How is a price target and stop calculated on a 4h or 1d chart?
The green and red zones the BT365 Scanner draws on a perps chart are one trade setup: where it would enter, where it would stop out, and where it would take profit. The levels come from fixed arithmetic on candle data. No model picks them, and the setup carries no probability of working.
This article sets out what is drawn, what each number is based on, how the 4h and 1d versions differ, and what the zones do not tell you. It covers trend setups. Rows the Scanner labels "Reversal" come from a separate rule set and are not described here.
What are the green and red zones on the BT365 chart?
They are a setup you loaded, from a row in the Scanner or from an Analyze request on the trade terminal. Loading one fills three things at once: a setup card, the zones on the chart, and the order form.
- A dashed line at the entry price, labelled "Entry".
- A green zone from entry to the target, with a badge such as "Target +4.2%".
- A red zone from entry to the stop, with a badge such as "Stop -1.9%".
The percentages are the price move from entry, before leverage. When the trade has a size, the badge adds the dollar amount.
Three things about the drawing are easy to misread:
- The width means nothing. The zones reach six candles back and sixty forward on every timeframe. That is a drawing choice, not a time estimate.
- It is a snapshot. The levels are fixed when you load the setup and do not follow the market. Reloading only re-sizes the same levels to your current risk amount. The zones go away when you dismiss the setup, run a new analysis or switch market.
- It follows the setup's timeframe. Loading a 4h setup switches the chart to 4h. The Scanner covers nine timeframes from one minute to one day; this article covers 4h and 1d.
What is the direction of a setup based on?
On indicators computed from the venue's candles: 200 candles of the chart timeframe, and 100 of a higher timeframe. A 4h setup needs at least 55 four-hour candles and 55 daily candles; a 1d setup needs 55 daily candles. Below that there is no setup.
Five readings are added into a direction score between -1 and +1:
| Reading | What it measures | Most it adds or subtracts |
|---|---|---|
| Moving averages (EMA 9, 21, 50, 200) | Whether the short averages sit above or below the long ones | 0.30 |
| MACD | Whether momentum has crossed up or down | 0.25 |
| RSI (14) | How far the price has run in one direction | 0.20 |
| Bollinger bands | Where the price sits inside its recent range | 0.15 |
| ADX (14), above 25 | Whether a trend is established | 0.15 |
The sum is then multiplied by 1.05 when the latest candle's volume is more than 1.2 times its 20-candle average, and by 1.15 when it is more than double. That pushes the score further in the direction it already had.
The rule is a threshold. At +0.20 or above the setup is long. At -0.20 or below it is short. In between there is no setup, and the Scanner says so.
The number shown as "Score" on the card runs from 0 to 100 and starts from the same readings. It is then adjusted twice:
- The higher timeframe. Agreement adds 8 points; disagreement takes 10 away. On a 1d setup the weekly check runs only when the market has 55 weekly candles, a little over a year of history; without it the score gets no adjustment.
- Funding. The current funding rate moves the score by up to 5 points either way, read against the crowd: a side that is expensive to hold loses points. Ordinary funding moves nothing.
The order book, liquidations, open interest and news are not inputs.
How are the stop and the target calculated?
From one measure: the average true range over 14 candles, or ATR. It is the typical range of one candle, gaps included, so it scales with how much the market is moving.
- Entry is the mark price at the moment the setup is read.
- Stop is a fixed multiple of ATR away from entry.
- Target is the stop distance times a fixed reward ratio.
Then two checks apply:
- Levels in the way. If a support or resistance level sits between entry and target, the target is pulled in to just before it, by a tenth of an ATR. Levels are price pivots the market has touched at least twice. A target pulled in this way must still pay at least 2 times the risk, or the setup is dropped.
- A sanity cap. On the 4h chart a setup is refused if the stop is more than 12% from entry or the target more than 42%. On the 1d chart the caps are 35% and 75%.
