The Hardware Ownership Paradox: When Sovereignty Gets Expensive

In my previous post about self-hosting for everyone I painted a hopeful picture: managed self-hosting where anyone can own their data. I glossed over something that bugs me. Hardware costs money, and the maths works out very differently depending on whether you’re a business or a person.

So let me lay out two rules that both sound completely reasonable. Then watch them collide.

Rule One: Own Your Hardware

This is the whole point of sovereignty. You buy the box, you run the box, nobody can pull your data out from under you. As I argued in Sovereign Infrastructure, I want to understand and control what I run. Renting that from someone else reintroduces exactly the dependency I’m trying to escape.

Nobody serious about owning their infrastructure is going to argue with this. Hardware ownership is non-negotiable.

Rule Two: Don’t Set Money On Fire

The second rule is just as sensible. Pay the lowest defensible price for the same hardware. Nobody chooses to spend more for an identical box out of principle.

Here’s where it stings. When a business buys a €1,000 server, here’s what actually happens:

A business effectively pays 40% less for the same hardware. The tax code rewards the entity, not the person. And it gets worse when you run the full numbers.

Total Cost of Ownership: A Real Example

Let’s price out a 5-year TCO for a modest self-hosting cluster. Three mini PCs (Lenovo ThinkCentre Tiny, HP EliteDesk Mini, that class of machine), which is the minimum for a proper high-availability setup.

Hardware Costs

Running Costs (5 Years)

Total 5-Year TCO

For an individual, self-hosting lands around €34/month over 5 years. That’s genuinely competitive with cloud services, right up until you add your own time to the bill.

The Time Problem

This is where the economics quietly fall apart for most people.

Value your time at €30/hour (conservative for a professional) and spend just 2 hours a month on maintenance:

Doing it all yourself costs nearly double what a managed service would. And that’s the rosy version, assuming a calm 2 hours a month. One serious incident eats that allowance in an afternoon. Following Rule Two, the answer is obvious: let someone else manage it, at scale, with the tax advantages and the cheaper hardware.

The Collision

Now put the two rules in the same room.

Rule Two says: get a business to buy and run the hardware, because they pay 40% less and amortise the time across many customers. Rule One says: own your hardware, because renting it back recreates the exact dependency you were trying to escape.

Walk the logic. To honour Rule Two I let a managed provider buy the box. But the moment that provider owns the box my data runs on, I’ve reinvented the cloud provider model with extra steps. My sovereignty now depends on their goodwill, their solvency, their terms of service. To honour Rule One I buy the box myself, pay the full consumer price, eat the 40% premium, and carry every hour of maintenance alone.

I need the business to make ownership affordable. But the business owning it is the thing that destroys the ownership.

That’s the hardware ownership paradox: the most economically rational choice undermines the entire reason for self-hosting. Each rule is correct on its own. Together they form a trap.

Living With It

I’ve spent more evenings than I’d like to admit drawing this on the back of an envelope. Here are the workarounds I keep circling back to. None of them are clean. All of them are people trying to thread a needle the tax code wasn’t designed to leave open.

1. The Cooperative Model

What if the customers collectively owned the management entity?

flowchart TD
    subgraph coop["Customer Cooperative"]
        note["Customers are shareholders"]
        buys["Buys HW wholesale"]
        provides["Provides mgmt"]
    end
    coop --> M1["Member Cluster"]
    coop --> M2["Member Cluster"]

How it works:

• Cooperative buys hardware at business rates
• Passes the savings through to members
• Members own their cluster hardware through the coop
• Management is a service, not a dependency

Economics:

• Hardware costs drop ~40%
• Bulk purchasing adds another 10-15% on top
• TCO could land around €22/month

Challenges:

• Legal complexity (varies by country)
• Governance overhead
• Minimum viable membership (~20-50 people?)

The clever bit: the business buying the hardware is you and the other members. Rule Two gets satisfied without breaking Rule One, because the entity that owns the box is collectively owned by the people whose data runs on it.

2. The Hardware Loan Model

The service provider buys the hardware, but ownership transfers to the customer over time.

Total paid: €1,740 over 5 years Result: Customer owns the hardware outright

Why it works:

• Provider gets the tax benefits early
• Customer builds equity over time
• After year 5 the cost drops to running costs plus management
• Real ownership at the end, just delayed

This one accepts a temporary violation of Rule One in exchange for satisfying it permanently. You rent for a while so that you can own forever.

3. The BYOH (Bring Your Own Hardware) Model

The simplest escape: customers buy their own hardware, the provider only handles management.

Provider responsibilities:

• Initial setup and configuration
• Ongoing monitoring and maintenance
• Security updates
• Disaster recovery

Customer responsibilities:

• Hardware purchase
• Physical location (home, office)
• Electricity costs

Pricing example:

• One-time setup: €100
• Monthly management: €15
• Customer buys €890 of hardware themselves

5-year TCO: €890 + €100 + (€15 × 60) + €1,140 = €3,030 = €50/month

More expensive than business-owned, because you swallow the full consumer price. But sovereignty holds from day one. This is Rule One winning outright and paying Rule Two’s tax as the price of admission.

4. The Refurbished Marketplace

This one I keep coming back to. A cooperative or non-profit that:

• Sources enterprise hardware at end-of-lease
• Refurbishes and tests it
• Sells to members at cost plus a small margin
• Reinvests the surplus into the community

Enterprise hardware sells for a fraction of its original price. A 3-year-old Dell OptiPlex that cost €800 new might go for €80 at auction. The cooperative absorbs the painful parts (sourcing, testing, warranty) and individual members get sovereign hardware at close to business prices. It chips away at the 40% gap from a completely different direction: cheaper hardware instead of cleverer ownership.

Breaking the Loop (Partly)

For my service cluster model I’m starting with a hybrid, because no single option breaks the paradox cleanly:

Phase 1: BYOH for early adopters

• Customers buy their own hardware
• I provide management at €15-20/month
• Real sovereignty from day one

Phase 2: Hardware loan option

• For people who can’t or won’t buy hardware upfront
• Ownership transfers over 3-5 years
• Higher monthly cost, no upfront barrier

Phase 3: Cooperative structure (eventually)

• If this grows beyond friends and family
• A formal cooperative for bulk hardware purchasing
• Democratic governance of the service

Every phase is a third rule bolted on to make the first two coexist. Less elegant than either, but the circle stops being a circle.

The Uncomfortable Truth

I won’t pretend there’s a perfect answer. The tax code favours businesses, and individual hardware ownership will always cost more than business ownership. Full stop.

Here’s what I’ve landed on:

Sovereignty has value that never shows up in a TCO spreadsheet. Peace of mind, principle, independence. They don’t have a line item, but they’re real.

The gap is smaller than it first looks. €34/month for genuine ownership against €25-30/month for rented sovereignty won’t be a dealbreaker for a lot of people.

Time is the variable that actually moves the needle. Kill the time cost through managed services while keeping the hardware in your name, and the numbers shift hard in your favour.

And delayed ownership still beats no ownership. A hardware loan that ends with you holding the title is a fundamentally different thing from renting forever.

The paradox is real. It’s also not unsolvable. It just asks for a bit of creativity, and for accepting that perfect sovereignty costs a little more than comfortable dependency.

I’m fine with that trade-off. Are you?