Canada’s C$50bn EV Battery Bet: Volkswagen St. Thomas, Stellantis-LG and the Northvolt Collapse

Canada’s EV battery strategy was designed to buy a place in the North American supply chain, and its first lesson is that subsidies promised are not the same as factories built. The federal government matched United States Inflation Reduction Act production credits dollar for dollar to win Volkswagen, Stellantis-LG and Honda, while Quebec backed Northvolt with equity and loans. Two years later the US credits are gone, Section 232 tariffs have cut the cross-border auto trade, and the EV demand curve that justified 60-to-90-gigawatt-hour plants has flattened. This article traces each project, counts what was paid, what was lost and what is still under construction, and asks what a smaller strategy might look like. It is part of the Canadian Company Stories hub.

Why did Canada decide to subsidise battery plants in the first place?

Because the United States had just done so, and Canada’s auto industry only exists as part of the US one. The Inflation Reduction Act of August 2022 offered battery makers a production credit of US$35 per kilowatt-hour of cells plus US$10 per kWh of modules, available only for production on US soil. For a country that assembles about 1.3 to 1.5 million vehicles a year, nearly all for export to the US, the risk was that the next generation of plants would simply skip Ontario.

The response, led by then-industry minister François-Philippe Champagne, was to match the IRA. Volkswagen’s PowerCo was the first to accept in March 2023, choosing St. Thomas, Ontario for a plant of up to 90 GWh a year with a capital cost of up to C$7 billion and federal production incentives of up to C$13.2 billion over ten years, plus about C$500 million in Ontario capital support. Stellantis and LG Energy Solution, who had announced the C$5 billion NextStar plant in Windsor in 2022, stopped construction until they received an equivalent package of up to C$15 billion in July 2023.

Northvolt followed in September 2023 with a C$7 billion plant on Montreal’s South Shore, backed by up to C$4.4 billion from Ottawa and C$2.9 billion from Quebec, including direct equity. Honda rounded out the list in April 2024 with a C$15 billion “value chain” in Ontario, comprising a retooled Alliston assembly plant, a battery plant and two materials facilities, supported by about C$2.5 billion of federal tax credits and up to C$2.5 billion from Ontario.

The logic was explained in the Kurums article on Canada’s critical minerals strategy: the country has lithium, nickel, graphite and cobalt, clean hydro power in Quebec and Ontario, and a trained auto workforce, so a battery industry would connect resources to assembly. The missing ingredient was demand certainty, and that was supplied by US policy that has since reversed.

What happened to Northvolt and what did Quebec lose?

Northvolt’s Quebec plant, branded Northvolt Six, never got past site preparation, and the Swedish parent’s bankruptcy in March 2025 turned Quebec’s equity into a total loss. The project had been announced with Justin Trudeau and François Legault in September 2023 as a 3,000-job, 60 GWh gigafactory on a 171-hectare site straddling Saint-Basile-le-Grand and McMasterville, about 25 kilometres east of Montreal, with first batteries due in 2026.

The warning signs arrived quickly. Swedish media reported heavy losses and low output at the Skellefteå flagship in late 2023; in 2024 BMW cancelled a roughly US$2 billion order over quality, Northvolt laid off 1,600 people, and on 21 November 2024 it filed for Chapter 11 protection in the United States with about US$30 million of cash against US$5.8 billion of debt. CEO Peter Carlsson resigned the next day. The Swedish bankruptcy filing on 12 March 2025 followed when no rescue investor appeared.

Quebec’s exposure was C$510 million: C$270 million invested directly in the Swedish parent, which became worthless, and a C$240 million guaranteed loan used to buy the land. Economy minister Christine Fréchette set a 1 September 2025 deadline for a viable plan, then withdrew the province’s remaining commitments. On 5 September 2025 Justice Janet Michelin of the Quebec Superior Court declared Northvolt North America insolvent, allowing Quebec to recover nearly C$200 million from frozen bank accounts toward a C$260 million claim including interest, and the subsidiary laid off its entire workforce. The Caisse de dépôt et placement du Québec had separately invested C$200 million; the Canada Pension Plan Investment Board and the Investment Management Corporation of Ontario also disclosed write-downs on Northvolt, reportedly of about C$400 million for IMCO.

Lyten, a California lithium-sulphur start-up, bought Northvolt’s Swedish and German assets in August 2025 and said it hoped to take over the Quebec site, but Quebec declined to put in the extra money Lyten wanted. In December 2025 The Logic reported the province was marketing the serviced land for C$315 million, which would more than cover the land loan. The net cost to Quebec taxpayers, excluding the pension funds, therefore looks like the C$270 million equity plus interest and transaction costs, a serious loss but far smaller than the C$2.9 billion headline commitment.

