Stefanutti Stocks Holdings Limited (JSE: SSK)

Deep Dive Research Report - September 2026

A note on reporting cadence before we begin. Stefanutti Stocks has a 28/29 February financial year-end and reports half-yearly, which is the JSE norm for a company of this size. It does not hold quarterly earnings calls, and there is no transcript aggregator coverage of it. What exists instead, for each of the last six reporting periods, is a full SENS results announcement, a results presentation, and an audio/visual webcast published on the company's investor relations page. The six periods used throughout this report are: H1 FY2024 (28 Nov 2023), FY2024 (23 May 2024), H1 FY2025 (26 Nov 2024), FY2025 (27 May 2025), H1 FY2026 (25 Nov 2025) and FY2026 (26 May 2026), supplemented by a 16 February 2026 operational update. The most recent full-year release, for the year ended 28 February 2026, was published on 26 May 2026. The next release, H1 FY2027 covering the six months to 31 August 2026, is due around late November 2026 and has not yet occurred.

1. What the company does

Stefanutti Stocks builds things that are hard to build, for people who cannot afford for them to fail.

Strip away the label "multidisciplinary construction group" and what you have is a contractor that shows up on a piece of ground in southern Africa and takes single-point responsibility for turning that ground into a functioning asset. A wind farm. A data centre. A tailings storage facility at a platinum mine. A pair of 36-metre-diameter fuel storage tanks at a Durban refinery. A bulk water pipeline. A dual carriageway. The company describes its own offering as giving clients "single-point responsibility for multidisciplinary projects" (Integrated Annual Report 2026 landing page), and that phrase is doing more work than it appears to.

Most construction projects of scale are not one job. A data centre is earthworks, then piling, then concrete structure, then structural steel, then mechanical/electrical/plumbing, then commissioning. A wind farm is access roads, foundations, cable trenching, a substation, an overhead line, and the turbine towers themselves. Each of those is a different trade with a different labour pool, different plant, and different failure modes. The default industry model is that a client appoints a main contractor who subcontracts most of this, or appoints five specialists and manages the interfaces themselves. Stefanutti Stocks' pitch is that it owns enough of those disciplines in-house that the client signs one contract and the interface risk sits with the contractor rather than the client.

The founding story, and why it explains the current shape of the business.

The group traces to 1971. Two acquisitions in 2007 changed what it could sell: a controlling interest in Skelton & Plummer, a mechanical and electrical engineering business, and a controlling stake in Civil & Coastal, a civil and marine specialist. Those two deals are the origin of the "multidisciplinary" claim, because before them the group was a civils and earthworks contractor that had to subcontract anything with a cable or a pipe in it. In 2008 the group merged with Stocks Limited, producing the entity that carries both names today (company history summary via Africa Outlook). Apollo Electrical followed in 2010, along with Cycad in pipelines.

So the corporate genealogy maps almost exactly onto the service catalogue: civils and earthworks from the original Stefanutti business, building from Stocks, mechanical and electrical from Skelton & Plummer and Apollo, marine from Civil & Coastal, pipelines from Cycad. The group is a stack of acquired capabilities that were bolted together to remove subcontractors from its own supply chain.

Then it nearly died.

This is the part that matters most for anyone looking at the company today. In 2019 Stefanutti Stocks hit a severe working capital constraint, which COVID then compounded. The proximate cause was a contract at Eskom's Kusile power station, taken in joint venture with Basil Read, on which the company performed work it was not paid for. Its claim against Eskom eventually grew to R1.614 billion; Eskom's counter-position was that Stefanutti had been overpaid by roughly R1 billion (Moneyweb). A contractor cannot survive a nine-figure receivable it cannot collect, and the share price fell to 13 cents in December 2019.

What followed was a four-year restructuring run with the corporate restructuring firm Metis Strategic Advisors, which concluded in late 2025 (Engineering News, 9 December 2025). The plan had four legs: sell non-core assets, resolve the contractual claims, keep the lenders on side, and evaluate the capital structure including the possible raising of new equity. Three of the four were executed. The fourth, the equity raise, never happened, which is the single most consequential thing that did not occur in this company's history and is dealt with in Section 10.

The asset disposals were relentless and unglamorous: the 49% interest in Al Tayer Stocks LLC in the UAE for an initial R92 million, formwork yards in Midrand for R28 million, the remaining units of a Pretoria residential development for R24.4 million net, vacant industrial land in George for R1.4 million, an industrial property in Kempton Park for R1.9 million, and R58 million of underutilised plant and equipment (Moneyweb). The Mozambique business, SS-Construções (Moçambique) Limitada, together with Stefanutti Stocks Construction Limited, was finally disposed of on 12 December 2025 to East Africa Enterprises SPV FZCO for roughly US$3.2 million.

The Kusile matter resolved in two steps. On 17 October 2025 the Dispute Adjudication Board issued its Sub-clause 20.4 decision awarding Stefanutti Stocks R685 million excluding VAT. On 24 November 2025 the parties instead signed a full and final settlement for R580 million excluding VAT, a mutual release of all claims on that contract, payable by 12 December 2025 (H1 FY2026 SENS, 25 November 2025). In October 2025 the historic lender arrangement was replaced with a new five-year R850 million Standard Bank facility at JIBAR plus 3.5%, repayable in 19 quarterly instalments from February 2026 to August 2030. Settlement of the old loan on 31 October 2025 formally terminated the Restructuring Plan.

The settlement cash and the Mozambique proceeds were then used to prepay R620 million of the new facility, leaving R223 million outstanding at February 2026 against an initial R850 million. Chief executive Russell Crawford framed the bank facility, not the Eskom cheque, as the turning point:

"The securing of the Standard Bank facility... reflects a significant turning point for the group, bringing Stefanutti Stocks into a normalised trading environment." (Engineering News, 26 May 2026)

That is the correct emphasis. A construction company does not trade on cash; it trades on the willingness of banks and insurers to issue performance guarantees on its behalf. Without them it cannot bid. A contractor with a going-concern qualification and a lender running a restructuring plan is a contractor that gets excluded from the tender list before the price is even opened. Getting a clean five-year facility from a major bank is what re-admits the company to the market, and the order book response is the proof: it went from R8.6 billion at February 2025 to R17.2 billion reported with the FY2026 results, of which R6 billion is work outside South Africa.

What the company actually does for a customer, step by step.

Take the data centre work, which is the clearest illustration of the integrated model. A hyperscale or colocation client wants a facility live on a fixed date, because the revenue clock starts at commissioning and every week of delay is a week of unrecovered capital. Stefanutti Stocks' building discipline engages early, before the design is frozen, to give buildability input. It then does the bulk earthworks itself. Its geotechnical discipline does the piling itself, using its own rigs. It builds the structure. Its in-house mechanical, electrical and plumbing capability installs the services, which on a data centre is the majority of the value and virtually all of the commissioning risk. It models the whole thing in BIM so that clashes between the structure and the services are found on a screen rather than on site. The company states it has "an established track record of constructing and delivering Data Centers and associated infrastructure," using in-house MEP expertise and BIM platforms for early commissioning (Building discipline page).

The value proposition is not that any one of those steps is uniquely well done. It is that when the piling contractor blames the earthworks contractor who blames the structural contractor, on a Stefanutti Stocks job all three of them work for Stefanutti Stocks, and the client's programme does not get renegotiated.

The credentials that gate this work are specific and checkable: a CIDB Grade 9 rating, which is the highest grade in South Africa's Construction Industry Development Board system and carries unlimited tender capability, and B-BBEE Level 1 contributor status, which is the top empowerment score and is a scored criterion on virtually every public-sector and most large private-sector tenders in South Africa (About page).

2. Business segments

Stefanutti Stocks reports on a geographic basis, not a disciplinary one. There are four reportable segments: three South African regions (Inland, Coastal, Western Cape) and one Rest of Africa segment. This is a deliberate structure and it is worth understanding why before going through each.

The disciplines (building, civils, geotechnical, roads and earthworks, renewable energy, electrical and instrumentation, mechanical, oil and gas, materials handling, tailings management) are capabilities, deployed by regional offices. The company describes its operating structure as "regional offices that offer clients the full spectrum of their capabilities." A client in Cape Town deals with the Western Cape region, which can pull in the geotechnical fleet or the oil and gas pipe-spooling shop as the job requires. Reporting by region rather than by discipline reflects where the profit and loss actually sits: a regional managing director owns a portfolio of contracts across several disciplines, and it is the region that carries the tender risk, the plant utilisation, and the site overhead.

Approximate FY2026 shares of segment contract revenue, derived from the reported segment figures: Inland roughly 54%, Coastal roughly 19%, Africa roughly 15%, Western Cape roughly 12%. On a disciplinary rather than geographic cut, the group's revenue splits approximately 54% civil engineering infrastructure, 34% building construction, and 12% mechanical and electrical (MarketScreener company profile).

