AEM 100· ARX 96· CNQ 96· FTS 96· TOU 96· CNR 91· KMI 91· NTR 91· PPL 91· WFG 88· ENB 86· CVE 84· TRP 82· CCO 80· SU 76· FCX 71· OXY 70· TECK 70· MEG 68· AG 36· AEM 100· ARX 96· CNQ 96· FTS 96· TOU 96· CNR 91· KMI 91· NTR 91· PPL 91· WFG 88· ENB 86· CVE 84· TRP 82· CCO 80· SU 76· FCX 71· OXY 70· TECK 70· MEG 68· AG 36·

Reading 20 operators · Q1 2026 letter live · Apr 2026

Machine breadth. Human conviction.

Halvren is an AI-augmented research desk covering Canadian and U.S. operators in energy, materials, and infrastructure. Models read at scale. The principal signs the decision. The book is proprietary. The work is public.

At the desk Reading the desk's queue

New here? Start here →

Canadian & U.S. markets
AI-augmented · human-reviewed
Proprietary capital
  • Energy
  • Materials
  • Infrastructure
20

Operators in deep coverage

11

Long-form notes published, growing

17.1%

Annualized return since 2019, net of fees · every year positive

142

FY 2025 filings read this quarter

Our Philosophy

Three things we actually believe

Not aspirations. Not values off a slide deck. The convictions that determine what we read, what we write, and what we own.

01

The price of admission to real returns is the discomfort of looking wrong for years.

The consensus is already priced. Edge lives in the horizons others have abandoned, the industries that have embarrassed themselves, and the operators too boring to make a conference panel. Canada has a lot of those. We are comfortable being boring.

02

We don't predict cycles. We read the operators who survive them.

Macro forecasting is a prestige game. Operator quality is arithmetic. We spend our time on the second one: cost curves, reinvestment history, insider behaviour, the balance sheet through a 2015 or a 2020. That evidence tells you what a business is without asking the market what it thinks.

03

Risk is not volatility. Risk is the permanent impairment of capital, and it never announces itself.

A stock that falls 40% and recovers is not a loss. A business whose competitive position quietly erodes over four years while the chart trades sideways? That is a loss. We spend more time on the latter, because it never shows up in a risk model until it has already happened.

What we cover

Three sectors. Canadian and U.S. operators. Long horizons.

Energy

The businesses the market keeps giving up on, usually twice per cycle.

Canada runs on heavy oil, natural gas, and uranium. The operators worth owning have learned to live in the middle of the cost curve and still make money at the bottom of it. We look for low all-in sustaining costs, clean balance sheets, reserves that don't need a story to work, and boards that buy back shares when the stock is hated instead of issuing when it's loved.

Most of what gets written about this sector is a price deck in prose. We try to write about the business underneath: the one that was already profitable before the commodity moved, and will still be after.

Oil & gas · Uranium · Power · Midstream

Materials

Hard things with hard numbers.

Copper, silver, potash, lumber, specialty chemicals. The cycle is loud; the business underneath is quieter. We read cost curves, reinvestment history, and insider behaviour more than price forecasts. Most mining stories collapse on the drill results. The ones worth owning were already profitable before the story started.

Precious metals · Industrial metals · Fertilizers · Forest products

Infrastructure

The regulated cash machines, read honestly.

Pipelines, utilities, railways, ports. Boring on a chart, spectacular on a twenty-year dividend reinvestment. We care about regulatory jurisdiction, the quality of the asset underneath, and rate-base growth more than the payout ratio. Not every yield is a business. Some are a slow-motion return of capital.

Pipelines · Utilities · Rail · Ports & terminals

“

We don't trade sectors. We read operators. Price decks, macro calls, central-bank body language: noise that happens to other people.

— Halvren Capital, Research Notes

The digest is live·Updated this week

At the desk Reading the desk's queue…

What the desk is reading this week.

An AI-augmented read of every Canadian and U.S. operator earnings call on the Halvren coverage list. 142 filings, 4,820 pages, 7.4 hours of call audio processed this week — eleven model flags, eighty checklist questions evaluated, two names promoted to principal review.

Read the digest
Filings ingested142
Pages read4,820
Model flags11
Promoted to desk2

On the desk

Canadian names we're reading right now.

No price targets. No conviction levels. No pretending we know what the stock does next month. Just what we're actually working through under the hood: business health, who's at the helm, where we sit in the cycle.

Last reviewed May 2026 · FY 2025 filings (Feb–Mar 2026)

80 / 100
Halvren Read
CCO Cameco Corp
TSX / NYSE:CCJ

Uranium

Uranium has become a narrative asset. Cameco is still a mine with a cost structure. We're re-reading McArthur River restart economics, the Westinghouse integration, and what happens to contract-price realizations as the utilities' term book actually rolls. The bull case needs the commodity. The business is fine without it. That is exactly why we keep coming back to it.

What we trackAll-in mining cost · Westinghouse contribution · Long-term contract book · Capital discipline vs. peers

Read full research →

96 / 100
Halvren Read
CNQ Canadian Natural
TSX / NYSE

Oil & Gas

Murray Edwards built a business that doesn't need $100 oil to matter. Low-decline, long-life assets, a dividend that survived 2015 and 2020, and a management team that allocates capital like owners, because many of them are. The questions we keep returning to: how many more cycles does this same playbook have in it, and what does succession actually look like when it arrives.

What we trackSustaining capex · Free cash flow at strip · Buyback pace · Insider alignment · Succession

Read full research →

36 / 100
Halvren Read
AG First Majestic Silver
NYSE / TSX

Silver

The hardest of the three to own honestly. Silver is a narrative metal, and First Majestic has had more operational resets than closes at a new high. We're working through the real cost curve, the Mexico tax and royalty overhang, and whether management's capital decisions have generated per-share value, or just production growth. Sometimes the most useful output of research is “not yet.”

