Nshipyard Canada · Open data project

The Power Trade Margin Board

Canada sells electricity to the United States at one price. American buyers pay another. This board joins the Canada Energy Regulator's export prices with US retail electricity prices state by state, 2010 to 2025, and shows the margin on six real export corridors. Nobody publishes the spread. It is computed here from open data, corridor by corridor, year by year.

Explore the corridors Read the methodology

$59.6B

of cumulative margin captured across six corridors from 2010 to 2025: the gap between what Canada received per MWh and what the receiving US market paid

$102.60

per MWh margin on the East to New England corridor in 2025: Canada exported at $53.24 while New England retail averaged $155.85

2018

the year western export prices moved above eastern prices and stayed there, peaking at +$51.54 per MWh in 2023. For most of 1990 to 2016 the East was more expensive

$125.40

versus $58.87: the CER's published 2024 West and East export prices, reproduced from the source file to the cent as a check on the whole pipeline

Explorer

Every corridor, every year.

Pick a corridor to draw its margin spread.

Showing 0 of 0

Year	Corridor	Export price (USD/MWh)	US market price (USD/MWh)	Margin (USD/MWh)	Volume (TWh)	Margin value (USD M)

No rows match.

Showcase

Questions the raw files cannot answer.

The CER publishes what Canada received. The EIA publishes what Americans paid. Nobody publishes the difference. These cards compute it, with the source and the method printed next to every number.

How much margin does each power intertie capture?

New England captures the most. From 2010 to 2025 the East to New England corridor captured $20.44B on 260.1 TWh of exports. In 2025 alone the margin was $102.60 per MWh: Canada exported at $53.24 while New England retail averaged $155.85. The PJM corridor is the smallest at $0.41B cumulative.

When did western power get more expensive than eastern?

In 2018, and it stayed that way. Western export prices moved above eastern prices that year and the gap peaked at +$51.54 per MWh in 2023. For most of 1990 to 2016 the East was more expensive. The partial 2026 data (through July) shows the gap narrowing again, which is worth watching, not concluding.

What did the California energy crisis do to export prices?

It sent western export prices to $110.78 per MWh in 2001, the highest in the 36-year series, while eastern prices sat at $41.05. The crisis shows up in Canadian export revenue as a spike, not a blip: 2000 and 2001 are the only years before 2018 when the West priced far above the East.

What does the buyer pay versus what Canada receives?

In 2025, New England buyers paid $155.85 per MWh on average at retail while Canada received $53.24 for the power it exported to them. The honest caveat: retail prices include distribution and utility costs, so the margin is not pure profit on the wire. It is the full spread between the two published prices, which is exactly what neither government publishes.

CER Electricity Trade Summary XLSX (updated 2026-09-25) and archived quarterly CSV (1990Q1-2018Q2), via open.canada.ca and cer-rec.gc.ca. US prices: EIA Form 861 annual retail sales 2010-2025. Currency: Bank of Canada annual average 2017-2025, FRED DEXCAUS 1990-2016, BoC monthly average for 2026. Retrieved 2026-10-10.

Methodology

How the margin is computed, and where it is weak.