2.1 Inland Region

What it does. Headquartered at Kempton Park in Johannesburg, this is the group's home region and its largest by a wide margin. It covers Gauteng, Mpumalanga, Limpopo, North West and the Free State: the mining heartland, the industrial belt, the power generation complex, and the country's densest concentration of commercial and logistics property. Its work spans essentially the full disciplinary catalogue: bulk earthworks and roads, geotechnical work on mine infrastructure, mechanical and electrical installation, oil and gas fabrication and shutdown work, materials handling, tailings management, renewable energy balance-of-plant, and building including data centres.

The core capability. What Inland knows how to do that a smaller contractor cannot is mobilise several disciplines onto one industrial site simultaneously and sequence them. Mine infrastructure is the hardest version of this: a client wants a new plant tied into a live operation, which means the earthworks, the concrete, the steel, the piping and the electrical all have to happen inside shutdown windows measured in days. The oil and gas discipline's Sasol vessel replacement, three columns replaced during a 60-day, 24-hour shutdown, is the shape of the problem (Oil & Gas discipline page). You cannot buy that scheduling competence; you accumulate it over decades and it lives in a few hundred people.

Why it exists as a separate entity. Geography, plant logistics and client relationships. Heavy construction plant is expensive to move; a rig that can be trucked to a Rustenburg platinum mine in a morning cannot economically serve a job in Cape Town. Inland is also where the group's institutional client relationships sit: Eskom, Sasol, the platinum and coal miners, and Gauteng's industrial and logistics developers.

Competitive position. This is the most contested region in the country, because it is where WBHO, Concor, Raubex's construction arm and every mid-tier specialist also concentrate. Inland wins on breadth of discipline and on the CIDB Grade 9 / B-BBEE Level 1 combination that keeps it on the shortlist for the largest public and quasi-public work. It loses where the job is a single discipline at commodity price, because a focused specialist with lower overhead will undercut it.

How it fits into the group. This is the engine. It carried roughly 54% of FY2026 segment revenue and, critically, was the region that housed the Kusile settlement recognition. Its order book grew from R3.1 billion at February 2025 to R5.8 billion at February 2026, the largest absolute increase of any region (Property Wheel, May 2026). Management does not describe it as a growth bet; it is the region whose steady performance funded the survival of the group through the restructuring, having improved its operating result in every one of the six reporting periods reviewed here.

2.2 Coastal Region

What it does. Based in Pinetown, KwaZulu-Natal, Coastal covers KZN and the Eastern Cape. Its natural work is port and marine infrastructure, the Durban industrial and petrochemical complex around Island View and the refineries, KZN's logistics and warehousing property market, and Eastern Cape automotive and renewable energy.

The core capability. Marine and coastal civils, and refinery-adjacent oil and gas work. Building a tank farm or a fuel jetty is a different engineering problem from building the same structure inland: you are dealing with tidal working windows, marine corrosion, and, in the refinery case, hot-work permits inside a live hydrocarbon environment. The Transnet SAPREF R Tanks project at Island View, running from June 2019 to March 2022, involved two 36-metre-diameter storage tanks with aluminium dome roofs and internal floating roof systems, including hydrotest and finishing. That combination of large-diameter tank construction plus internal floating roofs plus hydrotest is a genuinely narrow skill set in southern Africa.

Why it exists as a separate entity. The client base is structurally different. Coastal sells to Transnet, to the refiners and fuel storage operators, and to KZN's industrial property developers. Inland's client list barely overlaps.

Competitive position. Coastal competes against national contractors on the property work and against a small number of specialist marine and mechanical contractors on the port and refinery work. On the latter it has real defensibility, because the qualification requirements (welding codes, hot-work discipline, tank construction references) are high and the reference list is short. On the former it is a price-taker.

How it fits into the group. Coastal is the region that has been most volatile across the six periods. In H1 FY2024 it was close to breakeven at the operating level on a small revenue base, a direct consequence of delayed project commencement. It recovered strongly through FY2025 as revenue nearly doubled, then saw revenue fall back in FY2026 while operating profit still improved, which is the signature of a region that has traded volume for margin discipline. Its order book rose from R1.6 billion at February 2025 to R2.6 billion at February 2026. It contributed roughly 19% of FY2026 segment revenue.

2.3 Western Cape Region

What it does. Based in Cape Town, this region serves the Western Cape's commercial property market, the Astron Energy refinery complex, the province's water and civils infrastructure programme, and, increasingly, renewable energy. The Western Cape is where South Africa's utility-scale wind resource is concentrated, and where several of the earliest Renewable Energy Independent Power Producer Procurement Programme projects were built.

The core capability. Renewable energy balance-of-plant, and integrated building. On renewables the group offers civil and electrical balance-of-plant design, procurement and construction; overhead line construction up to 132 kV; high-voltage substation construction for both main transmission and IPP applications; underground cable networks; battery energy storage system construction across all trades; and, distinctively, manufacture of pre-cast concrete wind turbine generator towers, which it did at the Nxuba Windfarm near Somerset East (Renewable discipline page). It also built the power block civils on the 100 MW Kathu concentrated solar plant, and delivered a complete substation and 132 kV line at the Sere Wind Farm in consortium with Powertech.

Pre-cast concrete towers are worth pausing on. The standard wind turbine tower is fabricated steel, usually imported. A concrete tower is cast on or near site, which removes an import line item and a set of abnormal-load transport constraints, and it allows greater hub heights. It requires a casting yard, moulds, and post-tensioning competence. Very few contractors in southern Africa have done it.

Why it exists as a separate entity. Distance and a distinct market. Cape Town is roughly 1,400 km from Johannesburg, which makes shared plant and shared site management impractical. The Western Cape's construction market also behaves differently from Gauteng's: more commercial and residential property, more water infrastructure, more renewables, less mining.

Competitive position. The renewable balance-of-plant market is where this region has the most defensible position, because the qualification set (grid-code-compliant substations, 132 kV overhead line, and heavy civils in one contractor) is unusual. On general Cape Town building it competes with the same national contractors plus a healthy local mid-tier.

How it fits into the group. Western Cape is the smallest South African region by revenue, roughly 12% of FY2026 segment revenue, but it carries the group's clearest exposure to the energy transition. Its FY2026 revenue was held back by delays in project implementation and its operating result included a R14 million impairment on land and buildings. Its order book nonetheless rose from R2.7 billion at February 2025 to R3.1 billion at February 2026, meaning it entered FY2027 with more forward work than any region other than Inland relative to its size.

2.4 Rest of Africa

What it does. After the December 2025 disposals, this segment is essentially Botswana, Zambia and Eswatini, with cross-border project work into Namibia and Zimbabwe. Stefanutti Stocks Botswana, Zambia and Eswatini are established local operating companies rather than project offices, described by the group as mature local contractors in their respective markets. The work is building, civil works, bulk earthworks and mining infrastructure. In Botswana the group built what it describes as the continent's first 6-star Green Star rated building.

The core capability. Being a local contractor rather than an expatriate one. In Botswana and Zambia, public and mining clients heavily favour registered local entities with local labour, local plant and a local track record. Having spent years building those entities is the capability, and it cannot be replicated by flying in a South African team for a single tender.

Why it exists as a separate entity. Currency, tax, regulatory regime and local content rules are all different, and the segment carries a risk the South African regions do not: contracts are frequently denominated in US dollars while costs are incurred in local currency and in rand.

Competitive position. In Botswana and Eswatini the group is a genuine incumbent. In Zambia it competes with Chinese state-linked contractors on price, which it will generally lose, and wins on mining-client work where technical execution and safety record matter more than headline price.

How it fits into the group. Africa is simultaneously the group's biggest FY2026 disappointment and its biggest FY2027 opportunity. It swung to an operating loss in FY2026, and the group attributes this to the strengthening of the Zambian kwacha against both the rand and the US dollar, which compressed gross profit on dollar-denominated contracts (Property Wheel, May 2026). At the same time its order book went from R833 million at February 2025 to R4.1 billion at February 2026, a roughly fivefold increase and the largest proportional move in the group. This is the segment that turns the "R6 billion of the order book is outside South Africa" statement into something real.

Segment comparison

3. Products and business detail

Construction "products" are capabilities plus plant plus certifications. Here is the full catalogue.

Building, including data centres

The group offers what it describes as "the full scope of traditional and niche building expertise," covering industrial and commercial buildings (warehousing, production facilities, distribution centres, offices), retail, hospitality and leisure, healthcare, and data centres. Critically it bundles into a building contract the things most building contractors subcontract: bulk earthworks, piling and geotechnical work, structural steel including specialised structures, MEP systems, civils, BIM modelling, and commercial management.

