If you've ever wondered what a serious trade idea actually looks like — not a guess, not a hunch, but an actual structured setup the way an institutional desk thinks about one — this article is for you.
A setup is not a stock tip. It's not "Apple looks good." A setup is a structured argument. It has a starting point, a stopping point, an ending point, and a reason for being.
Here's what's inside one.
The five elements of a setup
Every serious setup answers five questions:
- What's the stock? The instrument the setup applies to.
- What's the pattern? The configuration that made it worth looking at.
- Where do you get in? The entry level.
- Where are you wrong? The invalidation level — the price that proves the setup failed.
- Where are you right? The target — the price that defines success.
If a setup is missing any of those, it isn't a setup. It's a guess wearing a setup's clothing.
What the engine looks for
The Sentinel engine is constantly scanning for configurations where these five elements line up cleanly. It's not looking for setups where one or two things look interesting. It's looking for setups where everything agrees.
That usually involves several signals all pointing the same way:
- Price building toward a meaningful level.
- Volume confirming the move (see our essay on volume).
- The sector behaving consistently with the stock.
- Macro indicators not actively working against the move.
- Pattern recognition matching configurations the model has seen succeed before.
When most of these line up — and that's not often — the engine surfaces the setup to the quantitative desk for review.
What the desk does next
The desk reviews every setup the engine surfaces. They're looking at things the engine can't see on its own:
- Recent news on the stock or the sector.
- Earnings calendar — is the company reporting soon?
- Overall market behaviour — is this a healthy tape or a fragile one?
- Any specific events that might invalidate the technical picture.
If the desk is comfortable, the setup gets approved and lands on client dashboards. If they're not, it doesn't. Some days the desk approves a lot of setups. Some days they approve none. That's the point of having them.
An engine without a desk is reckless. A desk without an engine is too slow.
What you see on your dashboard
When a setup reaches your dashboard, every element is laid out for you:
- The ticker — which Fortune 100 stock it is.
- The thesis — what the engine saw, in plain English.
- The entry level — the price the setup begins at.
- The invalidation level — the price the setup is wrong at.
- The target — the price the setup is aiming for.
- The desk note — why a person signed off on it, including any context the engine alone couldn't have weighed.
What you do with that information is entirely your call. Your account. Your name. Your decision.
The point isn't to tell you what to do. It's to give you the same kind of structured argument an institutional desk would give itself — so when you make a decision, you're making it from the same starting point as the professionals.
A real setup answers five questions. If a "trade idea" is missing any of them, it isn't a trade idea — it's a guess. Sentinel's job is to make sure that when you see a setup, all five are in front of you.