HE Group Berhad (0296.KL) - Deep Dive Research Report

Sector: Industrials / Capital Goods (Electrical Engineering Services) Listing: ACE Market, Bursa Malaysia (listed 30 January 2024) Principal operating subsidiary: Hexatech Engineering Sdn Bhd Headquarters: Puchong, Selangor, Malaysia Fiscal year-end: 31 December Report date: 6 June 2026

A note on sourcing and the "concall" requirement. HE Group is a recently listed ACE Market micro-cap. Like almost all Bursa companies of its size, it does not host quarterly earnings conference calls or publish transcripts. There are no four concall transcripts to retrieve, and pretending otherwise would be fabrication. In their place I have used the four most recent statutory quarterly result announcements filed on Bursa Malaysia, plus the associated press coverage and the one initiating analyst note (Phillip Research) that exists on the stock. The four reporting periods used throughout are: Q1 FY2026 (ended 31 Mar 2026, filed 18 May 2026), Q4 FY2025 (ended 31 Dec 2025, filed ~Feb 2026), Q3 FY2025 (ended 30 Sep 2025, filed ~Nov 2025), and Q2 FY2025 (ended 30 Jun 2025, filed ~Aug 2025). Where a specific quarter's detail could not be verified from a primary or reputable secondary source, I say so rather than invent it.

1. What the company does

HE Group is the holding company for Hexatech Engineering, a Malaysian electrical engineering services contractor. In plain terms: when a semiconductor fab, a medical-device plant, or now a data centre is being built or expanded in Malaysia, somebody has to design and install the entire system that takes high-voltage grid power coming into the site, steps it down through substations, and distributes it safely to every machine, clean room, and rack inside the building. HE Group is the specialist that does that work. They design the power distribution system, supply the switchgear and equipment, install it, then test and commission it so the customer can switch the building on.

The company handles the full voltage ladder - high voltage (HV), medium voltage (MV), low voltage (LV), and extra-low voltage (ELV). A typical scope runs from the incoming MV intake and the indoor/outdoor substations, through transformers and switchboards, down to the cable trays, busbars, and final distribution boards feeding production equipment. On top of the core power work they also do "building systems" (mechanical and instrumentation systems, plus civil, structural, and architectural works on the electrical rooms), the hook-up and retrofitting of electrical equipment for plants that are reconfiguring lines, and a small amount of trading in electrical products.

The business was incorporated in 1995 (Hexatech Engineering, registration 199501008607), so it is a roughly 30-year-old engineering firm that spent most of its life as a private contractor before its founders took it public on the ACE Market in January 2024 to raise growth capital. The IPO raised RM24.33 million at 28 sen per share; the stock debuted at 43 sen, a 53% first-day premium, on the back of a 63x retail oversubscription - a sign of how strongly the market wanted exposure to Malaysia's factory-building cycle.

The core value proposition is reliability and qualification in a setting where electrical failure is catastrophic. A semiconductor fab cannot tolerate an unplanned power trip; a single excursion can ruin millions of ringgit of wafers in process and take days to recover clean-room conditions. So the multinationals building these plants do not award the electrical package to the cheapest bidder - they award it to a contractor who has already been vetted, has a track record on similar critical facilities, and can be trusted to deliver to schedule during a fast-moving construction program. HE Group's whole position rests on being inside that approved-contractor circle for blue-chip industrial clients.

What makes the work hard to replicate is not a single patented technology but the accumulation of qualified project references, licensed engineers, and the contractor accreditations (CIDB grade, Suruhanjaya Tenaga / Energy Commission electrical contractor registration, ISO quality systems) needed even to bid. A new entrant can buy switchgear, but it cannot instantly manufacture a portfolio of completed fab and medical-plant substations or the relationships with the EPC and EPCM managers who hand out the electrical subcontracts.

A concrete example of the company in action is the data-centre job it won in 2025: an EPCM (engineering, procurement and construction management) contractor building a data centre in Cyberjaya needed a partner to execute the complete electrical services package. HE Group's Hexatech signed an eight-month subcontract worth roughly RM56.6 million to design, supply, install, test, and commission the electrical distribution for that facility. That single contract is larger than a typical year of the company's historical order intake and is the reference project that opens the door to the rest of Malaysia's data-centre wave.

