MY-DC Tactical Atlas — 2026-06-10

Top 3 Signals This Cycle

1. Johor Land Arbitrage vs Singapore Is Structural, Not Cyclical

Why it matters: Iskandar industrial land at RM25–40 psf vs Singapore SGD800–1,200 psf is a ~30x differential. 42 new Johor DC projects approved in a single quarter. Sedenak Tech Park (STeP) already surging to RM38–45 psf. This is the bedrock thesis for Malaysian DC capital deployment.

Voices: Western-A, Western-B, Western-C, China-C (4/6 voices, cross-camp)

Date: 2026-06-05 to 2026-06-08 (fresh)

Horizon: Quarter

Risk Grade: B | Confidence: 5/5

2. Alibaba Cloud Johor Region Confirms Malaysia as Regional AI Hub

Why it matters: Alibaba launched a new public cloud region in Johor with two new DCs, part of a global $53B cloud/AI buildout, pushing Johor to 41 total DCs. China hyperscaler commitment signals demand durability beyond US tenants — important hedge against any US export-control turbulence affecting AWS/Azure/Google AI workloads.

Voices: Western-A, Western-B, Western-C, China-C

Date: ~2026-06-09 (fresh)

Horizon: Month to Quarter

Risk Grade: B | Confidence: 5/5

3. Hard Grid Cap: 500 MW Johor Southern Grid Limit Until 2028

Why it matters: Energy Commission imposing aggregate 500 MW cap on new Johor southern grid connections until 2028 transmission upgrades. Klang Valley substations (Bandar Sunway, Shah Alam, Puchong) at 87% utilization — flagged stressed. This converts land-arbitrage thesis into a power-allocation race: whoever locks ESA early wins; latecomers get nothing or pay premium.

Voices: Western-B, Western-C, China-B, China-C (cross-camp consensus)

Date: 2026-06-04 to 2026-06-07 (fresh)

Horizon: Year (binding until 2028)

Risk Grade: A (binding constraint) | Confidence: 5/5

Top 3 Blindspots

1. TNB 7.1 GW ESA Pipeline + 85% Utilization Mandate — Western-Only Visibility

The gap: Western-A/B/C are tracking TNB's 49 signed Electricity Supply Agreements (7.1 GW = ~13–25% of Peninsular peak) and the EC's 85% utilization enforcement target on declared maximum demand. China voices are NOT picking this up. This is the single most important supply-side signal — it's the enforcement mechanism that prevents speculative land/power banking and will trigger ESA cancellations within 12–18 months.

Why blindspot matters: If you're courting Chinese hyperscaler tenants, they may not yet appreciate that 85% utilization is non-negotiable. Cancelled ESAs = secondary market for power allocation = arbitrage opportunity for nimble operators.

Date: 2026-03-25 to 2026-06-04

Risk Grade: A | Confidence: 4/5

2. Singapore Moratorium Is a Strategic Pivot, Not a Shutdown — Limited Voice Coverage

The gap: Only Western-B and China-C are framing Singapore's IMDA DC-CFA2 (200 MW green allocation, >50% green power requirement) as a deliberate filter rather than a hard cap. Implication: Singapore is keeping high-margin AI/green workloads and exporting commodity compute to Johor. The IMDA also reportedly delayed lifting the moratorium by 6 months citing cross-border carbon accounting disputes with Malaysia.

Why blindspot matters: Malaysian DC operators chasing premium AI workloads may be competing against a revitalized Singapore, not a vacated one. Carbon accounting disputes are an under-priced regulatory risk.

Date: undated_estimate for moratorium delay — DATE_UNKNOWN, verify before action

Risk Grade: B | Confidence: 3/5

3. Specific Hyperscale Construction Starts (GDS, PDG) — Single-Source Risk

The gap: GDS Holdings 100 MW Johor Bahru campus (MIDA-approved, July 2026 construction start) and PDG MY03 / Cyberjaya 30 MW — reported by single voices (China-C, Western-A) without cross-corroboration. Western-A itself flagged PDG MY03 as "provisional, treat with caution."

Why blindspot matters: These are the competitive set for Vincent's Perak/Alor Pongsu thesis. If GDS lands 100 MW in JB on the southern grid, it eats meaningfully into the 500 MW cap. Need independent verification before site selection.

Date: 2026-06-08 (provisional)

Risk Grade: C | Confidence: 2/5

Top 3 Contrarians

1. Perak / Northern Corridor as Grid-Cap Hedge (Implicit Outlier)

Signal: Notably ABSENT from consensus — no voice is yet flagging Perak (where Vincent's Alor Pongsu plot sits) as the natural overflow from Johor's 500 MW cap. The market is still chasing JB/Sedenak. Selangor stressed at 87%. By process of elimination, Perak's northern grid headroom + TNB Manjung baseload proximity becomes the contrarian secondary market.

Why watch: First-mover advantage in Perak is currently un-priced. Voices haven't connected "JB capped → Selangor stressed → Perak next."

Date: undated_estimate (inference from absence)

Risk Grade: C | Confidence: 2/5

2. Singapore-Malaysia Carbon Accounting Dispute as Regulatory Wildcard

Signal: China-C alone flagged that the Singapore moratorium delay is tied to cross-border carbon accounting disputes with Malaysia. If Singapore demands Scope 2 emissions accounting for workloads served from Johor, Malaysian DCs lose their cost arbitrage on premium green-mandated tenants.

