Formulas

Reference math

Technical formulas and calculations used by the bot. Reference material for advanced users who want to understand the exact math behind grid spacing, grid distance , rebalancing, indicators, and risk limits.

Grid Spacing

Grid levels use geometric (percentage-based) spacing, not linear spacing:

level_price = anchor_price × (1 + spacingPct/100)^n

Where n is the level number (positive for sells, negative for buys).

Example: With anchor at $2,000 and spacingPct: 0.37:

Level +1: $2,000 × 1.0037 = $2,007.40
Level +5: $2,000 × 1.0037^5 = $2,037.14
Level -3: $2,000 × 1.0037^-3 = $1,977.92

Grid distance

Grid distance determines how far price can move before one side depletes:

slots = floor(inventoryUsd ÷ orderSizeUsd)

long grid distance  = ((1 + spacingPct/100)^slots - 1) × 100
short grid distance = (1 - (1 + spacingPct/100)^-slots) × 100

Example: With $500 seeded inventory and 0.37% spacing:

Long:  $500 ÷ $10.00 = 50 slots → 20.3% upside distance
Short: $500 ÷ $5.50  = 90 slots → 28.3% downside distance

Short grid distance cannot exceed 100% because price cannot fall below zero.

Rebalancing Distribution Rate

The rate at which excess position is corrected:

rate = clamp(baseRate × pivotRatio / ratio, baseRate, maxDistributionRate)

Where:

baseRate = 2.5% (fixed)
pivotRatio = config value (default 10.0)
ratio = current imbalance / orderSizeUsd
Result is clamped between baseRate and maxDistributionRate

Behavior:

Small imbalances (low ratio) → higher rate → resolve quickly
Large imbalances (high ratio) → lower rate → unwind gradually

Example: With pivotRatio: 10.0, maxDistributionRate: 20.0, baseRate: 2.5%:

Setup: baseRate = 2.5% · pivotRatio = 10.0 · maxDistributionRate = 20.0%

Imbalance ratio	Raw result	Clamped?	Applied rate
0.8	2.5% × 10.0 / 0.8 = 31.25%	Yes (cap at 20%)	20.0%
1.5	2.5% × 10.0 / 1.5 = 16.7%	No	16.7%
2.0	2.5% × 10.0 / 2.0 = 12.5%	No	12.5%
5.0	2.5% × 10.0 / 5.0 = 5.0%	No	5.0%
50	2.5% × 10.0 / 50 = 0.5%	Yes (floor at 2.5%)	2.5%

ROE% (Return on Equity)

Used by the Position Balancer to check profitability:

Long position:

ROE% = (markPrice - avgEntryPrice) / avgEntryPrice × 100

Short position:

ROE% = (avgEntryPrice - markPrice) / avgEntryPrice × 100

When avgEntryPrice is zero (no position), ROE returns zero.

Position Utilization

Used by Position Balancer tiers. Utilization is based on net exposure:

netExposure = abs(positionUsd - counterpartPositionUsd)
utilization = netExposure / maxNetExposureUsd

Example: Long $760, Short $200, with maxNetExposureUsd: 1000:

netExposure = abs($760 - $200) = $560
utilization = $560 / $1,000 = 0.56 (56%)

This would trigger the 50% tier (multiplier 1.25) but not the 75% tier.

Hedge Guard Activation

The Hedge Guard activates based on position ratio:

Entry condition:

longUsd < shortUsd × entryThresholdPct

Exit condition:

longUsd > shortUsd × exitThresholdPct

Setup: entryThresholdPct = 0.667 · exitThresholdPct = 0.9

Your long	Your short	Long/Short ratio	Condition	Result
$400	$700	0.57	0.57 < 0.667 (entry)	Activates
$600	$700	0.86	0.86 < 0.9 (exit)	Stays active
$650	$700	0.93	0.93 > 0.9 (exit)	Deactivates

Indicator Calculations

RSI (Relative Strength Index)

Standard RSI formula with configurable period (rsiLength):

RSI = 100 - (100 / (1 + RS))
RS = average gain / average loss over period

Bollinger Bands

middle = SMA(price, bbLength)
upper = middle + (stdDev × bbStdDev)
lower = middle - (stdDev × bbStdDev)

Price below lower band = oversold signal (for longs) Price above upper band = overbought signal (for shorts)

WaveTrend

A momentum oscillator combining channel and smoothing:

ap = (high + low + close) / 3
esa = EMA(ap, wtChannelLen)
d = EMA(|ap - esa|, wtChannelLen)
ci = (ap - esa) / (0.015 × d)
wt = EMA(ci, wtAverageLen)
signal = SMA(wt, wtSmaLen)

WT below wtOversoldThreshold = oversold WT above wtOverboughtThreshold = overbought Momentum = direction wt is moving relative to signal line

PnD Detection

Pump-and-dump protection triggers when:

closeFillCount ≥ closeFillsThreshold within withinSeconds

Example: With closeFillsThreshold: 8, withinSeconds: 60:

If 8 close orders fill within 60 seconds (~3% move at 0.37% spacing), the bot enters cooldown.

Effective Leverage

In Bybit hedge cross margin, long and short positions net each other. Only the difference between them drives effective leverage:

effectiveLeverage = abs(longUsd - shortUsd) / walletBalance

Setup: wallet balance = $500

Your long	Your short	Net exposure	Effective leverage
$500	$500	$0	0x
$1,000	$600	$400	0.8x
$1,500	$500	$1,000	2x

Note: Bybit's displayed per-position leverage is not the same as effective leverage across both sides.