2. Business segments

HE Group reports as a single operating business (electrical engineering services through Hexatech), but within that business the revenue splits across four service lines. They are not separate reporting segments with their own P&Ls in the way a conglomerate's would be; they are service categories within one contracting operation. The split below is from the Q1 FY2026 disclosure, where power distribution dominated at roughly 90% of revenue.

2.1 Power distribution systems (~90% of revenue)

This is the company, effectively. It covers the design, supply, installation, testing, and commissioning of HV/MV/LV/ELV power distribution systems and substations for industrial plants, commercial buildings, and (increasingly) data centres. The core capability is the engineering and project-management knowledge to take a customer's electrical load requirement and turn it into a working, commissioned, code-compliant distribution system delivered to a construction deadline. What took years to build is the qualified-reference portfolio with multinational semiconductor and medical-device clients and the licensed-contractor accreditations that gate the work. This is simultaneously the cash engine and the growth bet - everything else is ancillary. Within this segment the competitors are other Malaysian electrical sub-contractors of similar grade plus the local arms of multinationals; HE Group wins on track record, qualification, and the ability to mobilise quickly on critical-facility schedules.

2.2 Building systems and works

Adjacent scope that often travels with the electrical package: mechanical systems, instrumentation systems, and the civil, structural, and architectural works associated with substations and electrical rooms. It exists because customers prefer a contractor who can take responsibility for the full electrical-room build rather than coordinating three vendors. It is a smaller share of revenue and lower-differentiation than the core distribution work, but it raises the value of each project and deepens the customer relationship.

2.3 Electrical equipment hook-up and retrofitting

Work for existing plants rather than greenfield builds: connecting up new production equipment to existing distribution, and retrofitting or upgrading electrical infrastructure when a fab or plant reconfigures or expands a line. This is more recurring and less lumpy than greenfield substation work, because operating plants continually add and move equipment. It ties HE Group to the installed base of customers it has already served.

2.4 Trading of electrical products

A small trading line in electrical products. It is incidental to the contracting business - useful for sourcing and margin on pass-through equipment, not a strategic pillar.

3. Products and business detail

HE Group's "product" is a delivered, commissioned electrical system, so the catalogue is best understood as a scope ladder rather than a shelf of SKUs.

The full scope ladder. At the top sits the HV/MV intake and substation work: incoming medium-voltage supply, indoor and outdoor substations, transformers, ring main units, and MV switchgear. Below that, the LV distribution: main switchboards, sub-distribution boards, busbar trunking, power and control cabling, cable trays and containment. At the building edge, ELV systems and the final connections to production equipment. Surrounding the electrical core are the building-systems works (mechanical, instrumentation, and the civil/structural/architectural build of the electrical rooms), the hook-up and retrofit service for live plants, and the trading of electrical components.

What makes it hard to do. Three things gate the work. First, accreditation: an electrical contractor in Malaysia needs Energy Commission (Suruhanjaya Tenaga) registration, the appropriate CIDB contractor grade, and competent/chargeman-licensed personnel to legally execute and energise HV/MV installations; quality and safety certifications (ISO systems) are table stakes for multinational clients. Second, qualified references: semiconductor and medical-device multinationals run vendor-qualification processes - you must have done comparable critical-facility work, demonstrably and safely, before you are allowed to bid the next one. Third, schedule and coordination capability: fab and data-centre builds run on compressed timelines with many trades on site, and the electrical contractor who can mobilise crews and commission on time without compromising safety is the one who gets re-invited.

Delivery and geography. HE Group is a Malaysia-focused contractor headquartered in Puchong, Selangor. As part of its IPO growth plan it set out to establish physical offices in Kedah (to be near the northern semiconductor cluster around Kulim/Penang) and Johor (to be near the Johor-Singapore data-centre and industrial corridor). These two locations are exactly where Malaysia's two biggest electrical-demand engines sit - the established E&E/semiconductor belt in the north and the explosive data-centre build-out in the south - so the regional-office strategy maps directly onto where the contracts are.