Why did Honda suspend its C$15 billion Ontario project?

Because Honda’s global EV strategy collapsed before the Ontario plan had left the drawing board. The project announced in April 2024 was the largest single auto investment in Canadian history: an Alliston assembly plant retooled to build up to 240,000 EVs a year from 2028, a 36 GWh battery plant next door, a cathode plant with POSCO Future M and a separator plant with Asahi Kasei. It promised 1,000 new jobs and protection for the 4,200 already at Alliston.

In May 2025 Honda paused the project for about two years, citing slowing EV demand and US tariff uncertainty, and forecast a 59 per cent fall in profit. On 14 May 2026 CEO Toshihiro Mibe went further at a Tokyo briefing, announcing an indefinite suspension and a three-year restructuring that shifts resources to hybrids. Honda had just reported its first full-year loss, about US$2.7 billion, driven largely by EV write-downs after the US removed consumer tax credits and rolled back emissions rules.

Crucially for Canadian taxpayers, no public money had been paid. Ontario’s package of up to C$2.5 billion and the federal tax credits were tied to construction and production milestones that were never reached. Premier Doug Ford, who had promised in 2025 to hold Honda accountable, could do little beyond public pressure; Prime Minister Mark Carney called the decision disappointing and pointed to a five-point auto strategy that included a new EV rebate and the replacement of the 2035 zero-emission sales mandate with emissions standards. Finance minister Champagne, the same minister who negotiated the original deals, said the world was in a hold on the EV transition.

The Alliston plant continues to build the Civic and CR-V, and Honda says current employment is unaffected. But the episode shows the limit of performance-based subsidies: they protect the treasury, not the jobs that were promised.

What is the status of Volkswagen’s St. Thomas plant?

It is under construction, it is late, and it is now aimed at a market that may not exist on the original schedule. PowerCo broke ground on the 370-acre St. Thomas site in December 2023, and through early 2026 local officials reported steel framing and foundations rising. In April 2026, when Volkswagen ended ID.4 production at its Tennessee plant, PowerCo insisted there was no change to timeline, budget or scope and called the plant a long-term commitment to domestic jobs.

Five months later, on 24 September 2026, PowerCo Canada announced that cell production would begin in 2029 instead of 2027. Chief procurement officer Joel Karlsberg framed it as getting the pacing right rather than stepping back, citing next-generation battery technology and the need for flexibility as EV demand shifts. Construction continues under general contractor EllisDon, with about 60 workers on site and a peak of roughly 1,300 expected. Trade press linked the delay to Volkswagen’s search for an outside investor in PowerCo and to a possible Gotion stake in its Spanish plant.

The economics for Ottawa are straightforward: the up-to-C$13.2 billion in production incentives is paid per kilowatt-hour produced, so a two-year delay simply defers the cost. The economics for St. Thomas are harder, because the town and province built roads, rail and power connections against a 2027 start and 3,000 permanent jobs. The nearest firm customer is Volkswagen’s Scout brand, whose South Carolina plant is slated to begin production in 2028.

What did Stellantis do in Brampton and Windsor?

Stellantis kept the Windsor assembly plant and the NextStar battery joint venture running while effectively abandoning the Brampton plant that Ottawa and Ontario had helped fund. Brampton stopped building the Chrysler 300 and Dodge Charger in 2023 and was to be retooled with a US$970 million investment to produce the next-generation Jeep Compass in both combustion and electric versions from late 2025, with Unifor’s 2023 contract built around that promise.

On 15 October 2025 new Stellantis CEO Antonio Filosa announced a US$13 billion investment programme in US plants over four years, including the reopening of Belvidere, Illinois, which is slated to build the Compass. Canadian reporting at the time, and Prime Minister Carney’s public statement that the Compass was scheduled to be built in Canada and that Stellantis was expected to honour its commitments to Brampton workers, confirmed that the Compass programme had left Ontario. More than 3,000 Brampton workers, most already on layoff since 2023, were left without a product. Stellantis subsequently told a House of Commons committee and wrote to the prime minister opposing Canada’s EV availability standard, and Ottawa’s repeal of the 2035 mandate in 2026 was widely read as a response.

Windsor is the counterweight. The assembly plant builds the Chrysler Pacifica minivan and the Dodge Charger Daytona on two shifts with more than 4,200 workers, and the NextStar plant next door, a Stellantis-LG Energy Solution joint venture, began module production in 2024 with cell production following. It is the only one of the four flagship projects actually making batteries in Canada in 2026, and the one most exposed to Section 232 duties, because its cells are destined for vehicles assembled on both sides of the border.

