Playing a Different Game (Investor Series): “Configurable Risk” in Action — The Energy + Data Example (Power → Bytes)
If the shipping example was about corridors, this one is about something even more basic: power. Because in the next decade, “energy” isn’t just energy. It’s energy + data — electricity turned into compute, storage, inference, and revenue. Power becomes a monetizable input to the digital economy.
And that makes it a perfect second demonstration of the configurable idea. Not because it’s “safe.” Because the knobs are visible.
The setup: a small power-to-bytes project
Imagine a project with three physical elements:
- A renewable generation asset (hydro / solar / wind depending on the site)
- A battery or flexible load layer (optional but powerful)
- A small data center / containerized compute (or a long-term contract selling power to someone else’s compute)
You can do this at multiple scales. The point here isn’t the size — it’s the structure.
Because the deal isn’t one blob called “an energy project.” It’s a set of separable components with separable risks.
The core idea: you don’t “invest in a country.” You place the stack.
In power + data, you’re placing:
- the asset (generation + site)
- the offtake (who buys the power / compute)
- the ownership (where equity and governance live)
- the cashflow path (where money lands and sits)
- the contracts and dispute resolution (what happens when things go wrong)
The knobs: how you configure exposure
Put the physical asset where the economics are real
AssetCo / OpCo lives where the plant is — often in a corridor / frontier market where:
- power is cheap (or can be produced cheaply)
- there’s unmet demand or constrained grid capacity
- land and build costs are reasonable
- the “peace premium” is investable (continuity matters more than headlines)
That’s where you earn the real edge: physical economics. But you do not have to let that location dictate your entire risk profile.
Put ownership + governance where rule-of-law is strongest
Your HoldCo (the entity that owns the asset company) does not need to be in the same place as the plant. You choose a jurisdiction that gives you:
- clear shareholder rights
- predictable enforcement
- strong corporate governance tools
- workable dispute mechanisms
Translation: you want your ownership to live where rules are hard to bend.
Anchor key contracts in neutral law + neutral dispute venues
Energy projects don’t fail because the sun stops shining. They fail because:
- counterparties change their mind
- pricing becomes politically sensitive
- payments get delayed
- promises get “reinterpreted”
So you configure:
- Governing law (English law or similar)
- Dispute resolution (arbitration in a neutral seat)
You’re not trying to win arguments in advance. You’re trying to avoid being forced into the worst possible forum when the relationship breaks.
Split the offtake: avoid single-counterparty dependence
This is one of the cleanest “optionalities” in the model. Your revenue does not have to come from one buyer.
You can structure multiple offtake paths:
- Grid sales (baseline)
- Private offtake (PPA) to an industrial buyer (stability)
- Power-to-bytes (compute buyer or your own compute) for upside
That’s corridor diversification, but expressed as revenue channel diversification. If one channel gets politically constrained, another can carry the project.
Decide where the cash lands (and in what currency)
This is where “configurable” becomes more than a slogan. You can structure so that:
- local OpCo holds enough to run the plant
- upstream distributions flow to a HoldCo-controlled account in a strong banking jurisdiction
- revenues can be denominated or hedged in a basket (EUR/USD + local currency exposure kept bounded)
You’re reducing “trap risk” — the risk that your liquidity gets stuck in the wrong place at the wrong time.
Modularize the tech risk
Data infrastructure is a fast-moving world. So don’t weld your entire project’s fate to one compute bet.
You configure modularity:
- containerized data center
- swappable hardware cycles
- third-party compute buyer option
- ability to pivot between “sell power” and “sell compute”
The project becomes a flexible platform, not a single thesis.
Why this model is a great “configurable” demonstration
Because the risks are different — and separable:
- country / permitting / grid risk (local)
- contract / counterparty risk (configurable via law + venue + structure)
- currency + banking risk (configurable via cashflow architecture)
- technology risk (configurable via modularity and optionality)
- demand risk (configurable via multiple offtake paths)
You’re not eliminating risk. You’re moving from one big correlated risk pile to a designed portfolio of risks. That’s the whole game.
The punchline
I’m not saying: “Frontier energy + compute is safe.” I’m saying: you can structure it so you’re not hostage to any single failure mode.
The investor question isn’t “Is this risky?” It’s: Which risks are designable — and did we actually design them?
And here’s the litmus test: If your deal can’t answer, clearly and in one page:
- where the cash lands
- where disputes go
- who controls governance under stress
- what happens when the main offtaker fails
…then it isn’t configured. It’s just exposed.
