own two columns and both years are now stated on the FY2026 note's basis.

Independent credit view — agency-style analytical framework

This section is SMID Research’s independent, unsolicited public credit analysis, prepared from public information. It is not a rating issued by S&P Global Ratings, Moody’s or any other credit-rating agency, and it does not copy proprietary agency grids, weights, mapping or thresholds. Regional-bank credit-research brevity and issue framing inform the presentation; no third-party Issuer Profile construct is used. The familiar letter-grade debate and scenario migration register sit only in the authenticated private vault.

Blue team and red team

Lease-adjusted credit ratios

The same gap persists at the full year. On borrowings alone Civmec is A$5.4m net debt at 30 June 2026 — A$54.6m of cash against A$60.0m drawn — having been A$42.9m net cash a year earlier. Including the A$66.4m of lease liabilities the figure is A$71.7m. The lease-inclusive number is the consistent creditor measure and it is the one that moved: A$54.2m of deterioration in twelve months, of which A$48.3m is the cash draw-down and A$5.9m the lease-liability increase from remeasurement and additions. FY2026 EBITDA is on the derived total-finance-cost basis and ties to the issuer’s published A$107.3m.

Cash waterfall — cash quality matters more than the EBITDA line

FY2025’s A$97.8m “operating cash flow” in the results release is cash generated from operations before interest and tax; statutory CFO was A$60.9m. In 1H FY2026, about A$53.0m of working capital was absorbed and cash fell despite a A$16.5m net acquisition inflow. The full year absorbed A$115.3m and the outflow more than doubled to A$23.1m, so the second half consumed a further A$12.8m rather than reversing the first. That acquisition inflow is not recurring FCF: acquired cash arrived with A$22.2m of contract liabilities and A$128.3m each of receivables and payables, and those two are the opposite sides of one Commonwealth-to-vendor pass-through. The A$5.5m annual release of the acquired contract liability that the interim statements’ stated policy implied did not occur: note 13 of the FY2026 accounts recognises no release at all in FY2026 and leaves the carrying amount at A$22.195m. FY2026 earnings therefore need no adjustment for it, and the credit is still ahead of the company on an unstated schedule.

Liquidity, bonding and claims-quality lens

• Cash and facilities. Cash was A$102.9m at June 2025, A$87.6m at December 2025 and A$54.6m at June 2026. The facility is now a single A$160m revolving multi-option limit, 39.0% utilised at 30 June 2026 (December 2025: 39.2%; June 2025: 40.5% of A$156m), leaving roughly A$97.6m unused. Gross unused limits are not treated as fully committed cash because one limit now serves cash borrowing, letters of credit and bank guarantees together, and draw conditions, maturity and security terms are incomplete. Total headline liquidity of roughly A$152m is therefore an upper bound, not a committed cash figure.
• Sureties are capacity, not cash — and this is the FY2026 line that moved most. Surety-bond utilisation went from A$183.6m of a A$400m limit (45.9%) at June 2025 to A$282.3m of a A$417m limit (67.7%) at June 2026 — a 53.8% increase in bonds outstanding against a 4.3% increase in the limit. Undrawn bonding capacity fell from A$216.4m to A$134.7m. Bank guarantees are separately A$2.4m, so total off-balance-sheet performance obligations are A$284.7m against A$60.0m of drawn debt. The exposure is not funded debt, and rising utilisation is the normal consequence of a larger order book. It matters because bonding capacity is a real constraint on how much work a contractor can carry, and because a project dispute can produce a bond call at the same time as customer withholding, liquidated damages and working-capital stress. On the current trajectory the limit, not the balance sheet, is the binding constraint to watch.
• Contract assets need collection evidence, and FY2026 raised the stakes. Contract assets were A$155.0m at June 2025, A$178.8m at December 2025 and A$277.7m at June 2026 — 112 days of revenue against 70 a year earlier. Net contract assets (contract assets less contract liabilities) rose from about A$83.5m to A$174.9m. This is not a misstatement allegation. It is why construction-contract accounting is the first of the auditors’ three key audit matters for FY2026 — the others are the recoverability of A$237.2m of trade and other receivables and A$277.7m of contract assets, and the valuation of property, plant and equipment and investment properties: A$786.8m of FY2026 revenue, 87% of the group total, is recognised on the cost-to-cost input method, and the audit procedures disclosed include challenging management’s judgement on variations and claims and assessing onerous-contract provisions on low- or negative-margin projects. EBITDA should be read alongside certification, ageing, collections and the judgement used for transaction-price adjustments.
• Asset value is not immediate liquidity. FY2026 PP&E of A$619.6m supports operating capacity, but A$252.8m of the freehold carrying amount is revaluation above historical cost, the valuations are specialised Level 3 depreciated-replacement-cost marks, and the facility is secured by a general security deed over certain of those assets. All three require a material haircut in any recovery analysis.

