MY-DC Tactical Atlas — 2026-05-28

Top 3 Signals This Cycle

1. KL-Johor Corridor Land Repriced on Compute Density, Not PSF

Why it matters: Industrial land valuation methodology is structurally shifting — secured power allocation and fiber proximity now dominate over area-based pricing. For Vincent's Alor Pongsu plot (Perak), this means the upside is asymmetric IF power can be secured; without it, the land trades at conventional industrial discount. Voices: Western-A, Western-B, Western-C, China-C (4/6, cross-camp) Dates: 2026-04-06, 2026-05-26 (fresh) Horizon: Quarter Risk Grade: B | Confidence: 4/5

2. Privasia/Inspur Bagan Datuk RM569M EPCC Active in Perak

Why it matters: This is the only hard-dated, priced, China-EPCC comparable INSIDE Perak. Phase 1 = 10.06MW, RM569M EPCC, completion end-2027, ops early 2028. Implies ~RM56.5M/MW all-in EPCC benchmark. Bagan Datuk sits ~40km from Alor Pongsu — same TNB substation catchment logic, same state-level permitting regime. Vincent now has a real comp, not a Johor extrapolation. Voices: Western-A, Western-B, Western-C, China-C (4/6, cross-camp) Dates: 2026-04-10 (fresh) Horizon: Quarter Risk Grade: A | Confidence: 5/5

3. Northern Corridor Forming — Kulim, Penang, Perak as Overflow Geography

Why it matters: Princeton Digital 72MW Kulim (construction Q4 2026), GDS/K2 into Penang, TNB fast-tracking northern substations because Johor/Selangor grid is saturated. The northern corridor is no longer speculative — it is now state policy and operator behavior. Alor Pongsu is geographically inside this corridor. Voices: Western-B, China-C (cross-camp, smaller cluster but directionally reinforced by Signal 2) Dates: 2026-02-24, 2026-05-25 (fresh) Horizon: Year Risk Grade: B | Confidence: 4/5

Top 3 Blindspots

1. Western-Only View on Grid Capacity Cliff (Post-2030)

The split: All three Western voices flag the Peninsular grid as adequate to 2029 but facing a structural deficit post-2030 as coal retires — requiring either 23GW new gas (76% gas utilization rise) or aggressive clean-energy build. DC load already at 3% of grid and tripled in 9 months of 2025. China-C is silent on this. Why this is a blindspot: Chinese-camp analysts may be assuming Chinese EPCC + Chinese-financed captive power (solar+BESS, behind-the-meter gas) routes around TNB grid risk entirely. If correct, Western voices are overweighting a constraint that Chinese capital is already engineering around. If wrong, Chinese-backed Perak projects hit a wall in 2030. Date: 2026-05-25 (fresh) Risk Grade: B | Confidence: 3/5 Action implication: Vincent must ask any Chinese counterparty explicitly: captive power plan? PPA structure? Or TNB ESA-dependent?

2. China-Only Read on Strategic Capital Routing via JS-SEZ

The split: China-C frames GDS Iskandar Puteri 100MW as "Beijing-capital strategic deployment into JS-SEZ" — geopolitical positioning, not just commercial overflow. Western voices treat it as straightforward Singapore IMDA moratorium overflow. Why this is a blindspot: If China-C is right, the deal flow into Malaysia is driven by capital-controls / sanctions-routing logic, which has different durability and policy-risk profile than commercial overflow. US export control tightening could either accelerate or kill this depending on direction. Date: 2026-04-30 (fresh) Risk Grade: C | Confidence: 3/5

3. Singapore Overflow Durability — No Voice Stress-Tests It

The split: All voices assume Singapore IMDA moratorium = permanent tailwind for Johor/Malaysia. None model the scenario where Singapore relaxes the moratorium (it has done so partially before) or where Indonesia/Batam captures overflow instead. Date: undated_estimate — DATE_UNKNOWN, verify before action Risk Grade: C | Confidence: 2/5

Top 3 Contrarians

1. Western-A Flags Low-Confidence Signals (0.25–0.35) on Weekly Horizons

Claim: Two Western-A signals carry confidence 0.25 and 0.30 — unusually low. Both are short-horizon (week). This pattern suggests Western-A is tracking high-noise, high-velocity rumours (likely operator announcements or TNB internal leaks). Why watch: Low-confidence weekly signals from a single voice are exactly where edge lives if you have ground-truth verification capacity in Perak. Vincent does. Dates: 2026-05-26, 2026-05-27 (fresh) Risk Grade: D | Confidence: 2/5

2. Western-B: "Geographic Shift Is Now State Policy"

Claim: Northern Malaysia (Perak, Kedah) DC viability is no longer an operator-led overflow — it is being formally assessed by authorities for grid expansion. This is a stronger framing than the other voices use. Why watch: If true, Vincent's Alor Pongsu plot benefits from state-driven infrastructure investment (substation, transmission) rather than having to bootstrap it. Verify via TNB/MIDA bulletins. Date: 2026-02-24 (borderline-fresh, ~93 days old — re-verify) Risk Grade: B | Confidence: 3/5