How do the 4h and 1d setups differ?
| 4h | 1d | |
|---|---|---|
| Candles used | Four-hour | Daily |
| Higher timeframe checked | Daily | Weekly |
| Stop distance | 1.5 × ATR | 2.5 × ATR |
| Reward ratio aimed for | 2.25 | 3.5 |
| Target distance before any cap | 3.375 × ATR | 8.75 × ATR |
| Lowest score that shows a standard setup | 40 | 45 |
An illustration with round numbers, not a market: a price of 100 with a four-hour ATR of 2 gives a long setup with a stop at 97 and a target at 106.75. The same price with a daily ATR of 5 gives a stop at 87.50 and a target at 143.75.
A daily setup is wider in both directions. It risks more per unit and asks the market to travel further.
One exception to the score floor: a market that moved 10% or more in 24 hours can appear in the Scanner with a score as low as 25, if the setup points the same way as the move. Those rows are tagged "Mover".
The time estimate on the card ("~2–6 d to target") comes from the same inputs. It divides the target distance by the share of each candle's ATR that a market usually keeps as progress: 15% in a weak trend, up to 50% in a strong one, read from ADX. For an uncapped target the central figure is 7 to 23 four-hour candles on 4h, and 18 to 59 days on 1d. The card shows a range around it, from about half to about twice that figure, so a daily setup can read in weeks or months.
Does an AI model choose the levels?
No. The function that decides direction, entry, stop and target is plain arithmetic: it reads no database, makes no network call and calls no model.
A language model appears in one place. On the Analyze path it writes the short explanation at the top of the setup card, after the levels exist; its instructions state that the levels are already decided and that it cannot change them. Analyze reads six timeframes at once, and the model writes only for the highest-scoring one. If it fails or takes more than five seconds, the card shows a sentence built from the same indicator readings instead. The "Why this setup" breakdown lower on the card is arithmetic, not model output. The Scanner path calls no model.
Does the score mean the target is likely to be hit?
No. Three limits are written into the product itself:
- The score is not a win chance. The card's own hint says: "Higher is stronger, not a win chance." Nothing in the calculation estimates how often a target is reached before a stop.
- The time estimate is conditional. It is the time the target side implies if the price gets there. It is not a deadline, and it says nothing about whether the stop comes first.
- No hit rate is shown. Nothing in the Scanner or on the card says how often these setups reached their target.
Where this stops
- Nothing is placed for you. Loading a setup fills the order form with a limit order at the entry price, plus the stop and target. You still confirm the order, and edits you make in the form are not reflected in the zones.
- A market order moves the form, not the zones. If you switch the order to market, the stop and target in the form follow the live price at the same distances, while the zones stay where they were.
- A reading goes stale. A Scanner row read more than 30 seconds ago is checked again before it loads; if the setup is gone, it does not load. A row older than 15 minutes is marked as needing a live re-check.
- Loaded zones carry no age. A setup already on the chart is never re-read, so its zones can be hours old.
- The reading behind a row can be old too. The Scanner can keep a 4h reading for up to two hours and a 1d reading for up to half a day before rebuilding it.
- The Scanner does not read every market. On 4h and 1d it reads up to 250 of the most-traded markets, plus the day's large movers.
- A target can sit beyond anything the market has traded. This happens on new listings and on breakouts past the traded range. The setup then carries a warning; the target is a volatility figure, not a level.
- The Scanner and Analyze are gated. They need a logged-in wallet that holds at least 10,000 BT365 on Base, or an active access code.
- Autopilot is a separate feature. It reads the same scan results but has its own gates and rules. The zones on a chart never open or close a position.
- It is a tool, not advice. The setup is one mechanical reading of candles. It knows nothing about news, your other positions or what you can afford to lose.
Related
- Can you change the leverage on an open perps position, and what does it do to your liquidation price?
- Why did markets fall this week, and how can you hedge? for stops and hedges in practice
- How do you know what a market order will cost before you send it?
BT365 is not available in every jurisdiction, and nothing on this page is investment advice. Trading perpetual futures and prediction markets carries risk of loss.