Neutral credit watchpoints

One of these has now triggered. The previous version of this section said evidence would weaken if “negative operating cash flow persists alongside rising contract assets or receivables”. In FY2026 operating cash flow was negative for the full year, contract assets rose A$122.7m and receivables A$184.9m. That test is met on its own terms and is recorded as met, not reworded. What it does not yet establish is persistence: this is the first full-year outflow, it followed a 69.8% increase in second-half revenue, and the acquired pass-through inflates both the receivable and the payable side.

Evidence would strengthen if the receivable and contract-asset book converts in 1H FY2027 and rolling operating cash flow after working capital turns positive and covers reinvestment and dividends; if underlying margins hold as the acquired contract liability is eventually released; if customer concentration falls; and if facility and surety limits are extended ahead of order-book growth. Evidence would weaken if a second consecutive period of negative operating cash flow arrives alongside a still-rising contract-asset book; if surety utilisation continues toward its limit without the limit moving; if a material onerous-contract or bond-call event occurs; or if covenant and facility availability begin to constrain liquidity. The covenants are stated as met at 30 June 2026 and their thresholds remain unpublished, so covenant headroom cannot be tested from the public record in either direction.

The extended workbook, private shadow band, scenario migration analysis and blue/red-team disposition are available only in the authenticated Civmec vault. Public credit evidence is sourced to the FY2026 audited report of the directors and financial statements and FY2026 condensed statements of 27 August 2026, the FY2025 audited statements, 1H FY2026 interim statements, Q3 FY2026 update and 4 June 2026 order-book announcement.

The share price and what came with its moves

Over the window Civmec returned +141.9% on a dividend-adjusted basis; the S&P/ASX 200 returned n/a and the median of the 3 listed comparisons +241.8%.

Download

A print-ready PDF of this page, for reading away from the screen: Civmec evidence library (PDF). It carries the same content as this page — as-filed history, segment economics, the order book, the balance sheet, cash conversion, the credit view and open questions — and the same omissions: no rating, no fair value, no forecast.

Sources and corrections

Where the primary sources are filed. ASX issuer page for CVL — announcements · SGX issuer page for P9D — secondary listing announcements · Civmec investors. Exchange portals re-file and purge documents over time; the text above cites each document by name and date so it can be found again. Tables on this page as data: JSON · CSV — the page’s own as-filed and recomputed tables, cell for cell, with their section and caption.

This page is built from Civmec’s own filings: fifteen annual reports covering FY2012 to FY2026 (complete since listing), the full-year and half-year results announcements from FY2012 to FY2026, the 1H FY2026 Appendix 4D and media release of 12 February 2026, the Q3 FY2026 business update of 15 May 2026, the order-book announcement of 4 June 2026, the Luerssen Australia acquisition announcements of 26 June and 1 July 2025, and the complete FY2026 results set dated 27 August by the issuer and lodged on the ASX tape at 07:45 AEST on 28 August 2026 — Appendix 4E, condensed statements for the six months and full year, audited report of the directors and financial statements, media release, investor presentation and notice of record date.