It offers three procurement models, which is unusual for a southern African contractor:

• Traditional / turnkey design-and-build, where the contractor carries design responsibility.
• Equity participation and co-development, where the contractor takes a stake in the asset it is building. This is a meaningfully different risk posture: it converts a fee-earning contract into a partial development exposure, and it is a way of winning work when a developer is short of equity.
• Early contractor engagement, where the contractor is paid to advise on buildability before the design is frozen.

Named building references include Park Square in Umhlanga Ridge, KZN (2016-2018); Mercedes-Benz South Africa facilities in East London (2017-2019); the Newlands Cricket Ground development in the Western Cape (2019-2022); and the iThemba LABS cyclotron installation in the Western Cape (2021-2022). The cyclotron job is the interesting one: installing a particle accelerator requires millimetre-level tolerances and radiation shielding, which is a very different problem from putting up a warehouse.

Geotechnical

This is the discipline with the most identifiable physical asset base and it is worth being specific, because plant ownership is the barrier to entry in ground engineering.

Services: piling in all soil conditions including design; lateral support systems; rock anchoring and shotcrete; void fill and specialised grouting (consolidation, compaction and dolomitic treatment); sinkhole remediation; and geotechnical investigation and reporting.

Fleet: 11 percussion and rotary drill rigs in the 5 to 22 tonne range for slope stabilisation and investigation work; 11 crawler-mounted hydraulic piling rigs in the 30 to 80 tonne range capable of auger, cased and continuous flight auger piling; and 4 dedicated investigation and exploration rigs (Geotechnical discipline page).

Dolomitic ground treatment deserves a note because it is a distinctly South African competence. Large parts of Gauteng and the North West sit on dolomite, which dissolves and forms sinkholes. Building on it requires specialised grouting and a documented methodology, and it is a regulated risk. Contractors that can do it, and can put their name to the geotechnical report, have a market that contractors without the rigs simply cannot enter.

References include piling for the Wilmar processing plant at Richards Bay (2020-2021), the Swakopmund River Bridge on a Namibian railway upgrade (2021), the Kazerne transit-oriented development in Johannesburg (2015), and Point Promenade in Durban (2018). Geographic reach spans Richards Bay, Mpumalanga, Cape Town, Durban and Limpopo in South Africa, plus Zambia (Livingstone and Lusaka), Namibia, Lesotho, Eswatini and Botswana.

Roads and earthworks

Bulk earthworks, blasting and excavation for all infrastructure types, rehabilitation of all road infrastructure, and pipeline construction. This is the group's oldest capability and the one with the most direct exposure to SANRAL and provincial roads budgets.

Renewable energy

Covered in Section 2.3. To restate the catalogue precisely: civil and electrical balance-of-plant design, procurement and construction on wind and solar farms; battery energy storage system construction across all trades; main transmission and IPP substation construction including earthworks, foundations, buildings and electrical installation; overhead line construction up to 132 kV; underground cable networks; pre-cast wind turbine generator concrete towers; in-house MEP execution; BIM 4D and 5D planning; and access to the group's pipe-spooling fabrication facility. The group cites experience on Eskom Self-Build projects.

Oil and gas

Fabrication, construction and maintenance of process plants. Services: structural steel erection, mechanical equipment installation, pipe spool fabrication, process piping installation, electrical and instrumentation, shutdown and maintenance, tank farms and fuel jetties, and scaffolding, painting and insulation.

The pipe spool fabrication facility is the group's one genuine manufacturing asset. Capacity is stated at 50,000 diameter inches per month. It is coded to ASME B31.1, EN ISO 15614-4, AWS D1.1 and API 1104, and handles carbon steel, stainless steel, chrome alloys, aluminium and ASTM A335 Grade P91.

That last material matters. P91 is a creep-strength-enhanced ferritic steel used in high-temperature, high-pressure steam service, which is to say power station and refinery piping. Welding it correctly requires tightly controlled pre-heat, interpass temperature and post-weld heat treatment; get it wrong and the weld fails in service years later. A fabricator qualified on P91 to ASME B31.1 is qualified for work that most fabrication shops cannot legally touch. This is a certification moat, narrow but real, and it is the reason the group can bid refinery and power plant piping packages.

References: Transnet SAPREF R Tanks at Island View (June 2019 to March 2022), Astron Energy Crude Flex in Cape Town (2021, interconnecting piping across refinery tanks with mechanical, civil and E&I scope), and the Sasol three-column vessel replacement in a 60-day continuous shutdown.

Mechanical, electrical and instrumentation

Mechanical work including water and wastewater treatment plant and mining water clarification; electrical and instrumentation installation. This capability is what allows the building and renewables disciplines to keep MEP in-house.

Civils, marine and water infrastructure, and concrete rehabilitation

Marine contracting, water infrastructure, and concrete repair and rehabilitation. Concrete rehabilitation is a quietly attractive niche: South Africa has a very large stock of ageing reinforced concrete infrastructure, and repair work is smaller, more frequent and less price-competed than new build.

Materials handling, tailings management and contract mining

Open-pit contract mining, mine residue disposal, materials handling installations and tailings storage facility construction. Tailings is a structurally growing market for a specific reason covered in Section 6.

Affordable housing

Retained as a discipline; the smallest and most policy-dependent of the group's offerings.

Geographies

Operations span all nine South African provinces, with regional offices at Kempton Park (Inland), Pinetown (Coastal) and Cape Town (Western Cape). Established local operating companies exist in Botswana, Eswatini and Zambia, with cross-border project execution into Namibia and Zimbabwe. Mozambique, Mauritius and the UAE have all been exited. Roughly 88% of project locations were in South Africa as at the February 2026 operational update, against 12% outside (Engineering News, 16 February 2026), though the forward order book is more international than that at roughly 35% outside South Africa.

Milestones that changed the business

• 1971: founding of the original Stefanutti business.
• 2007: Skelton & Plummer (mechanical and electrical) and Civil & Coastal (civil and marine) acquired, creating the multidisciplinary offering and opening the Western Cape.
• 2008: merger with Stocks Limited creating Stefanutti Stocks.
• 2010: Apollo Electrical acquired; Cycad added pipelines.
• 2019: working capital crisis; share price to 13 cents.
• 2021: Al Tayer Stocks (UAE) disposal; property and plant disposals begin.
• 17 October 2025: DAB awards R685 million on Kusile.
• 31 October 2025: historic lender loan settled; Restructuring Plan formally terminated; new R850 million Standard Bank facility in place.
• 24 November 2025: R580 million full and final Kusile settlement with Eskom.
• 12 December 2025: Mozambique and Stefanutti Stocks Construction disposals implemented.
• February 2026: Standard Bank facility reduced to R250 million from R850 million; annual interest charge expected down roughly 70%.

Scale

The group employed 6,336 people as at December 2025, of whom 4,473 were in South Africa, and distributes over R2 billion a year in wages (Engineering News, 9 December 2025). A more recent profile puts headcount at 6,489.

4. Customers

Who buys

Four broad buyer types, in descending order of importance.

Private industrial and commercial clients are the largest group. As at the February 2026 operational update, 78% of contract revenue came from the private sector and 22% from the public sector. This is an important and under-appreciated fact about the company: despite the Eskom history, Stefanutti Stocks is predominantly a private-sector contractor. Within this group sit mining houses (platinum, coal, gold, and Zambian copper), industrial and petrochemical operators (Sasol, Astron Energy, the fuel storage operators), logistics and commercial property developers, and data centre owners.

Independent power producers and their EPC contractors buy the renewable energy balance-of-plant work. This is a growing category driven by the REIPPPP and by private wheeling arrangements.

State-owned enterprises and government, principally Eskom, Transnet, SANRAL, water boards and provincial departments. This is 22% of revenue and is where both the largest single contracts and the largest single counterparty risk sit.

Foreign governments and mining clients in Botswana, Zambia and Eswatini, served through the local operating entities.

Who makes the decision, and on what criteria

The buying process differs sharply by customer type.

For public-sector work in South Africa, procurement is governed by a formal tender process with a scored evaluation. The gating criteria are CIDB grade (Grade 9 is required for the largest contracts and gives unlimited tender capability), B-BBEE level (Level 1 is the maximum score and is worth real points), financial standing sufficient to obtain performance guarantees, and demonstrated reference projects of comparable scale. Price then does most of the remaining work. The decision-maker is a bid adjudication committee, not an individual, and the process is documented and challengeable. Sales cycles from advertisement to award routinely run six to eighteen months, and can be extended by objections.

For mining and petrochemical clients, the decision typically sits with a project director and a supply chain function, working from a pre-qualified vendor list. Entry to that list requires audited safety performance (lost time injury frequency rate is a hard screen at most mining houses), the relevant technical certifications, and prior work on the site or a comparable one. Once on the list, individual packages are competitively bid but among a small field. Sales cycles are shorter, three to nine months, and repeat work is common.

