The bear case is the service.

If you read only one part of a research brief from this desk, read the risks. Not because we think every company we cover is about to disappoint, but because the risks list is the part we work hardest on and the part that most research does worst.

The usual version is a list of nouns. "Regulatory risk. Commodity price risk. Execution risk." It is written to be present rather than read, and it tells you nothing a reader could act on. A risk written properly names a mechanism: what would have to happen, to what, by when, and what it would do to the thing you own. "The merger could close later than guided, and until it closes the shares trade on the deal, not the copper." That sentence has a date in it, a dependency, and a consequence. It can be wrong, which is what makes it useful.

After a dozen briefs we have noticed that the risks list is also where a thesis gets honest. Writing the bull case is easy; the company has already written it for you. Writing the bear case means reading the same filing for the parts that argue against the headline: the customer concentration in a growth story, the deposits held by a foreign treasury, the guidance that moved twice, the backlog that depends on a budget someone else controls. By the time the list is written, we usually know whether we still believe the thesis. Sometimes we do not, and the brief changes.

This is also why the bear case gets the same type size as the bull case on the page. A warning in small print is a warning the author did not mean. We would rather a reader finish a brief less certain and better informed than more certain and less. The first is what research is for. The second is what marketing is for, and there is already plenty of that.

There is a practical point for the U.S. reader in all this. Canadian companies, especially the smaller ones, carry risks that do not appear in the same form in American filings: financing structures that dilute by design, provincial permitting that runs on its own clock, a currency that moves against the reporting line. Our risks lists try to name those specifically, and the primers in the library explain the mechanisms once so the briefs do not have to repeat them.

What would change our view: a reader telling us that a risks list talked them out of something that then went right. We would want to know whether the risk was wrong or merely did not happen. Those are different, and only one of them is a mistake.