Retrieval notes. The FY2013 annual report is served as a zero-byte file by the issuer’s own website; it was recovered from the Internet Archive instead. The 2012 IPO prospectus was also recovered from the Internet Archive (359 pages) after an earlier attempt failed against an archive.org outage; its disclosures are used in the concentration and input-cost sections above.

A separate revision and methodology note records in full how this page was built, every error found in the first two versions, how each was found, and what was corrected. It contains no investment opinion.

Corrections made after publication. This page was first published on 17 August 2026 and revised on 18 August 2026 following an independent blind review of the underlying pack, which returned 37 findings including six of the highest severity. Each was verified against the primary filing before correction. The material changes to this page were: the order-book trough (published as A$760m at 31 March 2025, corrected to A$633m at 31 December 2024, and the increase from 97% to 137%); the EBITDA basis (the profit-and-loss finance-cost line excluded lease interest booked in cost of sales, understating FY2025 EBITDA by A$3.8m and overstating interest cover); a statement that the Luerssen purchase-price allocation had not been disclosed, when it had been on 12 February 2026; and the FY2018 implied exchange rate. The 1H FY2026 segment note and the acquisition note were added, having been overlooked in the first build.

Second round of corrections, 18 August 2026. A further independent review — this one guided rather than blind, meaning the reviewer was pointed at the areas the author believed weakest — returned sixteen findings, two of the highest severity. Both were re-derived from the primary filing before acceptance and both held. The changes affecting this page were:

• The single-facility claim was wrong. This page described Civmec as operating “nearly all of it out of one 79,000 m² facility”. The issuer discloses a 70,000 m² hall on a 200,000 m² site at Henderson rated at 80,000 tonnes a year, plus a 227,000 m² facility at Tomago near Newcastle rated at a further 25,000 tonnes, with newer sites at Port Hedland and Gladstone. Corrected above.
• The segment-assets statement was too absolute and the note was misquoted. Corrected in section 2, with the correction stated there rather than only here.
• Free float. Where a derived figure of roughly 60% was used, the issuer’s own statement is that approximately 54% of issued shares are in public hands (AR-FY2025, printed p114). The issuer figure is now used.
• Board turnover now carries the issuer’s stated reason. The simultaneous retirement of all three independent non-executive directors on 30 October 2024 was described without their explanation. The FY2025 directors’ report states the retirements were “to ensure compliance with tenure requirements under the SGX Listing Rules” and that a transitional period existed during which the incoming independent directors had already joined the Board. That materially changes how the event reads, and it should have been quoted the first time.
• A joint-auditor change was never disclosed here. Moore Australia Audit (WA) and Moore Stephens LLP were appointed joint auditors following the change of domicile, with FY2025 fees of A$313,000. It is now stated.
• Depreciation. An average quoted as A$18.4m is A$18.1m across FY2021–FY2025; A$18.4m was a single year’s figure carried into a sentence about an average.
• 1H FY2026 margin. The half-year gross margin of 11.8% was described as achieved without a one-off. It was restated here as including approximately A$2.8m of release from the acquired Luerssen contract liability, giving approximately 11.1% excluding it. That restatement has itself been withdrawn on 27 August 2026: the FY2026 accounts recognise no release of that liability in FY2026, so the 11.8% stands as reported. See the correction in the acquisition section above.

Two of these — the facility description and the board-turnover framing — had been published on this page and were read by anyone who visited before 18 August 2026. Neither reviewer is independent of the author’s model family, which is a real limitation on how much assurance these passes provide, and is stated rather than argued away.

Corrections made during the build. Five transcription errors were caught by arithmetic identity checks before publication and corrected against the filings: a missing share-of-joint-venture line in FY2018, a missing other-expenses line in FY2019, a missing other-credits line in FY2023, an income-tax-receivable balance attributed to the wrong year across FY2024/FY2025, and a wrong non-controlling-interest balance in FY2022. Each was re-derived from the filing’s own subtotals. They are listed with their evidence in the working pack.

This page is an evidence library with independent public credit analysis. It contains no public letter rating, fair value, forecast or scenario migration register, and it is not a recommendation or personalised financial advice. Figures labelled derived are computed from reported figures with the computation stated.