How did US tariffs change the maths?

They removed the two assumptions on which every Canadian battery plant was financed: tariff-free access to US assembly plants and IRA-driven EV demand. The 25 per cent Section 232 tariff on imported vehicles from 3 April 2025 and on parts from 3 May 2025 exempts USMCA-compliant parts, so a battery made in Windsor can still cross the border duty-free if it meets the rules of origin. But the vehicle it goes into, if assembled in Canada, pays the tariff on its non-US content, which makes Canadian assembly less attractive and therefore reduces the local demand for Canadian cells.

The IRA side collapsed faster. The 2025 US budget law removed consumer EV tax credits, and the administration rolled back fuel-economy rules, which is why Honda, Volkswagen and Stellantis all cut their North American EV programmes within twelve months. US EV sales growth stalled, and carmakers that had sized Canadian plants for a 2030 market of several million EVs a year found themselves with contracted battery capacity and no vehicles to put it in.

Canada’s own retaliation was calibrated to protect the plants. The 25 per cent counter-tariff on US-built vehicles from April 2025 came with a remission framework that waived the duty for automakers that kept building in Canada, in proportion to their Canadian production and investment. That carrot was meant to hold Honda, Stellantis, GM and Ford in place; it did not stop Stellantis moving the Compass or GM ending BrightDrop van production at CAMI in Ingersoll, as discussed in the Kurums piece on Canada’s auto industry under the 2025 tariffs.

Which parts of the strategy are actually working?

The upstream and midstream parts in Quebec, and the module plant in Windsor. Bécancour, on the St. Lawrence between Montreal and Quebec City, has become the real cluster: Ford and partners’ cathode plant, GM and POSCO Future M’s Ultium CAM cathode facility, and precursor and recycling projects together employed close to 3,000 construction workers in 2025 according to the Quebec government, and they serve US as well as Canadian battery plants. Cathode material is a USMCA-compliant part, so it crosses the border without tariff.

Ontario’s parts base has also held up. Magna and Linamar reported record or near-record earnings in 2025 and 2026 because USMCA-compliant parts were exempt from Section 232, and both companies won new North American contracts from carmakers seeking supply certainty; the Kurums profile of Magna International covers the detail. Battery module assembly at NextStar fits the same pattern: it is parts, not vehicles, and it is already producing.

The hybrid pivot is the other quiet winner. Honda’s Alliston plant builds hybrid Civics and CR-Vs; Toyota’s Cambridge and Woodstock plants, which received no battery subsidy, run near capacity on hybrid RAV4s and Lexus models and were Canada’s largest vehicle producers in 2025. The demand that was supposed to flow to 90 GWh gigafactories is flowing to 1.5 kWh hybrid packs instead, and those are sourced largely from Japan and the United States.

What are the risks for the remaining projects?

The largest is that the United States uses its leverage over the CUSMA review to demand that battery and EV content rules favour US production. In July 2026 Washington declined to renew the agreement for a further 16 years, placing it on annual reviews; any tightening of the “melted and poured” or battery-origin rules would strand Canadian capacity. The second risk is technological: PowerCo’s stated reason for delay was next-generation cell chemistry, and a plant that opens in 2029 with 2023 technology would be uncompetitive against Chinese LFP cells that already cost well under US$60 per kWh.

What can founders and investors learn from Canada’s battery bet?

The first lesson is that matching a rival’s subsidy locks you into the rival’s policy cycle. Canada’s incentives were designed to mirror the IRA, so when Washington reversed course in 2025, Canadian projects lost their demand case and their competitive rationale at the same time. A strategy built on domestic demand, such as Quebec’s hydro-powered cathode cluster, has proved more durable than one built on exports to a single market.

The second is about counterparty quality. Volkswagen, Stellantis-LG and Honda are investment-grade groups whose delays cost Canada time but not money; Northvolt was a venture-backed start-up with US$5.8 billion of debt, and Quebec’s decision to buy its parent-company equity was a venture investment made with a public balance sheet. The Maple Eight pension funds that also lost money on Northvolt at least had diversified portfolios.

The third is that performance-based subsidies work as designed. Critics who cite C$50 billion as money spent are wrong; most of it was never disbursed. But founders and local governments should understand what that means in practice: the treasury is protected, and the communities that invested in roads, housing and training on the strength of the announcement are not. For a comparison with how a smaller economy handled the same trade-off, see the Australian Company Stories hub.

Frequently Asked Questions