For property developers and data centre owners, the decision is made by the developer's project executive and cost consultant. The criteria are programme certainty, price, and increasingly the contractor's willingness to take design responsibility. On data centres, ability to commission the MEP on time dominates, because the client's revenue starts at handover.

Why they choose Stefanutti Stocks

Four specific reasons, not generic ones.

1. Interface risk transfer. For a client running a multi-discipline project, the alternative to a single integrated contractor is managing five interfaces and arbitrating five sets of delay claims against each other. The group's whole pitch is that it absorbs that.

2. Certifications that narrow the field. CIDB Grade 9 and B-BBEE Level 1 together are held by a small number of South African contractors. On the oil and gas side, the P91 and ASME B31.1 pipe-spool qualifications narrow it further.

3. Owned specialist plant. A client who needs 80-tonne crawler piling rigs on a dolomitic site does not have many people to call. Owning 11 of them rather than hiring is both a cost advantage and a programme-certainty advantage.

4. Local incumbency outside South Africa. In Botswana, Zambia and Eswatini, being a registered local contractor with local labour is often a formal procurement requirement rather than a preference.

Switching costs

Honest answer: for most of the work, low. Construction is a project business. Each contract is separately tendered, and a client who is unhappy simply does not invite the contractor to the next tender. There is no installed base, no software lock-in, no consumable annuity.

Where switching costs do exist, they are of three kinds. First, pre-qualification: getting onto a mining house's or refiner's approved vendor list requires an audit process that takes months and that the client has no incentive to repeat. Second, site knowledge: on a live refinery or an operating mine, a contractor that has worked on the site knows the permit systems, the isolation procedures and the physical layout, which is worth real programme time. Third, framework and maintenance agreements: shutdown and maintenance work in oil and gas tends to be awarded on multi-year frameworks, which is the closest thing in this business to recurring revenue.

Concentration

The single most important concentration fact is historical and cautionary. Eskom's Kusile contract was a large enough exposure that non-payment on it nearly destroyed the company. That is what customer concentration risk looks like in construction: it is not that a customer stops buying, it is that a customer disputes what it owes on a contract already performed, and the contractor has to fund the working capital while an adjudication runs for years. The Kusile claim ran from late 2018 to a settlement in November 2025.

Today the exposure looks structurally better. The order book is spread across four regions, no single region has less than 15% of the total order book, and 78% of revenue is private sector rather than concentrated in one state utility. But no individual customer concentration disclosure is available in the sources reviewed, and the nature of the business means a single very large contract can always become a disproportionate share of revenue.

Contract structures and what they mean for revenue predictability

Most work is fixed-price or remeasurable lump-sum construction contracts, priced at tender and recognised as work progresses. This is the least forgiving contract form in existence: the contractor prices the risk of ground conditions, weather, labour, materials escalation and design change, and then owns the outcome. A variation to the scope requires the client to agree, in writing, before the work is done, and where the client does not agree the contractor either stops or performs and claims. The Kusile dispute is exactly this mechanism failing at scale.

Layered on top are:

• Framework maintenance and shutdown agreements in oil and gas, which are the most predictable revenue in the group.
• Design-and-build turnkey contracts, which carry design liability but usually better margin.
• Equity participation arrangements on building projects, where the group takes a development stake.
• Joint ventures, used on large or specialist contracts (Kusile with Basil Read; Kusile Package 28 with Izazi; the Hyvec JV in Mauritius). JVs spread risk but create their own exposure: the group recognised a R30 million loss on the Hyvec JV in FY2025.

Revenue predictability therefore comes almost entirely from the order book, which is why the company reports it prominently and why its trajectory dominates Section 7. As at the FY2026 results the order book stood at R17.2 billion, of which R8.5 billion related to periods beyond February 2027, meaning roughly half the backlog extends past the coming financial year.

5. Competitive landscape

The structure of this industry has changed more than the companies in it

South African construction is not a normal competitive market. It is the aftermath of a market. Over roughly the past decade, gross fixed capital formation fell to record lows and the industry's large contractors were destroyed one after another: Murray & Roberts, Group Five, Aveng and Basil Read among them (Foord, "The SA construction sector").

The most recent casualty is instructive. Murray & Roberts Holdings placed its South African subsidiary and its OptiPower trading division into business rescue in November 2024, its shares were suspended, and the holding company was delisted from the JSE on 19 January 2026 with the company acknowledging there was no prospect of returning value to shareholders (Business Day, 23 December 2025). Its construction business had already been sold in 2016 to the Southern Palace consortium and now trades as Concor.

The result is a market with far fewer credible bidders for large, technically complex work than there were ten years ago, at exactly the moment public infrastructure budgets are being restored. Foord's assessment is that the sector "has become less crowded, more stable, and potentially more profitable for the survivors."

Named competitors

Wilson Bayly Holmes-Ovcon (WBHO) is the benchmark. It is the largest surviving listed South African contractor, with revenue of approximately R28.5 billion and an order book of approximately R37.6 billion reported in 2026. It overlaps with Stefanutti Stocks across building, civils and roads, and it is the contractor most likely to be sitting opposite on any large South African tender. WBHO exited Australia after losses from aggressive pricing and problem contracts, and has since run a deliberately risk-averse strategy focused on margin rather than volume. Against WBHO, Stefanutti Stocks loses on balance sheet strength and on the sheer scale of contract it can absorb; it competes on multi-discipline breadth (WBHO subcontracts more of its MEP and geotechnical work) and on B-BBEE Level 1 status.

Raubex Group is a roads and infrastructure specialist with a materials business (quarries, asphalt, bitumen) that Stefanutti Stocks does not have. That vertical integration into materials is a structural cost advantage on roads work specifically, and it is why Stefanutti Stocks' roads and earthworks discipline is not the sharpest edge in the group. Raubex also has a large infrastructure division doing energy and telecoms work, which overlaps with the renewables balance-of-plant offering.

Aveng is a survivor rather than a thriving competitor. Its listed equity is now very small, though it reports approximately R37.5 billion of committed work in hand, largely through its Australian McConnell Dowell business and its Moolmans contract mining arm. Moolmans is the direct overlap: open-pit contract mining, where Stefanutti Stocks' mining services offering competes. On South African building and civils, Aveng is much less present than it was.

Concor (private, owned by the Southern Palace consortium) is the former Murray & Roberts construction business. It competes directly on civils, roads and building, and carries the same B-BBEE positioning as Stefanutti Stocks, which neutralises one of the group's advantages on public work.

Chinese state-linked contractors (CRJE, China Civil Engineering Construction Corporation, CHEC and others) compete aggressively for large public infrastructure and for work in Zambia and Botswana. They bid on price with access to concessional finance. Stefanutti Stocks does not generally win price fights against them and should not try.

Mid-tier specialists compete discipline by discipline: private building contractors like Tiber and Trencon in Gauteng, specialist piling contractors, specialist mechanical and E&I contractors, and marine specialists on the coast. Individually small, collectively they set the price on single-discipline work.

Market capitalisations sourced from AFX as at 8 September 2026 and included solely as a peer-size reference.

Barriers to entry

Higher than they look, and the barrier is financial rather than technical.

Performance guarantees are the gate. A client on a large contract requires a bond, typically 10% of contract value, issued by a bank or an insurer. The guarantor underwrites the contractor's balance sheet, not its skill. This means a new entrant, or a contractor in distress, simply cannot bid for large work regardless of competence. Foord identifies exactly this as the barrier: performance guarantees "create substantial barriers to entry by requiring firms with substantial balance sheets."

This is the single most important fact for understanding the Stefanutti Stocks investment case. During the restructuring the group's guarantee capacity was constrained, which capped the size of work it could pursue. The new Standard Bank facility and the removal of the going-concern uncertainty restored it, and the order book roughly doubled within twelve months. The barrier that nearly excluded the company is now working in its favour.

Secondary barriers: CIDB Grade 9 requires demonstrated financial capacity and completed reference projects of scale, and cannot be bought. B-BBEE Level 1 requires ownership, management, skills development and procurement structures built over years. The technical certifications (ASME B31.1 and P91 welding qualifications, dolomitic ground treatment methodology, marine works references) each take years and reference projects to establish. Specialist plant, particularly the piling fleet, requires capital and, more limiting, operators.

What is not a barrier: general building and general earthworks. Those are commoditised and always will be.

Market share and structural shifts

No reliable market share data is published for South African construction, and any figure would be misleading given the fragmentation below the top tier. What can be said with confidence is that the number of contractors able to take on a R1 billion-plus multi-discipline contract in South Africa is now in single digits, and was in low double digits a decade ago.

Three shifts are live:

1. Consolidation by attrition is largely complete. Murray & Roberts' delisting in January 2026 was probably the last major exit; the survivors are WBHO, Raubex, Concor, Stefanutti Stocks and a much-reduced Aveng.
2. Demand composition is changing. Energy, water and logistics are growing faster than commercial property, and renewables balance-of-plant is a category that barely existed ten years ago.
3. Client counterparty quality is improving. The shift toward private-sector and IPP clients, at 78% of revenue, reduces the payment risk that historically destroyed contractors here.

