DCA bot settings
The settings a crypto DCA bot exposes: how the averaging grid is built, what it commits, where leverage puts liquidation, and when the position comes off. They are listed apart from the forex trade settings because the two markets do not share them — there is no pip here and no lot, and a crypto grid merges every fill into one position at a weighted average rather than keeping a chain. The same descriptions appear as the `?` hints beside each field in the wizard.
The averaging grid
7 settings| Base Order (quote currency) | The order the bot opens the moment its entry signal fires — the only one that fills for certain.Martingale orders below it are conditional: they fill only if price moves against you. So the base order decides what you make when the bot is RIGHT — price runs to the target and nothing else fills — while the martingale orders decide what you commit when it is wrong. Example: a 100 base under a 31,100 grid earns about 1 USDT on a 1% target and risks 31,100. That shape is a deliberate probe for some traders and a surprise for others; the wizard shows both numbers so it is a choice. |
| Martingale Orders | How many conditional orders sit below the entry, averaging the position down as price falls.More orders survive a deeper move, but each one is larger than the last, so the money at risk grows much faster than the count. Going from 5 orders to 8 at a volume scale of 2 does not add 60% to the commitment — it adds 700%. Other platforms call these averaging orders (3Commas) or DCA orders (Gainium) — the same mechanism under a different name. This is how many rungs the grid HAS; how many of them rest on the exchange at the same time is a separate field, Max Pending Orders. |
| Max Pending Orders | How many of the grid's martingale orders rest on the exchange at once. 0 sends the whole plan.Easy to confuse with Martingale Orders, and the difference shows up on the exchange: that field is how many rungs the grid HAS, this one is how many are live at any moment, the next going out as each one fills. A 25-order grid with this set to 5 shows five open orders on Binance and nothing is missing. Left at 0 the whole plan rests at once, which is what a grid you want to see in full looks like — the trade-off is a long list of open orders and a slower book to manage. |
| Overlap (%) | How far price may move against the entry before the grid runs out of orders.The one dial you can actually answer — "this could fall 30%" — so the grid means it exactly: the last martingale order lands on that percentage, not near it. Everything else is derived from it. Note this is the TOTAL span, not the distance to the first martingale order; platforms that ask for the first step instead call their field price deviation, and the two numbers are not interchangeable. |
| Volume Scale | How much larger each martingale order is than the one before it, measured in money.It multiplies the AMOUNT, not the coin quantity. A scale of 2 on a 200 base gives 200 → 400 → 800 → 1,600, and the whole grid costs base × (scale^(orders+1) − 1)/(scale − 1) whatever the market does — a figure you can check against a balance before starting. Higher scales pull the average entry down faster, which brings the target closer, and commit far more capital to do it. |
| Step Scale | How much wider each gap between orders is than the one before it.1.0 spaces the orders evenly across the overlap. Above that, early orders sit close together and later ones far apart — cheap averaging while the move is small, reserves kept for a move that is not. The first gap is solved backwards so the last order still lands exactly on the overlap. |
| First Martingale Order (quote currency) | Size of the first martingale order, set independently of the volume scale. 0 derives it as base order × volume scale.It exists because the multiplier alone cannot express a small probe followed by a real entry. Getting 50 → 500 out of the scale needs a factor of 10, which then runs away: 5,000, 50,000. Setting the first martingale order directly gives 50 → 500 → 1,000 → 2,000 at a scale of 2, and the multiplier goes back to meaning what it says. |
Capital and leverage
2 settings| Leverage | Borrowed size. It does not change what the grid costs you; it changes how far price may move before the venue closes the position for you.Higher leverage brings liquidation closer to your average entry, and a DCA grid is the configuration most likely to reach it, since buying while price falls is the strategy. The wizard checks whether the grid can reach its own last order before liquidation and says so — a 40% grid at 10x cannot, and every field in it still reads as valid. |
| Margin Mode | Whether this position risks only its own margin (Isolated) or the rest of the balance defends it (Cross).Cross postpones liquidation and puts everything else on the account behind this one bot. Isolated caps the damage at the margin allocated to the position, which is what makes a grid's worst case a number you can state. |
Target and stop
5 settings| Take Profit (% of average entry) | The target, measured from the position's average entry rather than from where you first bought.Every martingale order that fills drags the average down and brings the target with it. That is what makes averaging down work, and it is why one percentage is misleading on its own: a 1.5% target measured from the average can sit 12% below where you started buying. The wizard shows both. |
