The Australian industrial real estate sector is currently navigating a structural discontinuity that challenges the foundational tenets of property valuation. For decades, the “Highest and Best Use” (HBU) of industrial land was predicated on “Logistics Utility”—a derivative of transport economics, measuring the efficiency of goods movement from port to warehouse to consumer. Our analysis confirms that this paradigm has fractured. We are witnessing a definitive HBU Reset where valuation has decoupled from transport access and re-anchored to “Power Utility”—a derivative of the energy market, measuring the capacity to energise high-density digital compute workloads.

The stress test of this hypothesis centres on the Goodman Group’s aggressive strategic pivot, which serves as the primary market signal for this transition. The data unequivocally validates the “Crowding Out” hypothesis. Goodman’s reported 68% allocation of its Work in Progress (WIP) to data centres is not merely a portfolio rebalance; it is a capital flight from the traditional logistics sector.^[1] This reallocation of capital is driven by a fundamental arbitrage opportunity: the “Residual Land Value (RLV) Gap™.”

Our analysis quantifies this gap, revealing that digital infrastructure operators, capitalising on the exponential margin potential of Artificial Intelligence (AI) and cloud computing, are acquiring land at premiums ranging from 50% to over 100% above the maximum residual capacity of traditional logistics developers.^[3] This creates a financial firewall around grid-proximate land, effectively displacing logistics tenants to the un-serviced fringe and fracturing the industrial market into two distinct asset classes: “Powered Industrial” (Digital) and “Passive Industrial” (Logistics).

However, the investigation also uncovers a critical counter-narrative. While the valuation pivot is supported by a massive supply-demand imbalance—quantified by M3 Property as a 1.7 GW forecast supply gap^[5]—the operational reality is constrained by severe “Grid Lock.” Connection delays, transmission congestion, and the emergence of “load shedding” clauses in connection agreements suggest that a significant portion of the industry’s “Secured Power” pipeline functions more as a valuation hedge than immediate supply.^[6] The “Power Bank” has replaced the “Land Bank” as the primary determinant of Net Tangible Assets (NTA), yet its liquidity is constrained by the physical laws of the electrical grid.

This report provides an exhaustive analysis of these dynamics, moving from the primary verification of the Goodman pivot through to the granular mechanics of the land valuation fracture and the technical constraints of the energy market.

This section requires the rigorous verification of the market signals that underpin the HBU Reset Hypothesis. Specifically, we must validate the magnitude of the Goodman Group’s strategic reorientation and the macroeconomic supply gap identified by M3 Property that justifies such a radical deployment of capital.

The Goodman Group Q1 FY26 Operational Update: A Structural Anomaly

The release of the Goodman Group Q1 FY26 Operational Update on November 5, 2025, marks a watershed moment in the history of Australian Real Estate Investment Trusts (REITs). The document explicitly confirms a capital allocation strategy that subordinates the company’s legacy logistics operations to its burgeoning digital infrastructure platform.

Verification of the 68% WIP Figure

The investigation confirms the primary datum: as of September 30, 2025, Goodman Group reported a total development Work in Progress (WIP) of $12.4 billion.^[1] Critically, the update reveals that data centres now represent 68% of this total WIP.^[1]

This figure represents a capital commitment of approximately $8.43 billion solely to data centre projects within the active development pipeline. To contextualise this magnitude, the remaining 32% of the WIP—allocated to traditional logistics, multi-storey warehousing, and other industrial uses—amounts to only $3.9 billion. This inversion of the development workbook indicates that for every dollar Goodman invests in moving physical goods, it is investing more than two dollars in processing digital data.

The update further projects that by June 2026, the total WIP will expand to exceed $17.5 billion.^[1] Management has indicated that this growth will be driven by major data centre commencements, with the explicit goal of lifting the data centre WIP power capacity to approximately 0.5 GW by that date.^[8] This forward guidance suggests that the 68% figure is not a peak, but potentially a stepping stone to an even higher concentration of digital infrastructure assets, structurally cementing the Group’s pivot.

