Investing in public markets
This page has been a living document since 2022. Drag the timeline to watch it — and my thinking — change over time.
The process I run today: a universe of ~500 Canadian microcaps, signals read off every release, and a rent-vs-own book where each name needs a falsifiable thesis or a dated catalyst — sized by conviction, and sold only on thesis, the sizing gate, or a better idea.
Sourcing
My universe is the 500 Canadian companies under $3B market cap with revenue and positive gross margins. I read every press release from these companies and look for signals (e.g., new management, acquisitions, NCIB announcements, spin-offs, insider buying, strategic reviews).
Watchlist
I'm looking for exciting business models at attractive valuations with easy to define catalysts not yet understood by the market. I look at trends in price, shares, margins, income statements, EV-to-sales-and-EBITDA. I read the last 5-10+ news releases and search for a pattern where news flow is appreciating faster than the share price.
Research
I download and read the latest interim and annual reports, shareholder letters, inflection documents (CEO change, acquisition, etc.), IPO or PP prospectus. I want to understand the management, business model, risks, industry, financials, valuation, and catalysts and answer:
- How strong and enduring is the earning power? Can the company predictably grow its revenues 20%+ per year? Is the company's TAM, team, and margins expanding?
- How is the management team incentivized and how good is their track record?
- What's the market's view vs my view? Can I be 5x right or only 25% wrong (asymmetry)?
Portfolio
| Quadrant | What it is | Sell rule | Sizing | Clock |
|---|---|---|---|---|
| OWN + edge | Your real alpha — compounders you found early | Hold hard; trim only on the sizing gate, never on "it ran" | Full 6–8 / 10% / 40% rules apply | Relaxed while falsifier holds |
| RENT + edge | Your highest-IRR hunting ground — special-sits nobody covers | Leave when the catalyst closes or the clock breaks | Sized to the catalyst, asset floor required | Ruthless (6-mo dead-money) |
| OWN + passenger | Beta-with-a-tilt — riding a jockey (CSU/FFH) | Benchmark vs an index, not your microcap return | Separate, lower cap — no edge, no concentration license | Never relaxes — a passenger going nowhere is dead beta |
| RENT + passenger | Doesn't exist for you — widely-covered arb | Don't play here | — | — |
Buying
I only initiate if there's a falsifiable thesis — or, for a special situation, an asset floor below the current price plus a dated catalyst. I size by conviction: a 3-5% starter while the thesis is unproven, building to 8-12% as the business executes, and anchoring higher only once it's proven — never more than 20% at cost, since appreciation beyond that I let ride. Before committing capital I confirm it's the best marginal use of it: it has to outrank my other ideas and cash on expected return and probability, offer comparable upside on a comparable timeline, and warrant ≥15% sizing to deserve fresh money. I buy in the most tax-efficient account, and in thin names I scale in over days with limit ladders.
Selling
I sell or trim for three reasons, and never just because a stock has run. First, thesis or catalyst: if the thesis breaks, or a rent's catalyst has played out, I leave — and if six months pass with no thesis progress on a name that should be working, I re-evaluate hard. The clock is harsh on rents and relaxed on owns while the falsifier holds. Second, the sizing gate: if a position grows past my survivability limit — a single illiquid name beyond ~40% — I trim the excess back toward the ceiling and keep the core, which is right-sizing, not selling the compounder. Third, a better idea: if a higher-ranked idea clears my four-condition test, I trim or sell to fund it, and the proceeds go to that idea or to cash, never back into the name I trimmed. When I execute, I go tax-first: harvest-loss lots before gain lots, registered before taxable, mind the 30-day superficial-loss rule, and scale out over days in thin names.