Where Stefanutti Stocks is strong, and where it is exposed

Strong: multi-discipline integration on complex industrial sites; oil and gas fabrication and shutdown work with certifications few competitors hold; geotechnical work backed by an owned rig fleet; renewables balance-of-plant including the unusual pre-cast concrete tower capability; genuine local incumbency in Botswana, Zambia and Eswatini; B-BBEE Level 1 and CIDB Grade 9.

Exposed: it is materially smaller than WBHO and therefore has less capacity to absorb a single bad contract; it has no materials vertical integration like Raubex; its balance sheet, while transformed, is only twelve months clear of a restructuring; and its Africa segment is currently loss-making on currency translation.

There is no wide moat here. This is a competent, specialised contractor in a consolidated market with a restored balance sheet. That is a real position, and it is not the same thing as a moat.

6. Industry

What drives demand

Demand for what Stefanutti Stocks sells is almost entirely a function of capital expenditure decisions by four types of buyer: the South African state and its state-owned enterprises, mining companies, energy developers, and private property and industrial investors. It is derived demand, it is lumpy, and it turns on the availability of capital rather than on any consumer behaviour.

Industry size and trajectory

The South African infrastructure cycle turned in the 2026 Budget. Public sector infrastructure spending over the 2026 medium-term expenditure framework was estimated at R1.07 trillion, with state-owned companies the largest single contributor at a projected R455 billion over three years, provinces at R217.8 billion and municipalities at R205.7 billion (IOL, 25 February 2026).

By sector over the same three years:

• Transport and logistics: R417.6 billion, the largest slice. SANRAL is focused on network resilience, maintaining roughly 27,000 km of roads and resurfacing 2,000 km annually. Transnet plans R76.6 billion.
• Energy: R213.6 billion.
• Water and sanitation: R185.2 billion. Phase 2 of the Lesotho Highlands Water Project is under way; the Trans-Caledon Tunnel Authority had raised R24 billion against a revised capital estimate of R53 billion.
• Health: R65.4 billion; education: R58.5 billion.

On the energy side specifically, 7,741 MW of contracted independent power producer capacity was operational by December 2025, with a further 1,610 MW under construction expected to reach commercial operation between 2026 and 2028.

Each of those line items maps onto a Stefanutti Stocks discipline: roads and earthworks to SANRAL; civils, marine and mechanical to Transnet; renewables balance-of-plant and substations to the energy allocation; mechanical, water treatment and pipelines to water and sanitation; building to health and education.

Where the company sits in the supply chain

Stefanutti Stocks is a tier-one main contractor. It sits below the client and the consulting engineer, and above the specialist subcontractors and materials suppliers. Unusually, it has pushed downward into work most main contractors subcontract (piling, MEP, pipe spool fabrication), and upward into design responsibility on turnkey contracts and into equity on co-development deals. It has not pushed into materials, which is the one direction Raubex has gone and it has not.

Import dynamics

Construction is largely a domestic-input industry: labour, aggregate, cement and concrete are local. The import-exposed inputs are steel (structural and pipe), specialist mechanical equipment, and imported wind turbine components. The group's pre-cast concrete wind turbine tower capability is a direct import-substitution play: it replaces imported fabricated steel towers with locally cast concrete ones, which removes both an import bill and an abnormal-load logistics problem.

The relevant import dynamic in the services market is different and more important: Chinese state-linked contractors bidding South African and SADC public work with concessional finance behind them. That is import competition for the contract itself, not for the input, and it is a permanent feature of the large public tender market.

Regulation

Three regulatory regimes shape this market:

• CIDB grading. The Construction Industry Development Board grades contractors from 1 to 9 by class of work. Grade 9 carries unlimited tender capability. Public bodies may not award contracts above a contractor's grade. This is a hard legal constraint on who can bid.
• B-BBEE. Broad-Based Black Economic Empowerment scorecards are a weighted criterion in public procurement and a common requirement in private procurement by large corporates. Level 1 is the top rating.
• Health, safety and environmental. Construction Regulations under the Occupational Health and Safety Act, plus mine-site regulation under the Mine Health and Safety Act for mining work. Lost-time injury frequency rate is a pre-qualification screen at most mining and petrochemical clients, so safety performance is directly commercial, not merely ethical.

Additionally, the industry carries the legacy of the Competition Commission's collusive tendering investigation and the resulting Voluntary Rebuild Programme, which sits in the background of every large contractor's relationship with the state.

Cyclicality

Severely cyclical, and with a lag. Construction demand follows fixed investment, which follows business and government confidence, which follows the economic cycle by roughly 12 to 24 months. Order books are a lagging confirmation of a turn that has already happened and a leading indicator of revenue 12 to 36 months out.

The cycle is asymmetric in a way that matters. On the way up, contractors add overhead and plant and win work. On the way down, revenue falls but fixed overhead does not, margins go negative, and the contractors that survive are those with the strongest balance sheets rather than the best operations. This is precisely why an entire generation of South African contractors failed: the downturn from roughly 2015 lasted long enough that balance sheets, not skills, determined survival.

Tailwinds and headwinds

Tailwinds: the R1.07 trillion public infrastructure programme; the energy transition and the IPP build-out; grid investment, since South Africa cannot connect new generation without transmission and substation capacity; water infrastructure backlogs that have become politically unavoidable; data centre construction driven by cloud and AI capacity demand in South Africa; and the reduced competitive field after a decade of consolidation.

Headwinds: chronic slippage between budget allocation and actual project award in South African public procurement, so the R1.07 trillion is an allocation and not a pipeline; municipal and SOE payment risk; skills attrition after a decade of industry contraction, since experienced site agents and engineers who left the sector do not come back quickly; Chinese contractor price competition on public work; and currency volatility on dollar-denominated African contracts.

7. Growth triggers

Extracted from the six results releases and the February 2026 operational update. Because Stefanutti Stocks does not hold quarterly earnings calls, "concall" here means the corresponding results announcement and presentation.

• Order book roughly doubled to R17.2 billion, from R8.6 billion a year earlier, with R6 billion of it outside South Africa. This is the largest single forward indicator the company has disclosed. (FY2026 results, 26 May 2026)

• R8.5 billion of the order book relates to periods beyond February 2027, meaning roughly half the backlog is multi-year work rather than the coming year's revenue. (FY2026 results, 26 May 2026)

• The Rest of Africa order book grew from R833 million to R4.1 billion, a roughly fivefold increase, giving the segment its largest forward book in years despite its current operating loss. (FY2026 results, 26 May 2026)

• Inland order book grew from R3.1 billion to R5.8 billion, the largest absolute regional increase. (FY2026 results, 26 May 2026)

• Short-term potential awards of R12.5 billion and identified prospects of R144 billion were disclosed as the pipeline behind the order book. This is the first time in the six periods reviewed that a pipeline of this scale has been quantified. (Operational update, 16 February 2026)

• Standard Bank facility reduced from R850 million at inception to R250 million by end January 2026, cutting the anticipated annual interest charge by approximately 70%. Interest saved flows straight to earnings without any new work being won. (Operational update, 16 February 2026)

  "Outstanding facility with Standard Bank has been reduced from an initial loan of R850-million to R250-million by the end of January." (Engineering News, 16 February 2026)

• New five-year R850 million Standard Bank facility signed, historic lender loan settled on 31 October 2025, formally terminating the Restructuring Plan. This is the trigger that restores guarantee capacity and therefore bidding capacity. (H1 FY2026 results, 25 November 2025; repeated at FY2026 results, 26 May 2026)

  "The securing of the Standard Bank facility... reflects a significant turning point for the group, bringing Stefanutti Stocks into a normalised trading environment." (CEO Russell Crawford, FY2026 results, 26 May 2026)

• Kusile settlement with Eskom concluded at R580 million excluding VAT, full and final, payable by 12 December 2025, releasing both cash and management attention that had been consumed since 2018. Flagged as pending across five consecutive releases before landing. (H1 FY2026 results, 25 November 2025)

• Disposal of SS-Construções (Moçambique) and Stefanutti Stocks Construction implemented 12 December 2025, completing the disposal programme and removing the last held-for-sale overhang, which had been carried since August 2020. Repeated as an outstanding item across all six reporting periods before completion. (H1 FY2026 results, 25 November 2025; confirmed FY2026 results, 26 May 2026)

• Exit from Mozambique and Mauritius refocuses the group on core South African operations and selectively higher-margin African projects. (FY2026 results, 26 May 2026)

• Revenue mix disclosed as 78% private sector, 22% public sector, with 88% of project locations in South Africa. Management disclosed this composition for the first time in the reviewed period, framing the private-sector weighting as a deliberate risk posture. (Operational update, 16 February 2026)

• Coastal region order book grew from R1.6 billion to R2.6 billion and Western Cape from R2.7 billion to R3.1 billion, giving both regions more forward work than at any point in the reviewed period. (FY2026 results, 26 May 2026)

Trigger summary

8. Key risks

1. Contract execution risk on a doubled order book (high probability, moderate to severe impact)

This is the central risk and it is a direct consequence of the good news. The order book roughly doubled in twelve months. A contractor that doubles its forward workload has to double its site management capacity: project managers, site agents, quantity surveyors, safety officers, plant. After a decade of industry contraction, that talent is genuinely scarce in South Africa. A contractor that wins work it cannot properly staff prices it correctly and then loses money executing it.