Milestones that changed the business.

• 1995: Hexatech Engineering incorporated; decades as a private electrical contractor to multinationals.
• January 2024: ACE Market IPO at 28 sen, RM24.33m raised; order book RM211.9m and tender book ~RM400m at listing.
• August 2025: Letter of instruction for its first major data-centre electrical job in Cyberjaya (~RM56.7m), formalised as an eight-month subcontract (~RM56.6m) in November 2025 - the company's entry into data-centre infrastructure and its single largest contract.
• Q1 FY2026: incorporation of Digitalflux Technologies Sdn Bhd, a new subsidiary to expand into IT and engineering services; declaration of a maiden interim dividend of 1 sen per share.

4. Customers

Who buys. HE Group's customers are the multinational manufacturers building and expanding plants in Malaysia, and the EPC/EPCM main contractors who manage those builds on the owners' behalf. At IPO the end-market mix of revenue was roughly 77% semiconductor, 16% medical devices, and 7% other electronics / end-user industries - a customer base of blue-chip foreign manufacturers operating in Malaysia's E&E heartland. The newest customer type is the data-centre operator and its EPCM contractor: the Cyberjaya job came not directly from the owner but from a foreign-owned EPCM firm specialising in construction, civil engineering and offshore oil-and-gas, which subcontracted the electrical package to Hexatech.

Who makes the decision and how. On a greenfield fab or plant, the electrical package is typically awarded by the project's main contractor or EPCM manager, with the end-owner's facilities and construction-management team setting the qualification bar. The buying criteria are, in order: prior qualification on comparable critical facilities, safety record, ability to hold schedule, and price - in that priority, because the cost of electrical failure or schedule slip on a fab dwarfs the cost of the electrical package itself. Sales cycles are tied to the customer's capex calendar: a contract follows months of pre-qualification and tendering, which is why the company tracks a tender book (bids submitted, awaiting award) separately from its order book (won, awaiting execution).

Why they choose HE Group. Existing qualification is the decisive reason. A multinational that has already approved Hexatech on previous projects can award the next package with far less requalification friction than bringing in a new contractor. The bundle (power distribution plus building systems plus hook-up) is the second reason - one accountable party for the electrical room.

Switching costs and lock-in. The lock-in is on the buy side: once a contractor is qualified and has performed, the owner's incentive is to keep using it, because re-qualifying a new vendor consumes time the construction schedule does not have, and carries execution risk. The hook-up and retrofit line deepens this - having built a plant's distribution, HE Group is the natural party to extend or modify it. There is no contractual recurring revenue, but there is a qualification moat around the installed base.

Concentration and contract structure. This is a project-based contractor, so revenue is inherently lumpy and concentrated by contract. A single award like the ~RM56.6m data-centre subcontract can dominate a year's billings, and revenue recognition is by progress billing over a project's life, which is why quarterly revenue swings with project timing (Q1 FY2026 revenue fell year-on-year specifically because projects completed and progress billings stepped down). The order book - around RM170 million as of May 2026, down from RM211.9m at IPO - is the key forward-visibility metric, and order-book replenishment (winning the next contracts) is the central operating question for the business. Contracts are milestone/progress-based, fixed-scope subcontracts of months to a couple of years in duration; there is no long-term recurring supply revenue.

5. Competitive landscape

The Malaysian electrical-engineering-services market is fragmented and competitive, with no single dominant specialist contractor at HE Group's tier. The structure has three layers, and HE Group competes mainly within one of them.

At the top sit the global multinationals - Siemens, ABB, Schneider Electric - who supply much of the actual switchgear, transformers, and distribution equipment and offer their own engineering. They are equipment principals more than site-installation subcontractors at HE Group's scale; HE Group is often installing their products rather than competing head-on for the same scope, though the multinationals' local engineering arms can compete on larger or more sophisticated packages.

At the utility level sits Tenaga Nasional (TNB), which owns generation, transmission, and grid distribution - it is the counterparty on the grid side, not a competitor for in-plant electrical installation.

