Why Data Centers are NOT a "Technology Business" ?

【98% of people misunderstand the real business model of data centers!】

Data centers are NOT a “technology business.”

So how do banks actually evaluate and approve loans for building data centers?

1. Why data centers are essentially an “industrial leasing business”

1️⃣ Core revenue model = rent + power + infrastructure

A data center doesn’t make money by “doing tech.” It makes money by:

* Renting out racks

* Leasing space (wholesale / colocation)

* Charging for power usage

* Charging for network connectivity

👉 In essence, it’s like:

* Factory = rent factory space

* Data center = rent “server factories”

✔️ Who are the customers?

* Cloud providers (AWS, Azure, Google Cloud)

* Financial institutions

* E-commerce platforms

👉 They don’t buy buildings — they lease long-term.

2️⃣ Contract structure = long-term leases (5–15 years)

Typical agreements include:

* 5 / 10 / 15-year leases

* With clauses like:

* Minimum commitment (space/power)

* Take-or-pay (pay regardless of usage)

👉 This is critical:

📌 Highly similar to real estate

📌 Predictable, stable cash flow

3️⃣ Investment structure = capital intensive (high CapEx)

Building a data center requires:

* Land

* Buildings

* Power systems (substations)

* Cooling systems

* Network infrastructure

💰 Cost:

👉 Hundreds of millions to billions (RM)

👉 Comparable to:

* Industrial parks

* Warehouses

* Commercial buildings

✔️ Therefore classified as:

👉 Infrastructure + Industrial Real Estate

4️⃣ Operating model = “serviced property”

A normal landlord:

* Just collects rent

A data center operator:

* Provides:

* 24/7 power

* Cooling systems

* Security

* Network reliability

👉 So it’s more like:

📌 “High-end industrial property + technical operations”

2. How banks view data centers

Banks do NOT treat them as tech companies. Instead:

1️⃣ Classified as “infrastructure assets”

Bank categories:

* Property

* Infrastructure

👉 Data centers sit in between

✔️ Key characteristics:

* Long-term stable cash flow

* High barriers to entry

* Strong customer stickiness

👉 For banks, this means:

💡 A high-quality lending asset

2️⃣ The 3 things banks care about most

(1) Tenancy strength

Banks will ask:

* Are there anchor tenants (hyperscalers)?

* Lease duration?

* Take-or-pay clauses?

👉 If tenants include:

* AWS / Microsoft / Google

✔️ Huge credit enhancement

(2) Power capacity ⚡

The core of a data center is NOT the building — it’s electricity.

Banks evaluate:

* Stability of power supply

* Approved capacity (MW)

* Scalability

👉 No power = asset value collapses

(3) Cash flow stability

Banks model it like real estate:

* EBITDA stability

* DSCR (debt service coverage ratio)

* Occupancy rate

👉 If:

* Pre-leased

* Long-term contracts in place

✔️ Banks are very willing to lend

3️⃣ Financing structure (similar to real estate)

Common structures:

* Project financing

* REITs (yield assets)

* Sale & leaseback

👉 In essence:

📌 Asset-based investment — NOT a tech startup

3. Why people mistakenly think it’s a tech business

Because it serves:

* Cloud computing

* AI

* Big data

But look deeper:

👉 Tech companies = use data centers to make money

👉 Data centers = make money by leasing

✔️ One is “operating technology”

✔️ The other is “monetizing infrastructure”

4. The key takeaway

👉 A data center is NOT a tech business. It is:

💡 “An infrastructure leasing business wrapped in a tech narrative”

5. A practical investment perspective (important)

If you’re an investor or property player, think of it this way:

👉 Data center = “selling power + space + reliability”

Final insight

👉 Bank lending is NOT about formulas — it’s about risk.

Formulas are tools. Cash flow is the answer.