The “Global Power Bank”: Redefining Asset Measurement

The operational update introduces and emphasises a metric that is alien to traditional industrial property valuation but central to the Digital Utility model: the “Global Power Bank.” Goodman reports a total power bank of 5.0 GW across 13 major global cities.^[9]

The composition of this Power Bank provides insight into the “realness” of the pipeline:

• Secured Power (3.4 GW): This tranche represents capacity where grid connections are approved and land is secured. It is the “hard” asset base of the digital strategy.^[11]
• Advanced Procurement (1.6 GW): This tranche represents capacity in negotiation. It is the “soft” or “option” value of the portfolio.^[11]

The shift in nomenclature is profound. By reporting in Gigawatts rather than Gross Lettable Area (GLA), Goodman is signalling to the market that the scarcity value of its portfolio lies in the grid connection, not the floor space. The report highlights that these sites are predominantly located in “highly supply-constrained, metropolitan markets”^[1], implicitly acknowledging that the value proposition is the convergence of land scarcity and power scarcity.

Geographic Distribution and “Tier 1” Focus

The Q1 FY26 update and associated documents reveal that this pivot is global, but highly concentrated in markets with the most acute power constraints. The project list includes:^[2]

• Sydney, Australia (SYD01): 90 MW in the Macquarie Park Availability Zone.
• Tokyo, Japan (TYO05): A massive 1,000 MW campus plan, with 50 MW in the first phase.
• Paris, France (PAR01/PAR02): Combined 305 MW capacity across North and Central Availability Zones.
• Frankfurt, Germany (FRA02): 90 MW in the Frankfurt South Availability Zone.

This geographic dispersion confirms that the “crowding out” phenomenon is not unique to Australia but is a global structural trend where high-value digital use cases are displacing logistics in every major Tier 1 market.

The M3 Property Baseline: Quantifying the Supply Fracture

To understand why Goodman is abandoning its traditional safe harbour of logistics for the volatile, capital-intensive world of data centres, we must look at the underlying market disequilibrium. The M3 Property “Data Centre Growth In Australia” report (November 24, 2025) provides the macroeconomic justification.

The “Critical Inflection Point”

M3 Property identifies that the Australian data centre market has reached a “critical inflection point where demand is outpacing supply”.^[5] This is not a cyclical tightness but a structural deficit. The report notes that for the first time on record, the volume of live capacity taken up by users in 2023 and 2024 exceeded the amount of new supply brought to market.^[12]

The 1.7 GW Supply Gap

The most critical data point for our stress test is the forecast supply gap. M3 Property forecasts a gap of up to 1.7 GW between projected demand and available supply.^[5]

• Current Occupancy: 1,315 MW.^[5]
• Historical Growth: A fortyfold increase from 37 MW in 2005.^[5]
• Future Requirement: The gap implies that the market needs to deliver more capacity in the next few years than has been built in the last two decades combined.

The AI Multiplier

The driver of this gap is explicit: “facilities designed specifically for hosting AI workloads have seen demand increase nearly 27 times”.^[12] This demand shock explains the “Power Premium.” AI workloads are not price-sensitive in the same way as traditional cloud storage; the revenue generated per rack of AI compute is so high that the underlying cost of land and power becomes a secondary consideration, allowing operators to bid up land values aggressively.

Pre-Commitment Levels

M3 Property reports that as of December 2024, 68% of all data centre capacity under construction has already been pre-committed, well above the long-term average of 38%.^[5] This statistic mirrors Goodman’s 68% WIP figure, suggesting a perfectly rational market response: developers are building exactly what the market has already bought. This high pre-commitment level de-risks the pivot, as the capital is being deployed into a vacuum of demand rather than a speculative glut.

Synthesis: Primary Source Verification

The primary source verification confirms the HBU Reset Hypothesis. The Goodman Group’s 68% pivot is a direct, calculated response to the 1.7 GW supply gap identified by M3 Property. The correlation between the “supply gap” and the “investment shift” is near perfect. Goodman is not “crowding out” logistics arbitrarily; it is acting as a rational economic agent, reallocating resources to the sector where the marginal utility (and thus the marginal value) is exponentially higher. The validation of the 1,315 MW occupancy figure against the 1.7 GW gap confirms that the market is structurally short of power, creating the prerequisite conditions for a valuation reset based on “Power Utility.”

To prove that the current market state represents a structural “crowding out” rather than a gradual evolution, we must analyse the velocity of the change. A slow transition allows the market to adjust; a rapid shock fractures the market, leaving incumbents (logistics tenants) stranded.