HE Group's real competition is the middle tier: other Malaysian electrical and M&E sub-contractors of comparable grade who hold the same accreditations and bid the same fab, plant, and data-centre packages. This is a crowded field of private and listed contractors (the broader peer set of Malaysian electrical/EPCC engineering names includes firms such as Pekat Group, Solarvest, and EIP/EI Power, though several of those skew toward solar/renewables EPCC rather than HE Group's critical-facility power-distribution niche). Against this tier HE Group wins on its specific qualification track record with semiconductor and medical-device multinationals and its early-mover reference in data-centre work; it loses, or is squeezed, when a job is awarded primarily on price or when a competitor holds the incumbent qualification for that particular owner.

Barriers to entry are real but not impassable. The accreditations (Energy Commission registration, CIDB grade, licensed chargemen) and the qualified-reference requirement keep out casual entrants and protect incumbents on critical-facility work. But there are many qualified mid-tier contractors already inside the circle, so the barrier protects the category of approved contractors more than it protects HE Group specifically. There is no patent, no proprietary technology, and no switching cost that prevents an owner from qualifying a second vendor over time. This is a track-record-and-relationships business, not a moat business - margins are decent on well-executed critical work but the company is a price-taker on commoditised packages and must continually win the next tender.

Structural shift in its favour. The one thing genuinely changing the competitive picture is demand, not supply: Malaysia's simultaneous semiconductor expansion and data-centre boom is enlarging the pie faster than new qualified contractors can be minted, which tightens capacity among the approved mid-tier and gives incumbents like HE Group more pricing and selection power than they would have in a flat market.

6. Industry

What drives demand. HE Group sells into industrial and digital-infrastructure construction in Malaysia, so its demand is driven by capex decisions of two engines: (1) the electrical and electronics (E&E) / semiconductor manufacturing build-out, where Malaysia is a long-established node in global chip assembly, test, and increasingly front-end and advanced-packaging investment; and (2) the data-centre boom, the fastest-growing source of new electrical-installation demand in the country.

Size and trajectory. The data-centre wave is the headline. Malaysia is attracting on the order of RM185 billion of data-centre investment across roughly four years, and total digital investment commitments reached over RM210 billion (~USD 51 billion) across 2023-2024. Year-to-date 2026 data-centre contract awards in Malaysia have already reached roughly RM7.4 billion. Installed data-centre capacity is projected to roughly double from about 1.26 GW in 2025 to ~2.5 GW by 2030 (mid-teens annual growth), with one market estimate putting the Malaysian data-centre market at USD 4.0bn in 2024 growing to ~USD 13.6bn by 2030. Johor (anchored by proximity to Singapore and the Johor-Singapore Special Economic Zone launched January 2025) is expected to host the majority of national capacity by 2030; Cyberjaya - where HE Group won its first DC job - remains a major established hub with dozens of facilities operating and under construction. Global hyperscalers (Microsoft, Google, AWS) and partnerships like YTL Power-Nvidia anchor the pipeline.

Where HE Group sits in the chain. It is a downstream installation specialist. The value chain runs: power utility (TNB) → grid connection → site substations and in-building distribution (HE Group's scope) → the customer's production or IT equipment. HE Group does not build the data centre or the fab; it electrifies it. That position means it benefits from the construction boom without taking technology risk on the chips or the servers themselves - but it is exposed to construction timing and to the grid/power-availability constraints that gate these projects.

The binding constraint - and why it matters to the thesis. The single biggest industry dynamic is power and water supply. Malaysian data-centre electricity demand is projected to rise roughly seven-fold, from ~8.5 TWh in 2024 to ~68 TWh by 2030, and Johor already faces large water-supply shortfalls against data-centre cooling needs. This cuts both ways for HE Group: power-distribution work is exactly what a power-constrained build-out needs more of (bullish for demand), but grid bottlenecks can also delay project awards and commissioning (the timing risk that the analyst note flagged when it trimmed near-term forecasts on "timing delays in new contract awards").