The mechanism is well documented in this industry, and its most recent local example is WBHO's Australian experience, where aggressive pricing and problem contracts produced losses large enough to force a market exit. Fixed-price contracts have no mechanism to recover a management failure. One badly executed R500 million contract can erase a year of profit from three good ones.

Calibrate this as the highest-probability material risk in the report. It is not a question of whether some contracts underperform; it is a question of how many and how badly.

2. Concentration of FY2026 earnings in a non-recurring item (certain, already occurred, forward-looking implication)

The Kusile settlement contributed R448 million to contract revenue, R388 million to operating profit and R132 million to investment income in FY2026. Excluding the settlement and a R27 million impairment, underlying operating profit was R328 million against R333 million the prior year (Business Day, 26 May 2026).

Read that again. Stripping out the settlement, the underlying operating result was essentially flat year on year. The Ghost Mail commentary made exactly this point, warning against taking the headline earnings at face value because "the settlement with Eskom is a non-recurring item that you would want to strip out" (Ghost Mail).

The mechanism by which this becomes a risk is straightforward: FY2027 has no settlement in it. Everything has to come from operations and from the interest saving. The order book says the volume is coming; the flat underlying FY2026 result says the operating leverage has not yet been demonstrated.

3. Africa segment currency exposure (medium probability, moderate impact)

The Africa segment swung to an operating loss in FY2026 because the Zambian kwacha strengthened against both the rand and the US dollar, compressing gross profit on dollar-denominated contracts. This is a structural mismatch: revenue in dollars, costs in kwacha and rand. It is not hedgeable in any complete way on multi-year construction contracts, and it now sits on a segment whose order book has grown roughly fivefold to R4.1 billion.

That combination is the risk: the segment is about to run substantially more volume through the same currency mismatch. If the kwacha stays strong, the loss scales with the volume.

4. Public sector payment and dispute risk (medium probability, severe impact)

The Kusile episode is the template. A state-owned client disputes what it owes on work already performed. The contractor funds the working capital gap while a dispute adjudication board process runs. In this case it ran from late 2018 to a November 2025 settlement, roughly seven years, and it took the company to the edge of business rescue.

22% of revenue is public sector, and there is a residual open matter: the group remains involved in a dispute with Eskom relating to Kusile Package 28, awarded to the Stefanutti Stocks-Izazi joint venture, which Eskom terminated in February 2019 due to its inability to provide site access for the JV to complete the works. The November 2025 settlement was a full and final release on the building contract; the Package 28 matter is separate. Its status was not clarified in the sources reviewed for this report.

The reason to rate this severe rather than moderate is that the loss is not the contract margin. It is the working capital, over years, plus the guarantee capacity consumed, plus the management attention.

5. Balance sheet is recently repaired, not yet proven through a cycle (medium probability, severe impact)

Management themselves flagged the position bluntly in the H1 FY2026 release:

"current liabilities exceed its current assets by R1 241 million" (H1 FY2026 SENS, 25 November 2025)

That was six months before the FY2026 results. The Kusile cash and the disposal proceeds materially improved it, and the facility is down to R223 million. But this is a company that as recently as November 2024 disclosed that total liabilities exceeded total assets by R94 million and carried a material uncertainty over going concern (Engineering News, 26 November 2024).

The mechanism by which this matters is the guarantee gate described in Section 5. If the balance sheet weakens again, guarantee capacity contracts, and the company is excluded from exactly the large work its order book now depends on. The recovery is real; it is also twelve months old.

6. Skills attrition and the cost of scaling (high probability, moderate drag)

Related to risk 1 but distinct. A decade of industry contraction pushed experienced construction professionals out of the sector, into mining, into consulting, or offshore. They do not return quickly. The constraint on the company converting a R17.2 billion order book into good margins is not plant and it is not capital; it is people who have run a R500 million multi-discipline project before. Expect this to show up as margin pressure and slower revenue recognition rather than as a single dramatic event.

7. Order book concentration in a small number of large contracts (unknown probability, potentially severe)

The order book grew by roughly R8.6 billion in one year across four regions. No disclosure available to this report breaks the order book down by contract, so it is not possible to say whether it is 40 contracts or 8. If it is concentrated, the execution risk in point 1 concentrates with it. This is a genuine information gap rather than a known problem, and it is the first question to put to management.

8. Public infrastructure allocation is not the same as awarded work (high probability, moderate impact)

The R1.07 trillion MTEF allocation is a budget line. South African public procurement has a persistent record of slippage between allocation, tender, award and site handover. A contractor that staffs up for an infrastructure programme that arrives two years late carries the overhead in the meantime. The 78% private-sector revenue mix limits this exposure, but the order book's growth and the R144 billion "identified prospects" figure both imply some dependence on the public programme actually converting.

9. Walk the talk

The six reporting periods used: H1 FY2024 (released 28 November 2023), FY2024 (23 May 2024), H1 FY2025 (26 November 2024), FY2025 (27 May 2025), H1 FY2026 (25 November 2025), and FY2026 (26 May 2026). The most recent is 106 days old, which is the correct and only answer: this company reports half-yearly, and the next release, H1 FY2027, is not due until roughly late November 2026.

A caveat before the assessment. Because there are no earnings calls, there is no forum in which management is pressed on guidance by analysts. What exists instead is a written commentary and outlook section in each SENS release, plus a results presentation. Management's "promises" are therefore narrower and more concrete than a typical guidance regime: they consist largely of stated outstanding items under the Restructuring Plan, each of which either got done or did not. That makes this an unusually clean credibility test, because the list was short, public, and repeated every six months.

Where it starts: November 2023. At H1 FY2024 the group had a R6.5 billion order book and had just been through a period in which the mechanical discipline was hit by a contract termination. CEO Russell Crawford's stated focus was on three things: completing the asset disposals, resolving the contractual claims on Kusile, and resolving the legacy arbitration matters (Engineering News, 28 November 2023). The specific expectation given was that a binding Dispute Adjudication Board decision on Kusile would come during the first half of 2024.

That specific promise was missed, repeatedly. At FY2024 in May 2024, the DAB decision was "anticipated in Q3." At FY2025 in May 2025, the DAB was "expected to issue a binding decision by mid-2025." Along the way, press coverage tracked the slippage: the decision was expected in Q3, then only in February. The decision finally landed on 17 October 2025, roughly eighteen months after the first stated expectation.

It is worth being fair about what this shows. The DAB timetable was never in management's control. It is an adjudication process with two parties and a panel. What management can be criticised for is having repeated a timing expectation five times without adequately signalling that they did not control it. What they cannot be criticised for is the outcome: the award, when it came, was R685 million excluding VAT, materially above what the market had priced, and the November 2025 settlement of R580 million converted it to cash within weeks.

The disposal promise was also chronically late, and also delivered. The Mozambique business had been classified as held-for-sale since August 2020. At FY2024 the transaction timeline was extended to 30 June for condition fulfilment. At FY2025 in May 2025, the disclosure was that the buyer was renegotiating the purchase price and terms, which is the kind of sentence that signals a deal in trouble. At H1 FY2026 in November 2025 the sale to East Africa Enterprises SPV FZCO was progressing with proceeds of approximately US$3.2 million expected by 31 December 2025. It was implemented on 12 December 2025. Five and a half years from classification to completion.

The operating promise, however, was kept every single time. This is the part of the record that deserves the most weight, and it is easy to miss underneath the disposal and adjudication noise. Across all six periods, operating profit before investment income improved year on year in every single one:

• H1 FY2024: improved on H1 FY2023, driven by Inland roughly doubling its operating result.
• FY2024: operating profit before investment income roughly doubled on FY2023, on revenue up 17%.
• H1 FY2025: operating profit before investment income up 113% on H1 FY2024, with Coastal moving from close to breakeven to a genuine contribution.
• FY2025: operating profit before investment income up 59% on FY2024, with every one of the four segments profitable.
• H1 FY2026: operating profit up 22% on H1 FY2025.
• FY2026: operating profit before investment income more than doubled on FY2025, though the underlying figure excluding the Kusile settlement and a R27 million impairment was approximately flat at R328 million against R333 million.