The Acceleration of the Digital Pivot (FY24 – FY26)

Tracing the trajectory of Goodman’s capital allocation over the past 18 months reveals a pivot velocity that is unprecedented for a REIT of this size.

FY24 Baseline: The Diversified Model

In the full-year results for FY24 (ended June 30, 2024), Goodman presented a balanced portfolio.

• Data Centre WIP: 40%.^[14]
• Logistics/Industrial WIP: 60%.
• Narrative: The Group described data centres as a “growing” part of the portfolio, but the core identity remained anchored in logistics. The 40% figure was significant but left the majority of development activity in the traditional sector.

FY25 Transition: The Tipping Point

By the full-year results for FY25 (ended June 30, 2025), the balance had shifted decisively.

• Data Centre WIP: 57%.^[9]
• Logistics/Industrial WIP: 43%.
• Narrative: For the first time, digital infrastructure became the majority component of the development workbook. The “Global Power Bank” metric began to take prominence in investor communications, signalling a shift in how the company measured its own value.

Q1 FY26 Realisation: Total Commitment

The November 2025 update confirms the completion of the pivot.

• Data Centre WIP: 68%.^[1]
• Logistics/Industrial WIP: 32%.
• Narrative: The “pure logistics” developer model has been effectively abandoned. With over two-thirds of development capital tied up in data centres, the logistics arm is now the subsidiary operation.

Table 1: The Goodman Group Pivot Velocity (FY24 – FY26)

Note: While the headline Power Bank figure remained stable at 5.0 GW, the composition improved significantly, with “Secured Power” rising from 2.7 GW to 3.4 GW, indicating the crystallisation of speculative options into tangible assets.

Analysis of Capital Displacement

The velocity analysis reveals a stark reality: in just 15 months, the implied capital allocated to logistics development collapsed by nearly 50% (from $7.80bn to $3.97bn), while capital allocated to data centres surged by 62%.

The Zero-Sum Game

Real estate development is a function of constrained resources: capital, land, and human expertise. Goodman’s shift is a zero-sum game within its own portfolio. Every dollar and every hectare allocated to a data centre is a dollar and hectare removed from the logistics supply chain.

• Land Banking Strategy: The stability of the “Power Bank” at 5.0 GW suggests that Goodman identified these sites early and has been systematically working to upgrade their utility from “land” to “power.”
• Crowding Out Effect: This internal crowding out at Goodman—the market leader—inevitably leads to external crowding out. As the largest developer withdraws from logistics supply, the shortage of warehousing space exacerbates, driving up logistics rents. However, even with rising logistics rents, they cannot catch up to the “Power Premium” of data centres.

The Velocity as a Market Signal

The speed of this transition (doubling exposure in roughly 18 months) serves as a distress signal for the logistics sector. It implies that the “Highest and Best Use” reset was not gradual. It was a step-change function triggered by the sudden explosion of AI demand (as noted in M3 Property’s “27x increase” in AI workload demand^[12]). Goodman’s agility in pivoting capital so rapidly indicates that the RLV Gap opened suddenly and widely, creating an arbitrage opportunity that demanded immediate and total capitalisation.

The “Residual Land Value (RLV) Gap” is the financial mechanism of displacement. It represents the difference between what a data centre developer can afford to pay for land versus what a logistics developer can pay. Our investigation confirms that this gap has widened to a point where traditional logistics is structurally priced out of grid-proximate locations.

Quantifying the “Power Premium” Fracture

The search for specific land valuation deltas in Western Sydney and Melbourne provides empirical evidence of the RLV Gap. The market is now bifurcated, with “Powered Land” trading at a massive premium over “Unpowered Land.”

The Western Sydney Pricing Fracture

The standard baseline for industrial land in Western Sydney provides the control group for our analysis.

• Standard Logistics Land: Cushman & Wakefield and Knight Frank report that prime industrial land values (1-5 hectare lots) in Western Sydney are averaging approximately $1,500 per square metre.^[15] In the Outer West (e.g., Erskine Park), values are slightly lower, averaging around $1,262/sqm.^[17]
• Englobo Land: Large, undeveloped (englobo) parcels typically trade at a discount, in the range of $575 – $625 per square metre^[18], reflecting the cost of subdivision and servicing.

Against this baseline, the data centre transactions stand out as statistical outliers, representing the “Power Premium.”