Regulation and cyclicality. The regulatory environment is the contractor-licensing regime (Energy Commission, CIDB) plus the safety and quality standards demanded by multinational clients - a qualification gate rather than a market-access barrier. The business is cyclical and capex-driven: it rides industrial and digital construction cycles, and in a downturn new plant and data-centre awards slow, the order book is not replenished, and revenue falls with a lag. The current cycle is firmly in an up-phase, but it is a cycle, not a secular annuity.

7. Growth triggers

Sourced from the four most recent quarterly result filings and the company's contemporaneous Bursa announcements (HE Group does not hold concalls; quarterly MD&A and contract announcements are the equivalent disclosure). Each item is dated to the filing or announcement it came from.

• Data-centre infrastructure entry, anchored by the Cyberjaya subcontract. The ~RM56.6m eight-month electrical-services subcontract for a Cyberjaya data centre is the company's reference project into the data-centre segment, explicitly framed by management as the gateway to "the fast-growing data centre infrastructure segment driven by AI and cloud demand." (Announced Aug 2025 as a letter of instruction; formalised Nov 2025; reiterated in Q1 FY2026 filing, 18 May 2026.)

• Order book of ~RM170 million providing near-term revenue visibility. Management cited the order book as the basis for near-term earnings visibility. (Q1 FY2026 filing, 18 May 2026.)

• New subsidiary Digitalflux Technologies Sdn Bhd to expand into IT and engineering services. A move to broaden the service base beyond pure electrical installation. (Q1 FY2026 filing, 18 May 2026.)

• Regional office expansion into Kedah and Johor. IPO-funded establishment of physical offices in the northern semiconductor belt (Kedah) and the southern data-centre corridor (Johor), positioning the company at both demand centres. (IPO prospectus / listing, Jan 2024; expansion allocation RM3.65m, restated in Q1 FY2026 IPO-proceeds disclosure.)

• Tender book skewed heavily to data centres. The pipeline of submitted bids was reported (via Phillip Research's initiating coverage) at roughly RM675 million, with ~80% data centres and the remaining ~20% across semiconductors and medical devices - the forward funnel from which future order-book wins are drawn. (Phillip Research initiation, 2025.)

• Anticipated earnings recovery as new project rollouts gather pace. Management's and the analyst's framing is that after a soft patch caused by project completions and award-timing delays, the ramp of new (especially data-centre) projects is expected to drive a meaningful recovery in the FY2026 reporting year. (Phillip Research note, 2025; consistent with Q1 FY2026 commentary, 18 May 2026.)

8. Key risks

Order-book replenishment / award-timing risk (high probability, moderate-to-high impact). This is the central risk for any project contractor. Revenue is recognised as projects are executed, and once a project completes, that revenue disappears unless a new contract has been won to replace it. The order book has already fallen from RM211.9m at IPO to ~RM170m by May 2026, and Q1 FY2026 revenue declined year-on-year precisely because projects completed and progress billings stepped down. The analyst note trimmed FY2025-26 forecasts specifically on "timing delays in new contract awards." If data-centre and fab awards slip - which they can, given the power-grid constraints in Malaysia - the order book thins and earnings follow with a lag. The mitigant is a large tender book (~RM675m), but a tender is not a contract.

Customer and contract concentration / lumpiness (high probability of volatility, moderate impact). A single award like the ~RM56.6m Cyberjaya job can dominate a year's billings, so quarterly results are inherently volatile and a delayed or cancelled large contract can swing a whole year. This is structural to the model, not a temporary condition.

Dependence on a continuing industrial/data-centre capex up-cycle (medium probability, high impact). The entire thesis rests on Malaysia's semiconductor and data-centre construction boom continuing. If hyperscaler capex moderates, if power/water constraints stall the data-centre pipeline (Malaysian DC power demand is projected to rise ~7x by 2030 and Johor already faces a multiple-times water shortfall), or if the semiconductor cycle turns down, the pool of new electrical-installation work shrinks and competition for what remains intensifies. HE Group has no recurring revenue cushion to ride out a downturn.