1
Sources, all keyless and retrieved 2026-10-10: the CER Electricity Trade Summary XLSX (updated 2026-09-25, monthly West and East export prices 2010-01 to 2026-07, annual volumes 2010-2026, annual province-to-destination flows 2010-2026); the CER archived quarterly imports-exports CSV (1990Q1-2018Q2, electricity exports by province and destination state with MW.h, CAN$ and CAN$/MW.h); EIA Form 861 annual files 2010-2025 (utility-level retail revenue and sales aggregated to annual state average retail price, USD/MWh); Bank of Canada FXAUSDCAD annual average 2017-2025, FRED DEXCAUS daily averaged annually for 1990-2016, BoC monthly average for 2026.
2
The missing artifact is the corridor crosswalk, hand-curated for this project and published as data/crosswalk.csv. Canadian export origins follow the CER's own grouping: West is British Columbia, Alberta, Saskatchewan, Manitoba; East is Ontario, Quebec, Newfoundland and Labrador, New Brunswick, Nova Scotia. US destinations map to five market regions: US West, US Midwest, PJM, NYISO, ISO-NE. Each region's price is the sales-weighted average retail price of its states from EIA-861.
3
Six corridors survive a 1 TWh minimum total flow over 2010-2025: East to ISO-NE, East to NYISO, East to PJM, East to US Midwest, West to US Midwest, West to US West. Four one-off pairings (for example West to PJM at 0.01 TWh total) are dropped as scheduling noise, documented here rather than hidden.
4
Currency: every export price is converted to USD with the annual average CAD-per-USD rate, so the margin compares like with like. US prices are already USD.
5
Quarantine: 2 rows in the archived CER CSV carry a blank destination and are excluded from every aggregate. The 2026 monthly CER data runs only through July and is marked partial wherever shown.
6
Validation: the pipeline reproduces the CER's published 2024 anchors exactly: $125.40 CAD/MWh West and $58.87 CAD/MWh East. If those two numbers had not matched to the cent, nothing downstream would have shipped.
7
Limits, stated plainly. US retail prices include distribution and utility overhead, so the margin overstates the pure wholesale spread. EIA-861 is annual, so the margin series is annual even where the CER publishes monthly prices. The 2019-2025 corridor export prices use the CER's West and East regional prices (the archived province-by-destination file ends in 2018Q2). Nothing here is a forecast, and the 2026 partial-year inversion flip is a signal to watch, not a conclusion.

The corridor crosswalk

Destination in CER data	Market region	Grid operator

For developers

Query it from code, or from an agent.

Three consumption paths, same normalized data. REST for applications, OpenAPI for integration, MCP tools over streamable HTTP for AI agents.

Endpoints

GET/api/v1/margins?corridor=East%20to%20ISO-NE&year=2025

Try it

Margin rows by corridor and year: export price, US market price, margin, volume

{
  "total": 1,
  "rows": [
    {
      "year": 2025,
      "corridor": "East to ISO-NE",
      "export_price_usd_mwh": 53.24,
      "us_market_price_usd_mwh": 155.85,
      "margin_usd_mwh": 102.6,
      "export_volume_mwh": 14234567.8,
      "margin_value_usd_millions": 1460.5
    }
  ]
}

GET/api/v1/inversion

Try it

West versus East export prices, 1990-2026, the inversion series

{
  "total": 37,
  "rows": [
    { "year": 2023, "west_usd_mwh": 97.94, "east_usd_mwh": 46.39, "west_minus_east_usd": 51.54 }
  ]
}

GET/api/v1/corridors

Try it

The six export corridors with origin provinces, market states, and grid operators

{
  "total": 6,
  "rows": [
    { "corridor": "East to ISO-NE", "origin_provinces_en": "Ontario, Quebec, New Brunswick", "iso": "ISO New England" }
  ]
}

GET/api/v1/crosswalk

Try it

The hand-curated destination-to-market-region crosswalk

{
  "total": 55,
  "rows": [
    { "destination_in_cer_data": "Maine", "market_region": "ISO-NE", "iso_rto": "ISO New England" }
  ]
}

GET/api/v1/summary

Try it

Headline aggregates: cumulative margin, 2025 ranking, inversion crossover and peak

{
  "cumulative_margin_usd_billions": 59.6,
  "top_corridor_2025": { "corridor": "East to ISO-NE", "margin_usd_mwh": 102.6 },
  "inversion_peak": { "year": 2023, "west_minus_east_usd": 51.54 }
}

OpenAPI spec →

Connect your agent

Put this data to work inside your AI tools.

Pick your harness, copy the prompt, send it to your agent. Your agent runs the setup itself.

Copy and send this to Claude Code

Set up the The Power Trade Margin Board MCP server so I can query it from here.
1. Run: claude mcp add --transport http canada-power-trade https://this-site.example/mcp
2. Run `claude mcp list` to confirm it connected.
3. what was the margin on the East to ISO-NE corridor in 2025, and show me the result.

Data