• Moorebank Record: A development-ready site in Moorebank sold for $2,315 per square metre.^[4] This transaction set a new suburb record. Compared to the standard Western Sydney average of $1,500/sqm, this represents a 54% premium. Compared to unserviced englobo land, the premium exceeds 300%.
• Eastern Creek Acquisition: NextDC acquired the “S7” site (25.8 hectares) for approximately $353 million.^[15] This implies a rate of roughly $1,368 per square metre.
  • Insight: While $1,368/sqm appears close to the $1,500/sqm “retail” rate for small lots, paying this rate for a massive 25-hectare parcel is extraordinary. Typically, bulk land trades at a significant discount (the $575-$625 range). Paying the “retail” rate for a “wholesale” parcel implies a hidden value driver: the grid connection. The site is described as being in “close proximity to a major electricity substation”^[19], validating the “Power Utility” hypothesis.

The Melbourne Multiplier

The fracture is even more pronounced in Melbourne, where land values are generally lower than Sydney, making the data centre premium more visible.

• Amazon Northern Melbourne: Amazon purchased a 12-hectare site for $71.2 million, equating to nearly $600 per square metre. The report explicitly notes that this is “double the usual rate” (approx $300/sqm) for that specific location.^[3]
• Direct Multiplier: The report states that operators like Amazon and Microsoft are paying “between two and three times the average market value for key sites”.^[3]

Table 2: The RLV Gap™ – Comparative Land Values (2025)

The Mechanics of the Gap: Why Logistics Cannot Compete

The “RLV Gap” is not an arbitrary bubble; it is a function of the divergent economic models of the two asset classes.

• Logistics Economics (Volume-Based): The logistics developer’s model is constrained by the tenant’s ability to pay rent, which is a function of the tenant’s operating margin (cents per box moved). Even with tight vacancy (2.1% in Sydney^[20]), logistics rents have a “natural ceiling.” If rents rise too high, the 3PL tenant’s business model breaks, or they move to a cheaper location.
• Data Centre Economics (Density-Based): The data centre developer’s model is driven by the “revenue per kilowatt.” In the AI era, a single rack of NVIDIA H100 GPUs can generate hundreds of thousands of dollars in annual revenue. The “land cost” component of the total stack (Land + Power + M&E + Servers) is negligible compared to the revenue potential.

Result: The data centre developer can absorb a land price of $2,300/sqm because the land is merely a vessel for the grid connection. The logistics developer, however, cannot make the numbers work at that price because their yield (e.g., 5%) on that cost base would require a rent far above the market ceiling.

This creates the structural “Crowding Out.” The data centre operator does not just “outbid” the logistics developer; they fundamentally reprice the asset class of “grid-proximate land” to a level that renders logistics development financially impossible.

Tenant Displacement: The Gentrification of Industry

The “Power Premium” fracture is actively displacing existing tenants, a phenomenon we can term the “Gentrification of Industry.”

South Sydney: The High-Value Squeeze

South Sydney (Mascot, Botany, Alexandria) is the epicentre of this displacement. The region is seeing a “widening rental discount” between Western precincts and South Sydney, with South Sydney rents growing by 14% in 2024 alone.^[17]

• Displacement Mechanism: Traditional industrial users (automotive repair, food services, standard warehousing) are being driven out by “higher intensity use outcomes,” which include both multi-story logistics and data centres.^[21]
• Rental Shock: As grid-proximate land is removed from the logistics pool for data centres, the remaining supply becomes hyper-expensive, forcing tenants to either pay the “scarcity premium” or migrate west.

Port Melbourne: The Rezoning Pressure

In Melbourne, the displacement is exacerbated by urban renewal. Areas like Fishermans Bend are being transitioned, and the remaining industrial pockets are highly contested.

• Vacancy: While vacancy in secondary markets is rising, prime vacancy in Melbourne remains extremely tight at roughly 5%.^[13]
• Migration: Reports indicate that businesses are being “priced out of inner and mid-ring industrial areas” and pushed to “fringe or poorly serviced zones”.^[3]

The “Fringe” Trap

The displacement forces logistics tenants to the “Outer West” (e.g., Kemps Creek, Western Sydney Aerotropolis). However, even these frontiers are not safe from the data centre pivot. The Goodman acquisition of land near the new airport^[22] and the NextDC acquisition in Eastern Creek^[19] show that the “Power Premium” is hunting for land in the exact same corridors where logistics tenants are seeking refuge. This suggests a future scenario where even the “fringe” becomes bifurcated into “Powered” (Data Centre) and “Unpowered” (Logistics) zones, potentially pushing pure logistics users to locations with poor transport access, damaging supply chain efficiency.