Thin competitive moat / margin pressure (medium probability, moderate impact). As Section 5 set out, this is a track-record-and-relationships business in a fragmented field, not a structurally protected one. Many qualified mid-tier contractors bid the same packages. On price-led tenders HE Group is a price-taker, and any softening of demand would quickly show up as margin compression. The Q1 FY2026 gross-margin improvement came from project mix and cost control - both of which can reverse.

Execution and key-person risk in a fast-scaling small company (medium probability, moderate impact). Scaling from a ~30-year-old private contractor into data-centre-scale jobs, opening new regional offices, and launching a new IT/engineering subsidiary (Digitalflux) all at once stretches a small organisation. Project execution slippage, cost overruns on fixed-price subcontracts, or stretched management bandwidth could erode the margin and the reputation the business depends on. The founder-managing director, Haw Chee Seng, is also the controlling ~24% shareholder, so the company is closely tied to one individual.

Recently listed, short public track record (inherent). The company has been public only since January 2024, so investors have a limited history of how management performs against guidance across a full cycle, and dividend/buyback policy is still being established.

9. Walk the talk

The four reporting periods used: Q1 FY2026 (filed 18 May 2026), Q4 FY2025 (filed ~Feb 2026), Q3 FY2025 (filed ~Nov 2025), Q2 FY2025 (filed ~Aug 2025). The most recent is 19 days before this report - well inside the 90-day window. The honest caveat repeated up front applies: there are no concall transcripts for a company this size, so this assessment is built from the quarterly MD&A statements, the contract announcements, and the IPO prospectus promises, cross-referenced against outcomes. It is thinner than a transcript-based credibility review would be, and I flag that rather than manufacture quotes.

Start at the IPO (Jan 2024). Management's listing-stage promises were concrete and trackable: use IPO proceeds to expand the business (including a deliberate push into data centres) and to set up offices in Kedah and Johor, supported by an order book of RM211.9m and a tender book of ~RM400m. On the single most important strategic promise - entering data centres - management delivered. They had said at IPO they were "confident in expanding to include data centres due to the growing presence of AI and IoT," and by August 2025 they had converted that into a real, sizeable Cyberjaya data-centre subcontract (~RM56.6m), formalised in November 2025. A promised strategic direction became a concrete contract within about eighteen months. That is a kept promise, and a meaningful one.

Through FY2025. Across the FY2025 quarters the company continued to execute its existing fab/plant order book while the data-centre pipeline matured. The Q3 FY2025 results were reported as a strong quarter with revenue up; the visible weak spot was Q4 FY2025, where profit before tax (RM2.03m) was the softest quarter in the recent run - consistent with the "project completion / award-timing" air-pocket that the analyst note had warned about. So management's narrative (a soft patch on timing, then recovery) lines up with what the numbers actually did, rather than contradicting it.

Into Q1 FY2026 (18 May 2026). Here the "walk" matches the "talk" reasonably well. Revenue was down year-on-year, and management attributed it plainly to project completions and lower progress billings rather than dressing it up - a candid, accurate explanation. At the same time profit before tax recovered sharply from the weak Q4 (PBT up ~67% quarter-on-quarter) and gross margin improved on better project mix, which is the early shape of the recovery they had pointed to. They also did two things that back words with action: declared a maiden interim dividend of 1 sen (returning cash rather than just promising to) and incorporated Digitalflux to broaden the business.

Assessment. On the evidence available, this is management that has so far done what it said on the things that matter most: the data-centre pivot promised at IPO became a real contract, the order book and tender book disclosures have been consistent with subsequent outcomes, and the soft patch was explained candidly rather than obscured. The important calibrations are that (a) the public track record is short (listed only since January 2024), so this is an early read, not a through-cycle verdict; and (b) the absence of guidance/concalls means there are fewer hard, datable promises to hold them to than for a larger company. There is no evidence of overpromising in the record so far. The strongest single corroborating signal that management believes its own story is in Section 11: the managing director has been buying his own stock aggressively in the open market.

