Bull case, bear case thinking

The problem with a pitch

Most stock commentary is a pitch: one direction, one story, all the evidence marching the same way. Pitches are persuasive precisely because they're incomplete. The evidence that didn't fit got left in the drawer, and you have no way of knowing what was in the drawer. A one-sided case tells you what the author wants you to conclude. It tells you almost nothing about the stock.

A bull case and a bear case, published side by side, fix this structurally. The bull case is the honest argument the stock rises. The bear case is the honest argument it falls. Both are written to be as strong as the evidence allows — and that word 'honest' is carrying real weight.

Steelman, not strawman

A bear case written to lose is worse than no bear case at all, because it manufactures false confidence. 'Some investors worry about competition' is not a bear case; it's a strawman wearing one's armour. A real bear case names the specific thing that kills the thesis: the customer concentration, the refinancing due next year, the margin trend the bulls keep calling temporary. Our internal test is simple, and we've published it in the glossary: a pick where the desk cannot write a credible bear case is a pick that has not been tested.

When you read research — ours or anyone's — grade the bear case first. If it's specific, uncomfortable, and would genuinely change your mind if it came true, the research process behind the pick is probably real. If it's vague throat-clearing, assume the bull case got the same rigour.

How to weigh the two

Start with the bear case. Reading it first inoculates you against the narrative pull of the bull story — the same reason a judge hears the defence before sentencing, not after. Then ask three questions of each case. First: how many things have to go right for this case to win? The case that depends on fewer independent things is usually the sturdier one. Second: what would I expect to see in the next few results if this case is playing out? A good case makes near-term, checkable claims, not just a distant destination. Third: which case does the current price already believe? A stock priced for the bull case offers little reward for being right and plenty of punishment for being wrong.

Notice that none of this requires you to pick a winner immediately. Holding both cases in your head at once — genuinely unresolved — is uncomfortable, and that discomfort is the skill. Certainty is cheap. Calibrated doubt is what you're training.

The bear case as an early-warning system

Here's the practical payoff most readers miss: a published bear case is a pre-registered list of reasons to change your mind. If you act on a pick and the bear case starts coming true — the named risk materialises, the checkable claim fails — you don't have to improvise an exit rationale under stress. The desk wrote it down for you on day one, before anyone was emotionally invested. Investors who lose big rarely lose because they lacked information; they lose because they kept redefining what would count as being wrong. A bear case makes that redefinition harder.

Disagreement is the product

Inside a research desk, the bull and bear cases are usually written by people who genuinely disagree, and the published pick is the survivor of that argument. That's not a bug in the process — it is the process. The conviction score you met in Lesson I is largely a measure of how the argument went: how much of the bear case survived contact with the evidence.

So when you see both cases published under one pick, read it as a transcript of a fight, with a scoreline attached. Your job isn't to adopt our conclusion. It's to check whether the fight was fair — and then have your own.