While the previous section confirms the valuation fracture, this section investigates the deliverability of the data centre pipeline. If Goodman has 68% of its WIP in data centres, but the grid cannot connect them, is the pipeline “real” or “phantom”?

The AEMO Connection Queue: The “Gated” Reality

The Australian Energy Market Operator (AEMO) data reveals a grid in crisis. The transition to renewables, combined with the sudden shock of data centre load, has created severe congestion in the connection queue.

Timeline Blowouts

The standard timeline for completing an interconnection agreement has stretched significantly. Sources indicate that delays can now stretch to “more than six years” in some jurisdictions due to study backlogs and the sheer volume of applications.^[23]

• Queue Volume: AEMO and Network Service Providers (NSPs) are handling “hundreds of applications simultaneously,” overwhelming the limited pool of power engineers.^[24]
• Forecasting Change: Recognising this disruption, AEMO has only recently (2025) updated its forecasting methodology to treat data centres as a “standalone component,” admitting they are a “major driver of electricity demand growth”.^[26] This belated recognition suggests the grid planners are playing catch-up to the property developers.

The “Phantom” Pipeline Paradox

Does a 6-year delay make the pipeline “Phantom”? In a traditional development model, yes. But in the “Highest and Best Use” reset, the delay creates value.

• Scarcity Value: If no new connections can be approved for 5 years, existing “Secured Power” (Goodman’s 3.4 GW) becomes infinitely more valuable. The “Grid Lock” acts as a moat, protecting the incumbents from new competition.
• Valuation Hedge: Goodman’s pipeline is likely valued not just on current cash flow, but on the option value of the connection. By securing the queue position now, they monetise the future capacity. Thus, the 68% WIP figure represents a “Power Bank” of future rights, which the market is willing to price in today because of the evident supply gap (1.7 GW).

The “Load Shedding” Clause: The Reliability Fracture

A critical finding in the investigation is the degradation of power “firmness.” Connection agreements are no longer guaranteeing the 100% uptime that Tier IV data centres require.

Schedule 9: Load Shedding

Investigation of connection agreements (e.g., Clarence Correctional Centre precedent, applicable to large loads) reveals explicit schedules for “Load Shedding”.^[7]

• Mechanism: TransGrid and other NSPs retain the right to interrupt supply during “Load Shed Events” or “abnormal operating conditions”.^[7]
• Manual vs. Automatic: Agreements cover both “Manual Load Shedding” (SCADA commanded) and automatic schemes.^[28]

The Curtailment Conflict

The AEMC (Australian Energy Market Commission) is consulting on rules to mandate “fast ramp down” capabilities for large loads to maintain system security.^[29]

• Industry Pushback: Data centre operators like NextDC have strongly opposed this, arguing that data centres are “critical infrastructure” and “not power stations”.^[30] They contend that mandatory curtailment would force them to switch to diesel backup, increasing emissions and operational risk.
• Valuation Risk: If a site is “Powered” but subject to a “Curtailment Clause,” its value is structurally lower than a site with “Firm Power.” This introduces a new valuation variable: Quality of Connection.
  • Hypothesis: The 3.4 GW of “Secured Power” in Goodman’s portfolio likely varies in quality. Sites with “firm” connections will trade at a premium to sites with “curtailable” connections. The “RLV Gap” may eventually fracture further based on this “Reliability Premium.”

Synthesis: The Grid Lock Reality

The “Grid Lock” reality does not invalidate the Goodman pivot; it calcifies it. The extreme difficulty of securing new connections ensures that the supply gap identified by M3 Property (1.7 GW) will persist for longer than forecast. This persistence supports the “Power Premium” on land that already has a ticket in the queue. The “Phantom” pipeline is actually a “Strategic Reserve” of grid capacity. However, the emergence of load shedding clauses presents a technical risk that must be monitored; if “Powered Land” cannot deliver “Reliable Power,” the valuation bubble could face a correction based on technical underperformance.