| Stop Loss Mode | What ends a losing cycle before the venue liquidates it: nothing, a stop tracking the average, or a stop that arms only after the grid is spent."None" leaves the venue's liquidation as the only floor — a deliberate choice for some DCA traders, but make it deliberately. "From the average entry" moves down with the average like the target does. "Only after the last martingale order" lets the grid place everything it has first, which is the mode that pairs with a wide overlap. |
| Stop Loss (% of average entry) | How far past the average entry the cycle is closed at a loss.Measured the same way the target is, so it moves as martingale orders fill. Set it inside the liquidation distance or the venue gets there first and the setting never fires. |
| Trailing Stop Activation (%) | How far past the average entry price must move before the trailing stop arms.Measured from the average entry the same way Take Profit and Stop Loss are — not from your first buy — so it moves as martingale orders fill. Set it at or beyond the take-profit percentage and the trail never arms: the position closes at target first, since both are measured from the same average, and the wizard warns when this happens rather than letting a dead setting pass as valid. |
| Trailing Stop Distance (%) | How far behind the best price reached the trailing stop follows, once armed.A tight distance locks in gains fast but gets tapped on small pullbacks; a wide one gives the position room but gives back more on a reversal. A distance wider than the gap between activation and break-even puts the trailing stop below your average entry the moment it arms — the wizard flags this combination, since neither number is invalid on its own. |
Liquidation price, leverage and margin ratio
The LIQUIDATION column is where the exchange would close the position, read from the exchange itself rather than worked out here. Beside it is how far away that is, as a percentage of the current price: grey above 15%, amber between 15% and 5%, red below 5%. Those two thresholds are our choice, not the exchange's — they are there so the cell can be read at a glance, and the number beside them is the one to trust.
This figure moves on its own, which is the reason it is on the screen at all. A grid averages down by design, so every martingale order that fills lowers the average entry AND drags the liquidation price up toward the market. A position that looked comfortable two rungs ago is a different position now. Under Cross margin it moves for a second reason: the whole wallet backs the position, so another symbol's losses shift this one's liquidation price with nothing happening here.
When the column shows a dash, nobody knows — and that is deliberate. We re-read the figure on every fill and once a minute, and if the last reading is too old to trust we show nothing instead of the old number. An out-of-date liquidation price is worse than none: it promises room that may already be gone. Under Isolated margin a reading stays valid until the position's size or its own margin changes, so it does not expire on a quiet position.
There is a case where the column is simply empty and nothing is wrong: under Cross margin the exchange often quotes no liquidation price at all, because the whole wallet stands behind the position and there is no single price at which it goes. Binance's own screen shows a dash there too. On a Cross position the number that answers "how close am I" is the margin ratio in the header, not this column.
WHICH PRICE the rest of the row is measured against is the exchange's MARK price, not the last trade. The mark is what the venue keeps score with — your unrealised profit, your margin ratio and the liquidation price above are all worked out from it — so Current, Change and P&L on a crypto row all come from that one number and agree with each other. On a thin market the mark and the last trade can sit some way apart, which is why the two are not interchangeable. If your figures differ slightly from the exchange's, check which price its own P&L is set to display: Binance offers Last Price and Mark Price, and only one of them is what it liquidates on.
LEVERAGE is per position here, not per account — two positions on one account can differ, which is why the header shows no account-wide figure. The column is off by default; turn it on from the column menu.
MARGIN RATIO in the header is maintenance margin as a share of your margin balance, the exchange's own measure of how close the account is to being closed out. It is not the margin LEVEL a cTrader account shows: level rises as an account gets safer, ratio rises as it gets more dangerous. Each account type shows the one its venue actually reports.
Editing the stop or target by hand
A running bot owns its TARGET. It works the target out from the position's average entry, so every time a martingale order fills the average moves and the bot replaces the target with a new one — a price you type by hand lasts until the next fill. That is what the grid is for: the target has to follow the average, or it asks for a profit measured from a price the position no longer has.
The STOP is different. If the bot has no stop of its own — stop-loss mode None, no trailing — it leaves yours alone. It used to cancel it and put nothing back, which left the position with no floor and said nothing; that is fixed. Clearing the field still removes the stop, and a bot that DOES have a stop configured still wins, because its own stop is worked out first.
One position, one set of levels: this venue keeps a single position per symbol and side and merges every fill into it, so there is no such thing here as a position the bot will not touch.