10. Shareholder friendliness index

Dividends. HE Group declared its maiden interim dividend of 1 sen per share (RM4.4 million total) in respect of FY2026, paid April 2026 - the first dividend since its January 2024 listing. Because the company has been public for barely two years, there is no three-year DPS trend to assess: FY2024 and FY2025 dividends were either nil or not separately verifiable from the sources reviewed, and the FY2026 interim is best read as a post-IPO initiation of returns rather than part of an established progression. Paying a maiden dividend while still in a growth-capex phase, and while sitting on a large net-cash balance (cash and equivalents of ~RM66.5m against minimal borrowings as of Q1 FY2026), is a shareholder-friendly signal from a newly listed small-cap that could just as easily have retained everything.

Buybacks and dilution. The MoatMap database records zero share-buyback filings for HE Group across the trailing three years - no buyback program was announced or executed, which is unsurprising for a company that only listed in 2024 and is funding growth. On dilution: the share count was enlarged to roughly 440 million shares at the IPO (via the public issue of new shares), and there is no evidence of further share issuance, option-driven dilution, or buyback-driven reduction since - the count looks flat post-listing. The 1-sen dividend on ~440m shares ties to the ~RM4.4m total, which is internally consistent with a stable share base.

Verdict: Neutral, tilting friendly - a newly listed company that has just initiated a cash dividend and is not diluting shareholders, but with too short a history (and an active growth-capex agenda) to be classed as a committed capital-returner yet.

11. Insider activities

Source and staleness disclaimer. All insider data in this section comes from the MoatMap database block injected with this assignment, which scrapes Bursa Malaysia Section 219 director/substantial-shareholder disclosures via klsescreener (the official bursamalaysia.com portal is Cloudflare-gated and not directly retrievable). The last MoatMap scrape was 2026-05-29 12:01 UTC, roughly 185 hours (8 days) stale as of this report - any filings in the most recent week may be missing. No SGXNet/Bursa/aggregator searches were performed, per the assignment's instruction that MoatMap is canonical for this ticker.

The picture over the last 12 months is overwhelmingly one-sided insider buying by the founder-managing director, against a single small director disposal.

Recent transactions (most recent first):

(All rows: Bursa Changes in Director's Interest / Substantial Shareholder's Interest, Section 219 CA 2016, dated as shown. The same transaction appears under multiple interest capacities in the raw feed - director, managing director, and substantial shareholder - because Haw holds all three roles; I have de-duplicated to the underlying lots.)

Buys - reading the signal. Across May 2026, managing director and ~24% controlling shareholder Haw Chee Seng made a sustained run of open-market purchases - at least eight distinct lots over a two-week span (12, 18, 19, 20, and 26 May), totalling on the order of 300,000+ shares for roughly RM120,000-150,000 of his own money, lifting his direct stake incrementally from ~23.95% toward ~24.00%. This is not a one-off token purchase; it is repeated, clustered, open-market accumulation by the founder-CEO who already controls the company. An insider who already owns a quarter of the business adding more, week after week, in the open market, is among the strongest conviction signals available - he is buying because he believes the stock is worth more than where it trades, at prices roughly RM0.37-0.46 per share. This is a very bullish signal. The buying is concentrated in one person, but that person is the single most informed individual about the order book, the tender pipeline, and the data-centre ramp.

Sells - working out the why. The only disposal in the window is Datuk Christopher Wan Soo Kee, a non-executive director, selling 100,000 shares for ~RM45,500 on 26 May 2026, which took his already-tiny holding down to ~0.01-0.02%. No reason is disclosed in the filing. Given the trivial size (a non-executive director trimming a negligible personal position to near-zero), this reads as routine personal portfolio housekeeping, not an informational signal about the business - it is too small, and from a non-executive with no operational visibility comparable to the MD's, to weigh against the founder's accumulation. Reason not formally disclosed; inferred as personal/immaterial from context.

Net assessment. Insiders are decisively net buyers, and the buying is the high-signal kind: the founder-managing director repeatedly adding to an already-controlling stake in the open market, clustered over a two-week period, while the only sale is an immaterial trim by a non-executive director. Activity is concentrated in one person (Haw), which slightly tempers the "broad-based cluster" reading, but the identity of that person - the controlling CEO with the best view of the pipeline - more than compensates. Read: bullish signal. (Caveat: data is ~8 days stale; any post-29-May activity is not captured.)