This analysis confirms the “Highest and Best Use” (HBU) Reset Hypothesis with a high degree of confidence. The valuation of industrial land in Australia’s major markets has fundamentally re-anchored from “Logistics Utility” to “Power Utility.”

The Mechanics of Displacement

The displacement of logistics by digital infrastructure is driven by a non-linear divergence in the “Residual Land Value” capacity of the two sectors.

• Logistics: Capped by transport margins. RLV ceiling ~ $1,500/sqm (Sydney).
• Data Centres: Driven by compute density and AI revenue. RLV floor ~ $2,300/sqm.
• Result: A structural “RLV Gap™” of >50% that traditional logistics cannot bridge.

The Goodman Signal

Goodman Group’s 68% Data Centre WIP is the definitive market signal of this reset. It represents a rational, aggressive capitalisation of the 1.7 GW supply gap. The rapid “Pivot Velocity” (doubling exposure in 18 months) confirms that this is a structural “crowding out” event, not a cyclical adjustment.

The New Asset Classes

We are no longer analysing a homogeneous “Industrial” sector. The market has bifurcated into:

1. Digital Industrial (Powered): High value, grid-connected, pricing based on $/MW. (e.g., Goodman’s 5.0 GW Power Bank).
2. Analog Industrial (Passive): Capped value, transport-connected, pricing based on $/sqm. (e.g., Traditional warehousing).

Implications for Valuation

Future valuation models for industrial REITs must incorporate “Power Utility” as a primary variable. “Land Banks” must be audited for their “Power Bank” conversion potential. Sites without a pathway to >50MW grid connection face a valuation ceiling, while sites with secured capacity possess intrinsic option value that exceeds their current physical utility.

Final Recommendation

The “Data-Pivot Fracture” is real. Logistics tenants are being structurally displaced to the fringe. Investors must recognise that the “Industrial” label now masks two divergent asset classes with fundamentally different risk/return profiles. The “Highest and Best Use” of any grid-proximate site is now, unequivocally, the storage and processing of data, not goods.

1. Q1 FY26 Operational Update – Goodman Group (ASX:GMG) – Listcorp.
2. Q1 FY26 OPERATIONAL UPDATE – Goodman Group.
3. Data Centre Demand Drives Land Price Surge – The Aussie Corporate.
4. Industrial land sold for record rate per sqm South West suburb – Parkside Commercial Real Estate South Western Sydney | The Industrialist.
5. Australia’s data centre sector expands fortyfold in the past 20 years – M3 Property.
6. Data Centre Connection Woes and Wins – Charles Russell Speechlys.
7. Load Network Connection Agreement – Corrective Services NSW.
8. Goodman Group delivers Q1 FY26 update: data centre growth drives outlook – Motley Fool.
9. Goodman 30 June 2025 Full Year results.
10. Data center property development – Goodman North America.
11. Data centre property development – Goodman Group.
12. Australia’s data centre sector expands fortyfold in the past 20 years – M3 Property.
13. Why More Australian Data is Staying Home | Property News.
14. Goodman Group delivers strong operating profit of $2,049 million and FY24 operating EPS growth of 14%.
15. sydney – logistics & industrial q1 2025 – Cushman & Wakefield.
16. sydney – logistics & industrial q3 2025 – Cushman & Wakefield.
17. Australian Industrial Review – Knight Frank.
18. Industrial Development Update – Colliers.
19. NEXTDC Announces Massive Expansion Plans with Acquisition of S7 Data Centre Site in Sydney.
20. Sydney facing critical industrial land supply deficit | CBRE Australia.
21. The outlook for Sydney’s Industrial & Logistics market | CBRE Australia.
22. Western Sydney Industrial Land Booms with Goodman’s $575M Airport Precinct Acquisition.
23. Powering Data Centers | Megawatts to Megabytes – Orrick.
24. ERC0393 – Final determination – AEMC.
25. Draft determination_Improving the NEM access standards – Package 1 – AEMC.
26. AEMO’s updated forecasting methodology targets rapidly growing electricity loads, following industry consultation.
27. CHAPTER 5 – AEMC.
28. Manual Load Shedding Standard – Australian Energy Market Operator.
29. National Electricity Amendment (Improving the NEM access standards – Package 2) Rule 2025 – AEMC.
30. Improving the NEM Access Standards Package 2 – AEMC.