Why watch: This is the kind of bureaucratic detail that becomes a structural ceiling on the cheapest-and-dirtiest Malaysian DC plays. Renewable PPA strategy becomes mandatory, not optional.

Voice: China-C only

Date: DATE_UNKNOWN — verify before action

Risk Grade: B | Confidence: 2/5

3. Cyberjaya Rack Pricing 40% Below Singapore + Johor Taskforce 12% Land Cut

Signal: Western-C reports Cyberjaya rack pricing 40% cheaper than central Singapore AND the Johor Taskforce just cut hyperscale land prices by another 12%. Single voice. If accurate, the arbitrage is widening, not stabilizing — implies either (a) demand softening, or (b) Malaysian states actively competing on price to win allocation before grid cap binds.

Why watch: Direction matters. Widening spread = race-to-bottom risk for late entrants; flatlining spread = oligopoly forming.

Voice: Western-C only

Date: 2026-06-07 to 2026-06-08

Risk Grade: C | Confidence: 3/5

Tracked Forecasts Update

FORECAST: Alibaba Cloud Johor region drives at least 2 additional China hyperscaler DC announcements (Tencent, Huawei, or ByteDance) in Malaysia | HORIZON: quarter | VERIFY_AFTER: 2026-09-10

FORECAST: At least 5 of TNB's 49 signed ESAs will be cancelled or restructured under the 85% utilization rule, freeing >500 MW for re-allocation | HORIZON: year | VERIFY_AFTER: 2027-06-10

FORECAST: Johor southern grid 500 MW cap will be exhausted by Q1 2027, forcing >1 GW of new hyperscale demand to Selangor/Perak/Penang | HORIZON: quarter | VERIFY_AFTER: 2027-03-31

FORECAST: Singapore IMDA will formally lift moratorium with green-conditional allocation by Q4 2026, NOT delay further | HORIZON: quarter | VERIFY_AFTER: 2026-12-31

FORECAST: Perak will see at least 1 hyperscale (>50 MW) DC announcement as Johor cap binds | HORIZON: year | VERIFY_AFTER: 2027-06-10

FORECAST: Iskandar industrial land prices will breach RM50 psf by mid-2027 as supply tightens | HORIZON: year | VERIFY_AFTER: 2027-06-10

Opportunity Map — Vincent-Specific

Signal 1: Johor Land Arbitrage (already in motion)

90-day action: Do NOT chase Johor land at current prices — it's pricing in 2028 demand already. Instead, commission a comparative valuation: Alor Pongsu (Perak) vs. Kulim Hi-Tech Park (Kedah) vs. secondary Sedenak parcels. Goal: identify Tier-2 land within 5 km of a TNB 275 kV substation with confirmed headroom.

Estimated RM cost: RM40k–80k (independent valuation + grid headroom letter from TNB regional office).

30-day disprove test: Request a TNB Tenaga Letter of Intent for ≥30 MW at Alor Pongsu site. If TNB cannot commit within 30 days, the northern-corridor hedge thesis is weak and you should redirect to Selangor brownfield.

Signal 2: Alibaba Cloud / China Hyperscaler Demand

90-day action: Engage Alibaba Cloud Malaysia BD team + GDS Holdings Singapore office as potential anchor tenants. China hyperscalers are under-served by Malaysian operators who default to AWS/Azure/Google channels. Position your site as China-friendly (data sovereignty narrative, RMB-denominated lease optionality).

Estimated RM cost: RM25k–50k (3 trips KL/Singapore + bilingual deck + legal review of cross-border data terms).

30-day disprove test: Secure one signed NDA + technical site visit from a China hyperscaler within 30 days. If no engagement, the China-tenant thesis is not actionable at your scale and you revert to Tier-2/colo operator tenants (Bridge, AIMS, NTT).

Signal 3: 500 MW Johor Grid Cap + Power Allocation Scarcity

90-day action: File an Electricity Supply Agreement (ESA) application with TNB for Alor Pongsu site at the maximum credible declared maximum demand (e.g., 20–40 MW initial, expandable). Even if you don't build immediately, the ESA queue position is the asset. Engage a former TNB executive as advisor.

Estimated RM cost: RM80k–150k (ESA application fees + advisor retainer 3 months + technical feasibility study for grid connection).

30-day disprove test: Submit ESA application within 30 days. If TNB indicates >18-month queue for Perak northern grid, the "Perak is hedge" thesis collapses and you should pivot to either acquiring an existing ESA holder or partnering with a brownfield operator.

Confidence & Coverage Note

Coverage: Strong on Johor/Singapore axis (4/6 voices converging). Weak on Perak specifically — no direct voice signal despite topic being a stated focus area; conclusions are inferences from grid-cap displacement logic. Weak on Penang/Kedah.

Date freshness: ~80% of consensus signals dated within last 14 days (2026-05-27 to 2026-06-09) — high freshness. Singapore moratorium delay and PDG/GDS construction starts have DATE_UNKNOWN or single-source flags.

Camp balance: Western voices over-represented (3 vs 2–3 China). Cluster-31 (TNB 7.1 GW ESA pipeline) is a pure-Western cluster — recommend China-side verification before treating as fact.

Confidence in brief overall: 4/5 — high on the macro thesis (Johor arbitrage, grid cap binding), medium on the contrarian Perak hedge (your specific deployment site), low on Singapore policy timing.