What we trackAll-in sustaining cost · Production-per-share · Mexico tax posture · Reserve quality · Hedge book

Read full research →

86 / 100
Halvren Read
ENB Enbridge
TSX / NYSE

Infrastructure

A pipeline is not an oil bet. Owning Enbridge is closer to owning a turnpike with a thirty-year toll structure. The corridor between the WCSB and the U.S. Midwest and Gulf Coast does not get rebuilt every cycle, and roughly 98% of group EBITDA is take-or-pay or cost-of-service. Thirty-one years of dividend raises is the receipt for that arithmetic. The honest question is whether management keeps treating it like the toll-road it is.

What we trackMainline apportionment · Dominion utility rate cases · U.S. gas backlog conversion · Net debt trajectory · Transition capital discipline

Read full research →

See the full coverage universe

Figures compiled from Cameco, Canadian Natural Resources, First Majestic, and Enbridge FY 2025 disclosure (released February–March 2026). Reviewed quarterly; next refresh expected after Q1 2026 earnings. None of the above is a recommendation, a solicitation, or a price call. The author may hold positions in any of these names and may transact in them at any time without notice. Full writeups live on Substack. See the footer for the full disclaimer.

The lens

The Halvren Checklist.

Ten questions we run through before a Canadian or U.S. operator earns the writeup. Nothing proprietary. Mostly the questions sell-side decks skip, because the honest answers are inconvenient. Screenshot it. Use it. Argue with it.

Pillar I

The business

01

Does it generate free cash flow through the full cycle, or only the top half of it?

Reported earnings lie in commodities. Ten years of FCF tells the truth.

02

Do the unit economics still work at the worst price of the last decade?

If the thesis needs a new commodity regime to work, it isn't a business yet.

03

What does the balance sheet look like at trough pricing: net debt, covenants, maturity ladder?

The next crisis won't ask what you projected. It asks what you owe and when.

04

When they reinvest a dollar (capex, M&A, or buyback), what actually comes back?

ROIC on incremental capital, not reported ROE on the whole book.

Pillar II

The people

05

How much of the operator's own net worth, bought and not granted, sits in this name?

Options are loyalty to a quarter. Open-market purchases are loyalty to a decade.

06

What did management actually do in 2015 and 2020: issue, buy back, or sit still?

Every operator on the desk has been stress-tested twice in the last decade. The record is public.

07

Is compensation tied to per-share value, or to production, revenue, and size?

Growth at the expense of the share count is a tax the operator quietly charges you.

08

Who succeeds the operator, and is that person already visible on the page?

Every owner-operator story ends. The question is whether the business does too.

Pillar III

The cycle

09

Where are we on the cost curve that matters: the real one, not the one in the pitch deck?

Every operator is first-quartile when they pick the quartile. Ask who sets the denominator.

10

What does a “normal” year look like a decade from now, and does this business still work at that price?

If the thesis only works at the top of the cycle, it's a trade. Halvren doesn't trade.

Ten questions. If a name can't survive the first four with honest answers, it never reaches the next four. Most don't. That's the point.

Read the full Checklist →

Powered by the Halvren engine

Or run it on a stock you actually own.

US tickers as-is. Canadian: TSX format with the suffix (e.g. CIX.TO). Or open the full page →

The Founding Memo

On patience, capital, and the long game.

There is a certain kind of investor who treats uncertainty as the enemy. They build models to defeat it, buy protection against it, and measure their competence by how seldom it appears in their portfolio. We are not that investor.

Uncertainty is the job. The question is never whether uncertainty exists (it always does), but whether the price you paid reflects it honestly. Our edge is not superior information. It is superior patience. We are willing to hold through the quarters where being right feels identical to being wrong.

We also believe in being honest about what we don’t know. We don’t know which AI model wins. We don’t know when Canadian real estate corrects, or by how much. We don’t know which of today’s platform businesses will look like railroads in thirty years: essential, regulated, and only mildly profitable. What we believe is that the businesses we own will still be operating, still compounding, and still relevant when those answers become obvious.

That, we think, is enough.

Read the full founding memo →
Amirali Karimi, founder of Halvren Capital

Amirali Karimi · Vancouver, BC

Principal

One desk. One principal.

Halvren is a small, AI-augmented research firm. Models read filings, transcripts, and pricing data across the Canadian and U.S. operator universes. The principal reviews every conviction, signs every position, and writes every public letter. Machine for breadth. Human for judgment. How the desk works →

Close to a decade in the work, split between managing investments and operating businesses, real estate development in British Columbia and software in Canada and the MENA region, including a successful exit to Digikala, Iran's largest e-commerce platform, in 2023. Economics at Simon Fraser University. Securities licence through Canadian financial services. CFA candidate. Fluent in English and Farsi.

Long before any of that, my family was building manufacturing businesses and developing real estate in British Columbia. Thirty-plus years in, and still at it. The work here draws on both sides of that education: how durable businesses actually get built, and how capital eventually learns to price them.

Halvren covers energy, materials, and infrastructure operators across Canada and the United States. The public research, the Halvren Checklist, the watchlist, and the company writeups, is free and will stay that way. The position-level commentary and the quarterly Halvren letter go out through the private letters.

Most of what we study never becomes a position. Most of what becomes a position is held for a long time. The writing reflects that pace.

Read the full bio →

Halvren manages proprietary capital and publishes research. We are not a registered investment adviser and do not manage outside capital.

Contact amirali@halvrencapital.com · LinkedIn · Substack
and inside the .desk-latest-list aside; no client-side fetch needed. -->