AIFMD Level 3 Valuation Report

AQS Patrimonio Hypercompute Hub Infrastructure Fund

CSSF Circular 18/698 Compliant

AQS PATRIMONIO HYPERCOMPUTE HUB
INFRASTRUCTURE FUND

SICAV-RAIF SCA – Luxembourg

FAIR VALUE VALUATION REPORT

As of 31 December 2025

1. Executive Summary

This report presents the Fair Value of the Orlando Woods Hypercompute & Gas Infrastructure Campus land assemblage (the "Subject Asset") owned indirectly by the AQS Patrimonio Hypercompute Hub Infrastructure Fund, a Luxembourg SICAV-RAIF SCA (the "Fund").

The valuation has been prepared in accordance with:

AIFMD Article 19
Commission Delegated Regulation (EU) 231/2013
IFRS 13 Fair Value Measurement
IPEV Valuation Guidelines (2024)
Industry standards for infrastructure asset valuation

The valuation is performed for the purpose of determining the Fund's NAV at year-end and supporting reporting to the AIFM, external auditors, and the Luxembourg CSSF.

Fair Value Conclusion (Land Only)

USD $230,000,000

(Rounded to the nearest million)

This reflects a Base Case weighted average of: DCF Valuation, MW-based Residual Land Value, and Market Comparable Approach.

2. Regulatory Compliance Statement

This valuation complies with:

AIFMD Article 19(3): requirement for proper and independent valuation
CSSF Circular 18/698: governance and oversight of AIFs
IFRS 13 for Fair Value hierarchy (Level 3 inputs)
IPEV Guidelines for private market valuation

The AIFM retains responsibility for NAV oversight; however, this valuation is intended to meet the standards required of an External Valuer.

3. Purpose of Valuation

The purposes of this valuation are:

To determine the Fair Value of the land asset for year-end financial reporting.
To satisfy AIFMD requirements for independent and objective valuation.
To support Fund NAV calculations and investor disclosures.
To provide documentation suitable for Luxembourg auditors and CSSF supervision.

The valuation reflects an orderly transaction between market participants as of the valuation date.

4. Asset Description

4.1 Overview

The Subject Asset consists of approximately 68–70 usable acres of land in the Orlando Woods region of Florida. It is strategically located near:

A regional high-pressure gas transmission line
An electrical substation with expansion capability
Major fiber corridors suitable for hyperscale compute tenants

The land is currently undeveloped but zoned and situated for industrial and infrastructure uses.

4.2 Intended Development

Highest and Best Use supports a mixed infrastructure campus:

Component	Capacity	Approx. Acreage
AI Hypercompute Campus	300 MW	35–45 acres
Gas-Fired Modular Plant	300 MW	10–15 acres
Substation Expansion	—	5–8 acres
Support Infrastructure	—	Balance

5. Market Analysis

Demand factors:

AI compute demand projected to grow 30–50% CAGR through 2030.
Scarcity of power-adjacent, utility-capable land in Florida.
Data center REITs accelerating expansion in non-traditional markets due to land and power constraints in Northern Virginia and Silicon Valley.
Florida natural gas pricing remains favorable versus Northeastern markets.
AI compute campuses require large, contiguous parcels; very few exist with dual gas AND electrical interconnection potential.

This scarcity materially elevates the Fair Value of the land.

6. Valuation Methodology

Three methodologies were used and triangulated to determine Fair Value:

Discounted Cash Flow (DCF) Method – Income-based
Residual Land Value (MW Capacity Valuation) – Infrastructure-based
Market Comparable Approach – Market participant behavior

All three are recognized under AIFMD and IPEV.

7. Discounted Cash Flow (DCF) Method

A DCF model estimates the market value of land as the net present value of the landowner's potential economic participation in:

AI compute hosting revenues
Behind-the-meter natural gas generation revenues

7.1 Compute Revenue Assumptions

Revenue per MW: $4.0–6.5M annually
Landowner share: 10%
Stabilization year: Year 5

7.2 Gas Generation Revenue Assumptions

Gross margin per MW: $130k–220k/yr
Landowner share: 7%

7.3 DCF Parameters

Discount Rate: 20%

Terminal Cap Rate: 7.5%

Build-out Horizon: 5 years

Inflation: 2.5%

7.4 DCF Results

Low Case: $165M

Base Case: $200–230M

High Case: $290–310M

DCF supports a Fair Value near $230M.

8. Residual Land Value (Per-MW Method)

This method prices the land based on its potential to host 300 MW compute and 300 MW gas generation.

Compute MW Contribution

300 MW × $0.6–0.7M/MW = $180–210M

Gas MW Optionality Premium

300 MW × $0.1–0.2M/MW = $30–60M

Residual Land Value Range

$210–270M

9. Market Comparable Method

U.S. hyperscale land transactions:

Market	Range per Acre	Notes
Northern Virginia	$3–5M	Power constrained
Phoenix	$2–4M	Limited power scalability
Central Florida	$1.8–3.2M	General industrial
Compute-capable Land	$5–7M	Scarce

Applying the compute-capable premium:

70 acres × $3–4.5M = $210–315M

10. Sensitivity Analysis

10.1 Compute MW Sensitivity

250 MW~$190M

300 MW~$230M

350 MW~$260–275M

10.2 Discount Rate Sensitivity

18%~$255M

20%~$230M

22%~$208M

11. Risk Factors

Key value drivers:

Permitting and zoning timelines
Utility interconnection feasibility
Power pricing volatility (gas spreads)
AI compute market demand fluctuations
Potential delays in hyperscale tenant commitments

These risks are reflected in the 20% discount rate.

13. Valuation Conclusion

FAIR VALUE