3. China-C Solo: Bagan Datuk as "China-Malaysia Commercial Precedent"

Claim: Frames Privasia/Inspur not as a commercial deal but as a template — implying more Chinese EPCC + Malaysian-sponsor JVs will follow this exact structure in Perak. Why watch: If this template thesis is correct, Vincent's optimal play is to position as the next Malaysian-sponsor counterparty for an Inspur/Sugon/H3C EPCC, not to build solo or sell raw land. Date: 2026-04-10 (fresh) Risk Grade: B | Confidence: 3/5

Tracked Forecasts Update

FORECAST: A second China-EPCC datacenter project (Inspur, Sugon, or H3C as EPCC) will be publicly announced in Perak with a Malaysian sponsor JV structure mirroring Privasia/Silverstreams | HORIZON: quarter | VERIFY_AFTER: 2026-08-28

FORECAST: TNB will publish or leak a formal Northern Corridor grid expansion plan (Perak/Kedah substation upgrades >500MVA) | HORIZON: quarter | VERIFY_AFTER: 2026-08-28

FORECAST: KL-Johor industrial land transactions will show explicit RM/MW or RM/kW pricing language in at least one major broker report or listed-co disclosure | HORIZON: month | VERIFY_AFTER: 2026-06-28

FORECAST: Princeton Digital Kulim 72MW project will commence physical site works on schedule in Q4 2026 | HORIZON: quarter | VERIFY_AFTER: 2026-12-15

FORECAST: Peninsular grid DC load will cross 5% of total demand (currently 3%) | HORIZON: year | VERIFY_AFTER: 2027-05-28

FORECAST: At least one Singapore-moratorium-driven hyperscaler will announce a Penang or Perak (not Johor) site selection | HORIZON: year | VERIFY_AFTER: 2027-05-28

Opportunity Map — Vincent-Specific

Signal 1: Land Repricing on Compute Density

90-day action: Commission a formal "compute density valuation" memo on Alor Pongsu — include TNB substation distance, available MVA at nearest 132kV/275kV node, fiber backbone proximity (TM, Allo, Maxis), and water table data. Use this as the marketing document for any JV / sale discussion, not a generic industrial land flyer. Estimated RM cost: RM 80,000–150,000 (independent power consultant + valuer + survey) 30-day disprove test: Request informal indicative power availability letter from TNB Grid Division for Alor Pongsu coordinates. If TNB returns "no capacity before 2030," compute-density thesis dies for this plot — revert to industrial valuation.

Signal 2: Privasia/Inspur Bagan Datuk as Live Comp

90-day action: Get in the room with Privasia, Silverstreams, or Inspur Malaysia BD. Bagan Datuk is 40km away — they know the Perak permitting timeline, the TNB constraints, and likely have surplus EPCC capacity. Position Alor Pongsu as a Phase 2 / sister-site offering, not a competing project. Estimated RM cost: RM 20,000–50,000 (travel, intermediary fees, legal NDA prep) 30-day disprove test: If Inspur Malaysia / Silverstreams will not take an introductory meeting within 30 days, the "template thesis" (China-C contrarian) is weak — Vincent should pivot to Western hyperscaler intermediaries or local Malaysian operators (YTL, Bridge DC, AIMS).

Signal 3: Northern Corridor as State Policy

90-day action: File a formal expression of interest with MIDA and Perak state EXCO (UPEN Perak) declaring Alor Pongsu as a DC-ready site. This costs little but creates a paper trail that puts the plot on state-level radar before TNB grid expansion plans are finalized. Estimated RM cost: RM 15,000–40,000 (legal drafting, MIDA submission consultant, one EXCO engagement) 30-day disprove test: If MIDA / UPEN response is generic or routes Vincent to standard industrial channels rather than DC-specific desk, the "state policy" framing is overstated — northern corridor is still operator-led, not policy-led, and Vincent has less tailwind than assumed.

Confidence & Coverage Note

Coverage: Strong on Perak-specific signals (Bagan Datuk anchor comp is gold-standard). Adequate on KL-Johor corridor. Weak on water/cooling specifics for northern Perak, weak on Singapore moratorium durability scenarios, weak on US export-control second-order effects on Chinese-EPCC Malaysia projects.

Confidence: Two true cross-camp consensus clusters (size 4, both camps represented) is unusually strong signal — Signals 1 and 2 carry high conviction. Most "outlier" clusters are Western-A artifacts (metadata clustering, not substantive divergence) and should be discounted as noise in the matrix, not as genuine analytical outliers. Genuine outlier voices are limited.

Freshness: 8 of ~12 dated signals fall within 90-day window. One Western-B northern-corridor signal (2026-02-24) is borderline — re-verify before acting on Signal 3 hard.

Gap to fill next cycle: Water allocation for Perak DCs (Sungai Perak abstraction limits), and explicit US export-control posture toward Malaysia-sited Chinese-EPCC AI infrastructure.