Six consecutive periods of improvement in the headline operating result, achieved while the company was insolvent on a current-asset basis, under a lender-imposed restructuring plan, litigating against its largest historical customer, and trying to sell a business it had been trying to sell since 2020. That is a genuine operational achievement and it is the strongest single piece of evidence on management's side of the ledger.

The debt promise was made quietly and then over-delivered. Through FY2024 and FY2025 the pattern was a series of six-month lender extensions: repayment extended to 30 June 2025, then to 30 June 2026, with the loan reducing from R997 million to R984 million to R947 million to R849 million. That is a company paying down debt at the pace its cash flow allowed while asking for more time each cycle. It is honest but it is not impressive.

Then it changed shape entirely. The new R850 million Standard Bank facility was signed in October 2025 with a first repayment due February 2026 and a final one in August 2030, which on the face of it was a five-year runway. By the 16 February 2026 operational update the facility was already down to R250 million, and by the FY2026 results the outstanding capital was R223 million, with the anticipated annual interest charge cut by approximately 70%. Management did not guide to that. They took a five-year facility and retired three quarters of it in four months.

The one promise that was quietly dropped, and it worked in shareholders' favour. The Restructuring Plan explicitly contemplated "an evaluation of the capital structure, including the potential of raising new equity" (FY2024 results commentary, May 2024). It appeared as an outstanding item in release after release. It never happened. Existing shareholders were never diluted through what was, for four years, an obvious candidate for a heavily discounted rights issue at a share price that spent much of the period below 50 cents. Management chose asset sales, claim resolution and operating cash generation over the easier route of issuing paper at the bottom. Whether by design or by inability to underwrite an issue, the outcome for shareholders was the better one and it is fair to credit it.

Promise versus outcome, side by side:

The assessment.

This is management that does what it says, but is consistently and materially over-optimistic on timing of matters it does not control, and under-promises on matters it does.

The distinction is sharp and consistent. Every single missed commitment in the six-period record is a date on a process governed by a third party: a dispute adjudication board, a foreign court, a buyer's financing, a lender's credit committee. Every single delivered commitment is something inside the company's own control: operating profit, cost, disposals of its own property and plant, debt repayment, and the decision not to dilute shareholders.

There is no instance in the record of management claiming an operational outcome that did not materialise, and no instance of a target being quietly restated downward and then declared met. When the DAB slipped, they said it slipped. When the Mozambique buyer started renegotiating, they disclosed it rather than burying it. When the balance sheet was in a hole they published the number: current liabilities exceeding current assets by R1,241 million, in the same release in which they announced the Kusile settlement that fixed it. A management team inclined to spin would have led with the settlement and buried the deficit.

The honest criticism is that they let a third-party timeline become a de facto guidance item five times in a row without adequately caveating it, and that a shareholder reading the FY2024 release would reasonably have expected the Kusile cash roughly eighteen months before it arrived. In a business where the entire equity value hung on that cash, that is not a trivial communication failure.

Verdict: credible on delivery, unreliable on dates. Weight what they say about operations and costs. Discount by a year, at least, anything they say about a process a court, a board or a counterparty controls.

10. Shareholder friendliness index

Dividends. Stefanutti Stocks paid no dividend in FY2024 (year ended 29 February 2024), no dividend in FY2025 (year ended 28 February 2025), and no dividend in FY2026 (year ended 28 February 2026). The FY2025 results state plainly that no dividend was declared, and the FY2026 results confirm that the board again did not declare one (Business Day, 26 May 2026). Third-party data confirms a trailing and forward dividend yield of 0.00%. The reason is not ambiguous and does not reflect on management's willingness to return capital: for the whole of this three-year window the company was operating under a lender-imposed Restructuring Plan, was insolvent on a current-asset basis, and carried a material uncertainty over going concern. Distributing cash to shareholders in that position would have been indefensible. The FY2026 cash windfall from the Kusile settlement went to prepaying R620 million of bank debt rather than to a distribution, which was the correct sequencing. No dividend policy has been announced for FY2027.

Buybacks and dilution. The MoatMap disclosure database records zero buybacks in the trailing ~90 days (window from 11 June 2026), and a search of the three-year record of JSE announcements, results commentary and financial press coverage found no share repurchase programme announced or executed in FY2024, FY2025 or FY2026 either. Every disposal, settlement and operating cash flow in that period was directed at debt reduction, which took the loan from R997 million at the start of FY2025 to R223 million at February 2026. The dilution picture is the more important half of this paragraph, and it is unusually good. The Restructuring Plan explicitly contemplated raising new equity, and it was never done: there was no rights issue and no placement at any point across the three years, despite a share price that spent much of the period below R1. Shares outstanding stood at approximately 167.2 million as at 26 May 2026, and the weighted-average share count implied by reported earnings and earnings per share across FY2024, FY2025 and FY2026 has been broadly stable at roughly that level, meaning the count is essentially flat over three years rather than shrinking or inflating. The one share-count event of note is the settlement on 31 August 2026 of the Forfeitable Share Plan awards granted in August 2023: 9,840,000 of 10,050,000 awards vested, and after net settlement for participants' tax liabilities, 5,412,000 treasury shares were transferred to participants rather than new shares being issued (SENS, 1 September 2026). Using existing treasury stock rather than issuing new equity limits the dilution to the release of already-created shares.

Verdict: Neutral. No dividends and no buybacks over three years, but that is the correct answer for a company that spent the entire period repairing a balance sheet, and management earned real credit by retiring debt and refusing to dilute shareholders through a rescue equity raise when the share price was at its lowest.

11. Insider activities

A note on sourcing. South Africa is an open venue: JSE SENS announcements are publicly accessible and are the primary regulatory source for director dealings under paragraphs 6.77 to 6.89 of the JSE Listings Requirements. The MoatMap database block supplied with this brief captured three transactions over the trailing twelve months. The SENS record shows materially more than that, and the MoatMap block therefore undercounts this venue. Every transaction below has been verified against the underlying SENS announcement, and the citations are to those announcements rather than to the database.

One SENS announcement titled "Dealings in Securities by a Director of a Major Subsidiary of the Company," dated 8 January 2026 at 10:41, appears in the announcement index but its detail could not be retrieved within the research budget. It is flagged here rather than omitted.

Recent transactions

Sources: JSE SENS announcements dated 1 September 2026, 28 August 2026 (Donato), 28 August 2026 (Wisse), 21 August 2026, 18 August 2026, 14 August 2026 and 15 December 2025.

Buys: reading the signal

Every single open-market transaction in the twelve-month window was a purchase. There were no sales at all.

The August 2026 activity is the part that matters, and it should be read carefully because there are two distinct things happening in the same month.

The first is the Forfeitable Share Plan purchases on 19 August 2026. These are structural rather than discretionary: participants bought shares under a plan approved by shareholders in April 2023, awarded on the basis of the year ended February 2026 results, vesting on 30 June 2029. They are compensation mechanics, not conviction, and the individual amounts (R48,630 for the CEO, R30,384 for the CFO, under R5,000 each for the subsidiary directors) are too small to signal anything. Their only real content is the three-year vesting date, which locks six senior people to the outcome through mid-2029.

The second is the genuinely discretionary buying, and it is the signal. Three separate senior operating executives made unforced open-market purchases within a fifteen-day window in August 2026, with their own money, at prices between R6.00 and R6.30. Mauro Donato bought 50,000 shares for R300,000 on 28 August. Shaun White bought 35,000 for R220,500 on 17 August. Eric Wisse bought twice, 6,500 shares on 13 August and 20,000 shares on 27 August, for R160,625 combined. That is roughly R681,000 of personal capital from three people in a fortnight.

This is cluster buying by operating insiders and it is a very bullish signal. Three points sharpen it further.

First, these are not non-executive directors buying a token holding for optics. Wisse, White and Donato are directors of Stefanutti Stocks Proprietary Limited, the main operating subsidiary. They are the people who know what is actually in the R17.2 billion order book, what it is priced at, and whether the sites are properly staffed. They are precisely the population whose buying carries information.

Second, the timing. The FY2026 results were published on 26 May 2026 and made explicit that underlying operating profit excluding the Kusile settlement was approximately flat. Anyone buying in August 2026 had read that disclosure, had seen the market's reaction, and bought anyway. They are not buying the settlement; the settlement is already in the price and already spent on debt. They are buying the order book.

Third, they are buying into strength, not weakness. The purchases were made at R6.00 to R6.30 after a share price that had risen roughly 555% over three years. Insiders adding at multi-year highs, with their own cash, is a materially different signal from insiders averaging down.

