Every project arrives with a list of things that could go wrong. The list is not the interesting part. The interesting part is who has agreed to be holding each item when it does.
Risk allocation is the quiet architecture of infrastructure. It decides which risks are priced by a contractor, which by a lender, which by a public balance sheet, and which are simply left with the sponsor because nobody else would take them. It also decides what the project costs before a single electron moves.
Risk is not a feeling. It is an exposure.
In project finance, a risk becomes manageable when someone can measure it, price it, monitor it and absorb the consequence. That is a high bar. A risk that cannot be measured cannot be priced; a risk that cannot be monitored cannot be enforced; a risk that cannot be absorbed is not really transferred, even if the contract says it is.
This is why contracts are so specific about small things. The definition of a reference index, the frequency of a meter reading, the date a payment is due, the currency it is paid in: each is an attempt to turn an uncertain future into something a lender can model and a court can interpret.
The risk ledger
The table below is an editorial map of common allocations. It is not a description of a specific transaction, and the weights are a way of reading, not a dataset. What matters is the shape: some risks stay with the project almost by nature; a few can be moved by contract; and a public counterparty is usually asked to hold the risks that private markets cannot price at all.
Editorial allocation
ProjectCapitalPublicDevelopment and permitting
Site control, environmental assessment, interconnection queue, local consent.
Project 90Capital 10Public 0Usually carried by the developer. It is rarely transferable before permits and site rights are real.
Construction and completion
Schedule, cost overrun, commissioning, performance at handover.
Project 55Capital 40Public 5Contracts can move much of this to a contractor and insurer. The sponsor still holds the residual.
Technology performance
First-of-a-kind behaviour, availability, efficiency, emissions intensity.
Project 70Capital 25Public 5Performance guarantees help, but a new technology can miss a specification nobody has tested at scale.
Price and market
Commodity, carbon or power prices moving away from the case that justified the investment.
Project 20Capital 20Public 60This is the risk a contract for difference is built to move. Without one, the project carries nearly all of it.
Policy and regulation
A carbon price, tax credit, standard or rule changing during the asset's life.
Project 15Capital 15Public 70A policy-price contract can absorb part of a change in a published price. It cannot promise that the policy survives.
Demand and volume
No buyer, a smaller buyer, or a buyer who defers delivery.
Project 30Capital 25Public 45An offtake contract provides revenue for a defined volume where private demand is still developing.
Counterparty and credit
The other side of the contract fails, disputes, or cannot post collateral.
Project 25Capital 35Public 40Every risk transfer creates a new exposure: to the party that accepted the risk.
Currency and interest rates
Debt costs or exchange rates move between commitment and construction.
Project 25Capital 65Public 10Normally a treasury problem, not an editorial one. Hedging has a price and a counterparty.
Social licence
Local opposition, consultation failure, lost trust after an incident.
Project 70Capital 10Public 20Hard to contract away. It is carried by whoever is still standing on the site in ten years.
Weights are editorial, illustrative of common structures. They are not data about any fund, project or contract.
Risk has a price
A party accepting risk wants to be paid for it. That price appears as a higher contract rate, a wider interest margin, a larger equity return requirement, or a contingency buried in a budget. It is rarely labelled "risk". It is simply the number the project has to clear.
The cheapest allocation is usually not the one that leaves the least risk with the project. It is the one where the party best able to measure and manage the risk holds it. A contractor who builds the same asset every year can price construction risk more accurately than a sponsor who builds one. A lender with a portfolio of similar loans can absorb credit risk more cheaply than a single project company.
Some risks, however, have no established price. A policy that might change after the next election, a market that may not exist at scale, a technology that has worked in a pilot but not in a Canadian winter: private capital may decline the project regardless of its expected return. This is the gap a public counterparty can fill — not by making the risk disappear, but by accepting it for a defined price, over a defined term, with a defined limit.
That is not the same as a subsidy. The Canada Growth Fund states that it does not provide grants or subsidies, and that it expects to recover its capital on a portfolio basis.1 The distinction matters: a contract can make an investment possible without making it free.
A contract does not delete risk. It names the party who will still be there when the risk arrives.
What a public balance sheet can absorb
The plan describes two contract families. An offtake contract buys a defined volume at a known price for a known term, providing revenue where private demand is still developing. A contract for difference exchanges the difference between a reference price and a strike price, without buying or selling the commodity itself. The plan describes both as tools to address demand and price risk, including policy risk.
The two-way structure is the detail that matters. If the reference price is below the strike, the provider pays the holder the difference. If the reference price is above the strike, the holder pays the provider. The public side is not only insuring the downside; it is participating in the upside, at least inside the contract's terms.
The capacity is real, and it is finite. The plan states that aggregate net exposure through contracts in a tail-risk scenario will be no more than $7 billion.1 That is a statement about the size of the promise, not the size of the fund. A public balance sheet can absorb a defined risk; it cannot absorb an open-ended one without turning a contract into a liability it did not price.
The published principles behind the strategy are also a useful test of intent: unlock projects, strengthen carbon markets, respect regional differences, prioritize efficient projects, accelerate technologies, and deliver value for money.2 Each is a claim about how the risk should be allocated, not only about what should be funded.
What a contract cannot move
A contract can move the consequence of a price. It cannot move the reason the price exists. Basis risk is the clearest example: if the contract references a national carbon index but the project sells into a regional market, the two can diverge while the contract performs exactly as written. The project is hedged on paper and exposed in practice.
Volume is another residual. A price contract does not pay for output that was never produced. A construction contract does not build the asset. A policy contract cannot promise that the policy survives a change of government. And every transfer creates a new exposure to the party that accepted it: a contract is only as strong as the balance sheet on the other side.
These are not drafting failures. They are the boundary of what a contract can do. Reading a project well means asking not only what was transferred, but what was left behind, and whether the party left holding it can survive the scenario it is now carrying.
Five questions to ask of any risk transfer
- What event triggers payment? A change in a published index is different from a change in the project's realised revenue.
- Who measures it, and can the measurement be disputed? A contract that depends on a number nobody can audit is not a hedge.
- How long does it last, and what happens at the end? A fifteen-year contract on a forty-year asset leaves a long tail.
- What does the accepting party get, and can it pay? Risk transfer without credit support is a promise about a promise.
- What is not covered, and who holds that residual? The answer is usually the sponsor, the community, or the public — sometimes all three.
The next explainer looks at the machinery underneath these questions: how a stack of contracts becomes the collateral for a project, and what a two-way contract for difference actually pays.
Sources
Method & source register- 01Corporate Plan Summary 2026-2030
Canada Growth Fund · Information current as of 23 October 2025 · Accessed 13 November 2026
- 02Our Investment Strategy
Canada Growth Fund Investment Management · Undated web page · Accessed 13 November 2026
Cited documents are linked, not reproduced. Figures are reported as published and are not live data.