The bottleneck is the asset
The Atacama and broader Pacific corridor host some of the highest solar irradiance on the planet, with capacity factors that comfortably exceed most of North America and Europe. That resource advantage is not the scarce input. Permitted megawatts are not scarce either - the pipeline of environmentally approved projects across the Chilean and Andean grid runs, in our estimate, into the tens of gigawatts. What is genuinely scarce is the physical right and the physical hardware to inject that power into a constrained transmission system at a point where it can actually be dispatched and paid for.
The result is a structural bifurcation. A project with a signed interconnection agreement, secured substation capacity and a defined injection point is a fundamentally different asset from an otherwise identical project holding only environmental permits. The market still prices them as if the gap were a matter of time and paperwork. In practice it is a matter of physics and queue position, and the two can diverge by years. That divergence is the arbitrage.
Why the spread persists
Grid interconnection queues in the region behave like any congested system with mispriced access: they fill with speculative applications that will never be built, which in turn lengthens study timelines and inflates the apparent backlog. Curtailment in the northern nodes - periods where generated power cannot be evacuated and is effectively spilled - further separates nameplate capacity from deliverable, revenue-generating capacity. An allocator who underwrites nameplate megawatts is underwriting a number that the grid may never honor.
This is why the interesting position is not building more panels. It is controlling the connection. Operators who already hold interconnection rights, land adjacency to substations, and the transmission studies that de-risk injection can acquire stranded permitted projects at paper valuations and re-rate them by supplying the one input the seller lacked. The margin is not a commodity solar margin. It is a bottleneck-clearing margin, closer to infrastructure than to generation.
Structuring the exposure
For a single-asset SPV the discipline is to underwrite the injection point before the irradiance. Diligence should center on queue position, substation headroom, curtailment history at the specific node, and the enforceability of the interconnection agreement - not on the theoretical yield of the array. A project that can prove firm evacuation capacity at a specific node deserves a materially different cost of capital than one that cannot, and the structure should isolate that risk rather than blend it.
Storage co-location changes the equation further. Pairing generation with batteries at a constrained node converts curtailed midday energy into dispatchable evening supply, which both lifts realized prices and, in some configurations, eases the evacuation constraint itself. Illustratively, moving even a modest share of output from spilled midday hours to priced evening hours can shift a project's realized capture price by a meaningful margin - enough to change the sign of the return on assets that screen as marginal on a generation-only basis.
What it means for an allocator
The mispricing will not last indefinitely. As transmission build-out and queue reform advance, the spread between permitted and shovel-ready capacity should compress. That compression is precisely the return - capital deployed today against connection-secured assets captures the re-rating as the wider market catches up to the distinction the grid already enforces. The window is defined by how long institutional capital keeps treating a permit as equivalent to a connection.
For a family office, the attraction is that this is a real-asset, hard-infrastructure position dressed as a renewables trade, with a return driver - clearing a physical bottleneck - that is largely uncorrelated with the direction of energy prices. The risk to respect is regulatory: transmission expansion timelines, queue reform, and curtailment compensation rules are all policy-dependent, and a single-asset SPV should be underwritten to survive slower reform, not to depend on it accelerating.