How this office decides.
These are the operating principles of Gord Holdings. They are not a strategy — a strategy written now will be wrong within a decade — and not a portfolio, which is an expression of a moment. They are the reasoning underneath both.
Each was paid for. Where one came from a particular episode the episode is on the record elsewhere on this site, including the ones that went badly. A principle without a scar is a slogan, and slogans do not survive a difficult year.
Never be a forced seller.
Every structural advantage this office holds reduces to one fact: nothing obliges it to sell. No fund life, no vintage year, no investor with a redemption right. That absence converts a downturn from a threat into a purchase. Debt is the mechanism that removes the choice, so it is used only where the obligation cannot force the sale of something intended to be kept.
Exit by combination.
An exit that ends the exposure ends the compounding with it. Where an exit has occurred here, the consideration has more often been a position in the larger entity than cash — a merger is a way to leave the operating seat and stay on the register. Cash is correct when the buyer is wrong about the future and this office is not.
Concentrate where there is an edge; diversify only against ruin.
Diversification is protection against ignorance, and ignorance is optional. Size the position to conviction and the downside to survival; they are separate questions and are answered separately. An edge is knowledge, access, or control. Enthusiasm is not an edge.
Own the layers everyone must use.
In any new market the applications are numerous, visible and mostly doomed, while the layers beneath them — identity, settlement, coordination, distribution — are few and dull. Ask what every participant must touch regardless of who wins. Owning that removes the need to pick the winner.
Measure in decades; report honestly in years.
A long horizon is a licence to ignore noise, not to avoid the scoreboard. Results are stated annually in writing, including the bad ones. An unmeasured long horizon becomes an excuse, and “it is a long-term position” is the most expensive sentence in private capital.
When speed and evidence conflict, evidence wins — then move faster on everything else.
Both halves are load-bearing. Evidence first, because one fabricated claim discounts every true one standing beside it. Faster afterwards, because caution applied uniformly is only slowness. Reserve the caution for what is published and committed; everything upstream should be quick to the point of discomfort.
Be early enough to be wrong in public.
Every position of consequence in this record was taken before the market had a word for it. Being that early guarantees being publicly wrong sometimes, which is the fee and is payable. Early is a position taken before the evidence is common; stubborn is holding it after the evidence has arrived and disagreed. The difference is whether you can say what would change your mind.
Price the wrapper, not the story.
Most of the distance between a good outcome and a bad one is structural: the entity, the domicile, the rights that transfer, the obligations that survive. Ask what remains if the operator disappears tomorrow. Compliance leads expansion — regulatory posture is a moat in a young market, not an overhead.
A system that pays participants in the thing it issues is a closed loop.
When the reward and its funding source are the same instrument, the arrangement must recruit perpetually or fail, and it fails. Ask where the money comes from one layer further than is comfortable, and stop only on reaching a customer. The same test applies to any business whose growth is funded by the enthusiasm it generates for itself.
Build it as a company, not a programme.
An initiative inside a larger entity, with a budget line and a rotating owner, produces effort and no revaluation. Four conditions are cumulative: separability, external revenue, disclosure, and credible management holding equity. Three of four produces a footnote. Four of four changes the multiple. Incorporate before announcing.
Sequence compounds; a list does not.
Acquiring several things is not a strategy. Acquiring them in the order where each lowers the cost or raises the value of the next one is. Choose the connecting substrate — the shared ledger, identity, and distribution — before the second acquisition. Retrofitted integration is the most expensive work there is.
Operate what you own.
This is an operator’s office. Nearly every outcome in the record came from building the business and running it closely rather than selecting it. Judgment about a business you operate is a different substance from judgment about one you read about.
Teach before you sell.
Building the education layer before a market exists creates the vocabulary, the practitioners and the trust the commercial phase later depends on. Whoever teaches a category defines it, and definition is the most durable position available. Teach it honestly, including what is wrong with it; education that is disguised selling is detected.
The building is the brand.
Physical presence, real institutions and things a person can visit are disproportionately persuasive in markets full of claims. They are also slow and expensive, which is precisely why they are credible. Prefer the proof that is inconvenient to fake.
No invented numbers, at any scale.
Every figure stated is sourced and defensible. This is a structural position before it is an ethical one: the first fabricated statistic retroactively discounts every true statement beside it, and the discount is applied to all of them at once. A rounded number is still a number. Round it honestly or omit it.
The vocabulary holds.
The words an institution uses are its position. Adopting the crowd’s vocabulary because it is easier surrenders the frame to whoever built it. One claim, one canonical home: say a thing once, properly, where it belongs, and reference it everywhere else. Repetition dilutes; reference compounds.
Publish what can be checked.
Claims that link to a live surface, a public record or a completed transaction accumulate into a reputation. Claims that link to nothing accumulate into a style. Over a long enough period the market distinguishes the two with complete accuracy. Prefer the receipt to the adjective.
Keep a written record of why, not only what.
The decision is recoverable from the accounts; the reasoning is not, and the reasoning is the part that transfers. Record the decisions declined and the reasons, which teach more than the acceptances. Record what was expected to happen, and compare it later — memory edits itself in favour of the person remembering.
Capital compounds fastest when it is never forced to leave. Judgment compounds only when it is written down.