Opportunity Radar
This is a research candidate, not investment advice, not an order, and not a profitability guarantee.
That is the platform's own disclaimer, returned by the service with every opportunity detail on this screen. The Radar does not currently render it, see the known gap. It applies regardless.
What is this?
A scanner that runs three technical detectors across an instrument universe and reports, per instrument, the setups it found. Each with a Setup Quality score, a trigger level, an invalidation level, and the evidence behind every point of the score.
It is the most concrete of the discovery screens: a Radar entry names an instrument and two prices.
Why does it exist?
The strategy builder is excellent once you know what you want to trade. Deciding what is a different problem, and the honest version of solving it is not "here is what to buy". It is "here is where a defined condition is currently present, here is how completely, and here is where it stops being true".
That framing is what makes the screen defensible and also what makes it useful. A score you can audit is a score you can disagree with.
How does it work?
| Step | Stage | Detail |
|---|---|---|
| 1 | Universe | The scanned instrument set for the selected market |
| 2 | Bars | OHLCV history per instrument, 1-minute or daily |
| 3 | Detectors | Breakout, momentum continuation, volatility expansion |
| 4 | Score | Five components, 100 points maximum |
| 5 | Penalties | Up to 30 points deducted |
| 6 | Band | STRONG, GOOD, WATCH or LOW |
| 7 | Floor | Candidates below 50 are not published |
| 8 | Publish | Stamped into a snapshot |
- Universe. The scanned instrument set for the selected market.
- Bars. OHLCV history per instrument. An instrument without enough history is skipped, and the skip is counted rather than hidden.
- Detectors. Breakout, momentum continuation, volatility expansion. Each is a defined technical condition. See the detector list.
- Score. Five components, 100 points: Setup Strength 35, Liquidity 25, Regime 20, Timing 10, Quality 10.
- Penalties. Up to 30 points off for event proximity, stale data, extreme spread or an invalid market state.
- Band. STRONG ≥ 80, GOOD ≥ 65, WATCH ≥ 50, LOW below.
- Floor. Candidates below 50 are not published. Only
READYandTRIGGEREDstatuses are. - Snapshot. The result set is stamped with a snapshot ID, an as-of time and a scoring version.
The list polls for a fresh snapshot roughly once a minute. It is a snapshot feed, not a tick feed.
How do you use it?
The list
Opportunities ranked by Setup Quality. Ties break on liquidity score, then symbol, so the order is stable rather than arbitrary.
Filters
| Filter | What it narrows |
|---|---|
| Segment | Asset segment |
| Setup type | Which detector, breakout, momentum, volatility |
| Timeframe | The detector's evaluation timeframe |
| Min liquidity | Floor on the liquidity component |
| Status | READY, TRIGGERED, and the rest of the lifecycle |
| Freshness | FRESH, DELAYED, STALE |
The detail panel
Selecting an opportunity opens:
| Section | What it gives you |
|---|---|
| Score breakdown | Each component, its points, its maximum, and its definition |
| Penalties | Each penalty applied, with the reason |
| Why now | Plain-language evidence statements, computed from the same numbers as the score |
| Trigger / Invalidation | Both levels with their labels, plus trigger distance as a percentage |
| Event risk | Scheduled events near the setup |
| Freshness | State and age in seconds |
| Change since last snapshot | Previous score, previous as-of time, and which components changed |
The change block is the one most worth reading. A score that rose from 61 to 82 because liquidity improved is a different situation from one that rose because the setup strengthened.
Three actions
| Action | What it does |
|---|---|
| Explore on chart | Opens the instrument's chart at the detector's timeframe |
| Build strategy | Hands off to the AI engine, see From a setup to a strategy |
| Watchlist | Saves the candidate to your opportunity watchlist |
Note what is absent: there is no order button, and no size suggestion.
Example
Illustrative. Not a real result.
RELIANCE, breakout, daily. Setup Quality 82, band STRONG, status READY.
| Component | Points |
|---|---|
| Setup Strength | 31 / 35 |
| Liquidity | 22 / 25 |
| Regime | 16 / 20 |
| Timing | 8 / 10 |
| Quality | 10 / 10 |
| Penalty: high-impact event proximity | -5 |
| Total | 82 |
Why now
- Close 2,847.30 is above 20-session high 2,841.00 (+0.22%).
- Volume 4.2M vs 20-session average 2.8M (1.50×).
- Trigger 2,841.00; distance +0.22%.
Trigger 2,841.00. Invalidation 2,769.50. Freshness FRESH, 42 seconds old.
Event risk Quarterly results in 3 sessions.
What to take from it: the setup is present and liquid, the regime is broadly aligned, and a scheduled event three sessions away cost it five points. Whether an earnings print in three sessions makes this tradeable is a judgement the screen cannot make for you, which is exactly why the event is listed rather than silently priced in.
Watchlist
/watchlist/opportunities holds candidates you saved. It is a different list from the
strategy watchlist, which tracks published strategies.
Saved candidates carry their snapshot reference, so you can tell later whether you are looking at the state you saved or a re-scored version.
Limitations
It is a snapshot, not a live feed. Between snapshots the market has moved. TRIGGERED
means triggered as of the snapshot.
Skipped instruments are counted, not scored. An instrument without enough bars produces no opportunity. Absence from the list is not a statement about the instrument.
The floor hides the tail. Candidates below 50 are never published. You are seeing a filtered set by design.
Detectors are fixed. Three defined conditions. A setup outside them will not appear, however good it is.
Coverage is uneven. Instruments differ in liquidity, history depth and data availability. The Quality component reflects that, which is why it exists.
Invalidation is not a stop-loss. It is the price at which the pattern is void. Using it as a stop is a decision with different consequences, position size against your own risk budget, not against a level a scanner computed.
The disclaimer gap
Stated plainly because it is a live compliance finding rather than a design choice:
The service returns a disclaimer string with every opportunity detail, and the field exists in the client's type definition. The Radar page does not render it. Trading Map, Cluster Density and Analog Events all render theirs correctly; Radar is the exception.
Radar is also the screen where it matters most. It names an instrument, a trigger and an invalidation level, then offers a Build strategy button. Rendering the disclaimer persistently in the detail panel is a P0 item on the compliance roadmap. Until it ships, the disclaimer is reproduced at the top of this page, and it applies whether or not the screen shows it.
Next
- From a setup to a strategy
- Trading Map (the same detectors, whole-market view
- Universe Scanner) your filters rather than fixed detectors