The pattern is also not new, which strengthens it. Outside the twelve-month window, Shaun White bought 16,593 shares at R3.75 in January 2025; Bharti Harie, a non-executive independent director, bought 36,300 shares across two trades at R4.55 and R4.56 in May 2025; and Eric Wisse bought 30,000 shares at R4.19 and 26,000 at R3.85 in late May 2025 (Moneyweb SENS). Shaun White has now bought at R3.75, at R4.50 and at R6.30. Eric Wisse has bought at R3.85, R4.19, R6.25 and R6.00. These are people who have added at every price point over eighteen months as the thesis played out, rather than people who bought once and stopped.

Sells: working out the why

There are none to explain. Across the full twelve-month window there is not a single disposal of shares by any director of the company or of its major subsidiary.

The one transaction that could be mistaken for a disposal is the 31 August 2026 Forfeitable Share Plan settlement, and it is worth being precise about what happened. 9,840,000 awards vested, but only 5,412,000 shares were actually transferred. The gap is not a sale. The plan operates on a net settlement basis: the number of shares delivered is reduced by an amount equivalent to each participant's tax liability on vesting, and the withheld shares are retained rather than sold into the market. Crawford, for example, had 1,155,000 awards vest and received 635,250 shares. No shares reached the market and no insider realised cash. This is the routine tax mechanics that this section's guidance says to skip, and it is included only because its scale (5.4 million shares) would otherwise look like a large insider transaction to anyone scanning the announcement index.

Net assessment

Insiders are unambiguously net buyers, and the buying is broad-based rather than concentrated in one person. Four distinct individuals made open-market purchases in the twelve-month window (White in December 2025, then Wisse, White and Donato in August 2026), plus the CEO and CFO through the plan-based purchases in August. Total insider disposals: zero shares, zero rand.

What has changed recently is the intensity and the price level. For eighteen months, insiders bought steadily in the R3.75 to R4.56 range while the company was still finishing its restructuring. Then, within three months of publishing a full-year result that explicitly told the market its underlying operating profit was flat, three operating directors bought at R6.00 to R6.30. The most reasonable reading is that the people running the sites believe the roughly doubled order book converts into earnings that the flat FY2026 underlying number does not reflect.

The one caveat worth stating: the buying is concentrated among subsidiary directors rather than the plc board. The CEO and CFO have added through plan-based purchases, but neither has made a large discretionary open-market purchase in this window, and no non-executive director has bought since May 2025. That is a mild qualification, not an offset.

Plain-language read: bullish signal. Zero sales, four buyers, cluster purchases within a fifteen-day window by the executives closest to the order book, made at multi-year high prices with personal capital, and three-year lock-in through the 2029 vesting date. This is close to the cleanest insider picture a company can present.

12. Scenarios

Bull case

The order book is the whole story, and in the bull case it converts cleanly.

Over the next two to three years, the R17.2 billion backlog runs through the business the way it was priced. The Inland region, which has improved its operating result in every one of the last six reporting periods, absorbs its enlarged book without straining, because it is doing more of what it already does on sites it already knows. Coastal, having deliberately traded volume for margin through FY2026, converts its enlarged book at the better margins that discipline bought. The Western Cape's renewables balance-of-plant capability finds itself in exactly the right place at the right time: South Africa cannot connect its new generation without substations and transmission, the energy allocation in the national infrastructure programme is real money, and there are very few contractors in the country that can do high-voltage substations, 132 kV overhead line and heavy civils under one contract. The pre-cast concrete tower capability turns from an interesting reference project into a repeatable product line as wind build-out accelerates.

Africa is the swing factor and in this case it swings the right way. The kwacha stops appreciating, the currency drag reverses, and a segment whose forward book grew roughly fivefold in a single year starts contributing rather than subtracting. The Botswana, Zambia and Eswatini entities, which are genuine local incumbents rather than expatriate project offices, take the work that the group's restored balance sheet now lets them bid for. The R6 billion of offshore order book turns out to be higher margin than the domestic work, which is the stated reason the group exited Mozambique and Mauritius in the first place.

Underneath all of it, the balance sheet keeps working in the shareholders' favour without anyone winning a single new contract. Debt that was R850 million in October 2025 was R223 million four months later, and the interest charge falls by roughly 70%. Every rand of that saving is earnings that require no execution. If the company stays disciplined and keeps paying down, it reaches a genuinely net cash position, which changes what it can bid for again, because guarantee capacity is the gate on contract size.

The industry backdrop cooperates. The consolidation is done. Murray & Roberts is delisted, Group Five and Basil Read are gone, Aveng is much reduced. A R1.07 trillion public infrastructure programme starts converting into awarded work rather than budget lines. The R12.5 billion of short-term potential awards lands, and a meaningful slice of the R144 billion identified prospects follows. With fewer bidders and more work, pricing improves for everyone left standing, and a contractor with CIDB Grade 9, B-BBEE Level 1 and in-house multi-discipline capability is on a very short list.

The tell that this is happening will not be a press release. It will be the H1 FY2027 result showing operating profit up meaningfully on an underlying basis, with no settlement in it, and the Africa segment back in the black.

Base case

The most likely path is that the order book converts, but more slowly and less cleanly than the headline number implies, and that this is fine.

The company delivers roughly what its six-period record suggests it delivers: steady, unspectacular improvement in the operating result, period after period, without dramatic surprises in either direction. Revenue grows as the backlog is worked through, but recognition lags the order book because South African projects start late, especially public ones, and because the group is staffing up into a market where experienced site agents are genuinely scarce after a decade of industry contraction. Some contracts in the new book underperform their tender margin, because on a book that doubled in twelve months some always do. Others outperform. The net is a business that grows and improves, but where the operating margin sits closer to where it has been than where a bull would put it.

Africa stays awkward for a while. The currency mismatch on dollar-denominated Zambian contracts does not resolve just because management would like it to, and the segment probably takes another period or two to work through the contracts priced before the kwacha moved. It stops being a loss-maker rather than becoming a star.

The debt keeps coming down, the interest charge stays low, and at some point in the next two years the board faces the question it has not faced since before 2019: what to do with genuine surplus cash. Given a management team that spent four years being told what to do with cash by a lender, the base case is that they stay conservative for longer than shareholders would like, keep paying down debt, and only then start thinking about a distribution.

Nothing in this path requires the infrastructure programme to be delivered in full or the energy transition to accelerate. It requires the company to be an averagely competent contractor in a market with fewer competitors than it had ten years ago, with a balance sheet that no longer excludes it from bidding. That is the reasonable expectation, and it is roughly what management has been quietly delivering every six months since November 2023.

Bear case

The bear case is not that demand fails. It is that the company wins the work and cannot execute it.

An order book that roughly doubles in twelve months is a management problem before it is a revenue opportunity. Every one of those contracts needs a project manager who has run something similar, a quantity surveyor who catches the variations before the deadline for claiming them passes, and a site agent who can hold the programme. South Africa lost a decade of those people to other industries and other countries when the sector collapsed, and they do not come back on twelve months' notice. In the bear case, Stefanutti Stocks staffs the new book with people who are one level less experienced than the job needs, and the consequences show up eighteen months later in the form of contracts that were priced correctly and executed badly. Because these are fixed-price contracts, there is no mechanism to recover it. One or two large jobs going wrong is enough to erase the profit from several good ones, which is precisely the pattern that took WBHO out of Australia and that has ended most South African contractors.

Compounding it, the FY2026 result already showed the underlying business standing still: strip out the Kusile settlement and the impairment and operating profit was approximately flat year on year. If FY2027 arrives with no settlement, a larger cost base built for the new book, and a handful of underperforming contracts, the reported numbers go backwards even as revenue grows. That is the worst possible combination for a company whose equity story is a turnaround.

The Africa segment is the second front. Its order book grew roughly fivefold while it was posting an operating loss caused by a currency mismatch it cannot fully hedge. If the kwacha stays strong, the segment simply runs more volume through the same broken economics, and a loss that was manageable at one scale becomes material at five times the scale.

Then there is the counterparty risk that has not gone away. 22% of revenue is public sector, and there remains an unresolved dispute with Eskom over Kusile Package 28, the contract awarded to the Stefanutti Stocks-Izazi joint venture and terminated by Eskom in February 2019. The November 2025 settlement was a full and final release on a different contract. The mechanism that nearly destroyed the company between 2018 and 2025 was not fraud or incompetence; it was a state client disputing what it owed on completed work while the contractor funded the gap for seven years. Nothing structural prevents a repeat, and a company that has only just repaired its balance sheet has far less capacity to absorb a second one.

Finally, the R1.07 trillion infrastructure programme is an allocation, not a pipeline. South African public procurement has a long record of slippage between budget, tender, award and site handover. A contractor that builds overhead for a programme that arrives two years late carries the cost in the meantime. In the bear case, the private sector work that is 78% of revenue softens with the economy at the same time as the public work fails to arrive on schedule, and the company finds itself with a large fixed cost base, an order book that is not converting, and a bank facility that has to be serviced regardless.