Overview
Dashboard
🛑
Trading halted by admin
No new trades will be placed until trading is resumed.
⚠️
Approaching daily loss cap
— REACHED
Today's PnL:
· Cap:
Model
Timeframe
trades
·
Net PnL
Win Rate
Total Active Capital
Total Trades
Cumulative PnL
Running net realized profit & loss
USDT
Strategy Mix
Open Positions
| Opened | Model | Pair | Side | Entry | Size | Lev | Status |
|---|---|---|---|---|---|---|---|
Recent Trades
Filtered live by your model + timeframe selection
| When | Model | Pair | Side | Entry | Exit | Net PnL | Status |
|---|---|---|---|---|---|---|---|
Configuration
Models
of
active
●
●
Notional: $
Risk: $ per stop
Position sized so hitting the stop loses exactly this $. A tight stop
buys a bigger position — capped at $200 notional.
Signals with a stop tighter than
are skipped — the fee eats too much of their risk.
● Unsaved
Records
Trade History
| Status | Model | Pair | Dir | Entry → Close | PnL % | PnL $ | Closed |
|---|---|---|---|---|---|---|---|
| → | |||||||
| No trades match your filters | |||||||
Analytics
Statistics
Cumulative PnL
By Model
Daily PnL
All models combined
Performance breakdown
All models
| Model | Trades | Win rate | Total PnL | Biggest win | Biggest loss |
|---|---|---|---|---|---|
|
|
Research
What-if
Replay the same history under different rules. The signals, the SL and the TP levels never change — only which trades you take and how much you stake.
Data
Every signal the models fired. Big sample, but the outcome is the generator's own
simulation — the fee below is applied on top.
Only trades that really executed on BingX. Real money, but a much smaller sample.
Round-trip fee (% of position)
BingX actually charged 0.102%
across 425 closed positions.
1 · Minimum stop distance
The fee costs fee ÷ stop of your risk.
On a 0.3% stop that is a third of it — more than the models' edge.
1b · Regime (circuit breaker)
Market data: volatility straight from exchange prices (15m + 1h ATR
across all pairs), independent of the models. 1.0
= a typical day. This is what would run live — the executor recomputes it from klines.
Model-based: median stop (= 1.5×ATR) of all signals over the trailing 24h.
Depends on when models fire and their ATR window — noisier, and slightly different
live vs backtest.
2 · Position sizing
Each model staked exactly as its Models tab says
.
This is the baseline everything else is compared against.
$
margin ×
leverage — same stake on every model
every stop costs exactly $
→ every TP wins ~2×
Portfolio
models selected — the numbers above describe exactly this set
Click a model to add or remove it. PF is its profit factor under the rules above —
anything under 1.0 loses money.
currently trading in your account
Recomputing…
Fixed risk with no filter is the worst setting.
To risk the same dollars on a tight stop you must open a huge position — and the fee
is charged on the position. You end up paying the most fee exactly where the stop is
tightest. Turn on a minimum stop.
Net PnL
Max drawdown
Profit factor
Return ÷ drawdown
Trades taken
Skipped by the filter
Win rate
Avg / trade
Actually made
Position size this would require
Smallest
Median
Largest
Needs leverage up to
Equity curve
Under these rules
Per model
Does the rule help everyone?
A rule that helps overall can still destroy one model — some models earn precisely
on the tight stops a filter would remove.
Scenario staked at
— your own stake is in the Your stake column, so the totals are not
directly comparable. Judge by return ÷ drawdown, not by the absolutes.
| Model | Your stake | Trades kept | Win rate | Real PnL all trades |
Scenario PnL under these rules |
Change | PF | Max DD |
|---|---|---|---|---|---|---|---|---|
| Trading now | ||||||||
|
Not trading
These have no stake configured, so they are simulated at
per trade. Judge them by PF, not by PnL.
|
||||||||
| not set | — | — | ||||||
Research
Model Analyzer
Deep-dive on one model, then compare it against every other model — and any combination — under the same filter and sizing rules as What-if.
Model
Filters (same as What-if)
min stop
regime ≥
sizing
margin $
×
x lev
$ per stop
Custom margin: every trade uses the same
notional.
PnL = notional × (price move − fee).
Fixed risk: every position sized so hitting the stop loses exactly —
SL/TP levels unchanged, so win rate is identical, only the $ amounts scale. Same as What-if.
Analyzing…
Total PnL
Win rate
Profit factor
Max drawdown
Return / DD
Trades
— equity curve
combination ( models)
Performance by market regime
Where this model actually makes its money.
| Regime | Trades | WR | PF | PnL |
|---|---|---|---|---|
| By band (disjoint) | ||||
Performance by pair
Which markets it trades well.
| Pair | Trades | WR | PnL |
|---|---|---|---|
Compare with other models
Tick models to build a combination (shown as the dashed line above). Corr = how this model's
monthly returns move with — negative is good for diversification.
| + | Model | PnL | WR | PF | Max DD | Corr |
|---|---|---|---|---|---|---|
| • (focus) |
Combination ( models):
PnL
WR %
PF
Max DD
Account
Settings
Profile
Your account information
Display name
Telegram username
Telegram ID
Status
Safety
Auto-suspend trading if today's realized losses exceed this cap. Leave empty for no cap.
of cap
✓ Saved
Market volatility filter
Now:
The models only have an edge when the market is actually moving. This pauses
all trading whenever market-wide volatility falls below your
threshold (1.00 = a typical day, straight from exchange candles — not model-based).
✓ Saved
Trading is paused — regime
is below your threshold.
Trading is active — regime is at or above your threshold.
BingX connection
Connected
Not connected
Paste your BingX API key + secret. Trades go through your account; we never withdraw.
API key
••••
Last verified
Confirm in Telegram
We sent a confirmation prompt to @Authentificator_nexus_bot. Tap Approve there to save.
Expires in s. Nothing saves until you confirm.
✓
BingX keys saved
Verified and stored. You can now run live trading once you enable models.
Failed