What the market pays for, and what it only seems to.

Most Canadian portfolios are accidental macro bets wearing a stock-picker's clothing.

1. The benchmark is a starting point for measurement, not for thinking.

   Three sectors — financials, energy, and materials — account for the better part of the S&P/TSX Composite, currently around XX%. Build a portfolio by reference to that index and you inherit a set of macro exposures you never consciously chose: a bet on the domestic credit cycle, a bet on the oil strip, a bet on the gold price.

   I'd rather own those exposures when I have a view on them, in the size that view justifies, than absorb them by default because the index does. Benchmark-relative risk is a reporting requirement. It is not an investment idea.

2. Inefficiency in this market is structural, not cyclical.

   Sell-side coverage thins out sharply below the largest names, and thins again below the liquidity threshold that most institutional pools can transact in. That isn't a temporary dislocation that arbitrage will close — it's a permanent feature of a market this size. The capital that could correct the mispricing is often structurally unable to participate in it.

   That is the most durable edge available to a Canadian manager, and it is only harvestable with genuine patience. It rewards doing primary work others aren't doing, and it punishes anyone who needs the thesis validated on a quarterly timetable.

3. Know whether you own a compounder or a price-taker.

   Canada has a small number of genuinely excellent businesses that reinvest capital at high rates for long periods, and a very long tail of companies whose economics are set by a commodity price they do not control. Both can be good investments. They demand completely different analysis, different holding periods, and different sell rules.

   The expensive mistake isn't buying a cyclical. It's buying a cyclical while telling yourself a compounding story — and then holding it through the part of the cycle where that story stops being true.

4. Position size is the only honest statement of conviction.

   Everyone describes their ideas as high conviction. Sizing is where the claim gets tested. A position should reflect the shape of the distribution — how much I can lose if the thesis is wrong, how correlated it is with everything else I own, how quickly I could change my mind — and not how interesting the idea felt in the meeting.

   Concentration is how you get paid for research. It's also how you get hurt. The discipline is in sizing to the downside, not the upside.

5. A bad outcome and a bad decision are different things.

   In a business with this much noise, results over any short window say very little about the quality of the reasoning that produced them. Good decisions lose money regularly. Bad ones are rewarded often enough to be dangerous.

   The only defence I've found is to write the thesis down before the outcome is known — including what would prove it wrong — and then grade the decision against what I knew at the time, not against the price six months later. Most of what I've learned came from the positions that worked for reasons I hadn't anticipated.

6. Being early is indistinguishable from being wrong.

   From the outside — and from a client's monthly statement — there is no difference. "Early" is a description you're only entitled to use afterwards.

   So the thesis has to carry pre-committed falsification criteria: the specific operating results, balance-sheet developments, or competitive facts that would tell me the reasoning was wrong rather than merely untimely. Without that, patience is just a word for being unwilling to revisit the decision.

What I don't do.

• Forecast the macro I have views on the credit cycle and commodity prices, because in this market you cannot avoid having them. I try very hard not to build a portfolio whose success depends on those views being right.
• Trade on momentum I can't explain If I can't articulate the economic reason a price should be higher, the fact that it has been going up is not an argument I can size against.
• Confuse activity with work Turnover is a cost with a certain outcome and an uncertain benefit. The default action on most days should be no action.
• Outsource the thesis Street research is an input and a useful map of consensus. It is not a reason to own something, and it is never the thing I put in the file.