12. Scenarios

Bull case. Malaysia's data-centre build-out continues at full tilt, and HE Group's Cyberjaya subcontract does exactly what management hoped - serves as the reference that unlocks a string of follow-on data-centre electrical packages from the same EPCM contractors and new hyperscaler-backed owners. The ~RM675m tender book, four-fifths of it data centres, converts steadily into order-book wins, and the order book climbs back above its IPO level and stays there. The new Kedah and Johor offices put the company on the doorstep of both the northern semiconductor belt and the southern data-centre corridor, so it wins work in both engines at once. Digitalflux adds an adjacent IT-and-engineering revenue line that smooths the project lumpiness. Margins hold or improve as capacity among qualified mid-tier contractors tightens against surging demand. The founder's open-market buying through 2026 turns out to have been well-timed. In this world HE Group graduates from a small fab-electrical contractor into a diversified critical-infrastructure electrical specialist riding the single biggest construction wave in Malaysia's history.

Base case. Management delivers roughly what the trajectory implies. The data-centre segment is real and growing but contract awards arrive in lumps with the occasional timing delay tied to grid and power-availability constraints, so revenue and profit are choppy quarter to quarter rather than smoothly compounding - much as FY2025's soft Q4 and the year-on-year revenue dip in Q1 FY2026 already showed. The order book oscillates around current levels as completed projects are replaced by new wins, with the net direction modestly up as the data-centre pipeline outpaces the maturing fab work. The maiden dividend is maintained or grown modestly. The company stays profitable, net-cash, and shareholder-friendly, but remains a cyclical project contractor whose fortunes track Malaysia's industrial and digital capex cycle, with a thin competitive moat that keeps it a price-taker on commoditised packages. A solid, unspectacular ride on a strong tailwind.

Bear case. The data-centre boom proves more constrained, or more delayed, than the headline investment numbers suggest - Malaysia's grid power and Johor's water shortfalls stall the project pipeline, hyperscaler capex moderates, and the award timing that already dented FY2025 stretches out further. The order book, already down from its IPO level, thins faster than new wins replace it, and because revenue is recognised on execution, profit falls with a lag and the quarterly lumpiness turns into a sustained slide. In a softer market the fragmented field of qualified mid-tier contractors competes harder on price, and HE Group's margin - improved recently only on project mix and cost control - compresses. A fixed-price data-centre subcontract overruns on cost or schedule, denting both the P&L and the qualification reputation the whole business rests on. The simultaneous stretch of scaling into larger jobs, opening regional offices, and launching a new subsidiary strains a small organisation built around one key person. None of this is a thesis-ending technology shift - it is the ordinary, specific way a small cyclical contractor disappoints when the capex wave it is riding pauses.

Sources

• The Edge Malaysia - HE Group sets IPO price at 28 sen
• BusinessToday - HE Group plans to raise RM24.33m on Jan 30 ACE Market debut
• The Star - HE Group climbs on ACE Market debut
• Minichart - HE Group Berhad Q1 2026 Financial Results
• The Star - HE Group unit bags RM56.7mil data centre electrical services contract in Cyberjaya
• The Edge Malaysia - HE Group secures RM56.6m electrical services subcontract for Cyberjaya data centre
• BusinessToday - HE Group clinches RM56 million subcontract for Cyberjaya DC job
• Hexatech Engineering (company website)
• KLSE Screener - HE Group Berhad (0296)
• I3investor - HEGROUP overview
• MarketScreener - HE Group (0296)
• White & Case - What is propelling Malaysia's data centre boom?
• Mordor Intelligence - Malaysia hyperscale data center market
• 360DigiTMG - Top electrical engineering companies in Malaysia
• Insider transaction data: MoatMap database (Bursa Section 219 disclosures via klsescreener), last scraped 2026-05-29.

Note on omitted Section 13: A targeted search of SemiAnalysis, Stratechery, and MBI Deep Dives returned no coverage of HE Group Berhad. None of these three independent analysts has written about this company, so the "Further Reading" section is omitted by design rather than padded.