CubeQuants Complete Parameter Guide

Parameter 1 (Protection Zone): The minimum safe distance between the current price and the average position price. If the price moves beyond this range, opening positions in the same direction is prohibited. 📘 Example: current price 2000, Parameter 1=500 → you cannot add to the long position when the long average price is ＞1500; you cannot add to the short position when the short average price is ＜2500. Parameter 2 (Leverage Cap): The maximum leverage multiplier actually used. It caps the notional position value. 📘 Example: capital 1000, Parameter 2=3 → you cannot add to the long position when the long notional value is ＞3000; you cannot add to the short position when the short notional value is ＜-3000. 💡 Risk implication: This combination keeps position leverage within 3x and can withstand an adverse move of about 33.3% measured from the average price.

Parameter 1 (Number of Orders): The maximum number of orders with unrealized PnL. Together with the per-order amount (5% of capital), it controls total leverage. 📘 Example: Parameter 1=60 → a full position is about 3x leverage and can withstand a 33% decline; Parameter 1=30 → about 1.5x leverage, can withstand a 66% decline. Parameter 2 (Tiered Limit): During unrealized losses, the cap is dynamically reduced according to the "Black Swan Index" — the higher the index, the fewer orders. 📘 Example: index 1 → 5 orders, index 5 → 25 orders. At the maximum of 25 orders, leverage is about 1.25x and can withstand an 80% decline. 💡 Design intent: Automatically reduce position size during unrealized losses, so that buying more as the price falls does not lead to liquidation.

Parameter 1 (Number of Orders): Symmetric to the long-side logic; controls the total leverage of the short position. 📘 Example: Parameter 1=60 → a full position at 3x leverage can withstand a 33% rally; Parameter 1=30 → 1.5x leverage, can withstand a 66% rally. Parameter 2 (Tiered Limit): Dynamically reduces the short position cap during unrealized losses. At the maximum of 25 orders, leverage is about 1.25x and can withstand an 80% rally. 💡 Design intent: Symmetric long/short protection — no matter which way the market moves, risk exposure is automatically reduced during unrealized losses.

Parameter 2 (Rebound Trigger Value): When the price rebounds while an order is in unrealized loss, that order is stopped out if the rebound reaches this value. 📘 Example: a long order is opened at 2000, falls to 1600 and then rebounds to 1960 (a 360-point rebound) → the system automatically stops out that order. 💡 Purpose: Eliminate orders that fail to rebound meaningfully after a deep unrealized loss, clean up weak positions, and reduce overall risk.

Parameter 2 (Deviation Index Retracement Value): When the deviation index of the entire position retraces from its extreme by this amount, close the whole position immediately. 📘 Example: the long position's deviation index is -1000 and the deviation range is -7400; when the index recovers to -1000 (a 6400-point retracement) → the system closes the entire long position. 💡 Core idea: "Take the rebound after a deep loss, then cut and exit." In an uncertain market, discarding losing exposure in time is a key risk control strategy.

Parameter 1 (Order Count Threshold): Once the number of orders with unrealized PnL exceeds this value, the system waits for overall PnL to return to zero and then closes everything and restarts. 📘 Example: 15 long orders + 15 short orders = 30 → the condition is triggered; when PnL returns to 0, everything is closed. Parameter 2 (One-Sided Unrealized Profit): When the unrealized profit of the long or short side exceeds 5% of capital, close that side to lock in the gain. 📘 Example: capital 1000U, long-side unrealized profit ≥50U → close the long position. Parameter 3 (Overall Unrealized Profit): When the account's total unrealized PnL exceeds 1% of capital, close everything to lock in the gain. 📘 Example: capital 1000U, overall unrealized profit ≥10U → close all positions. 💡 Strategy combination: Parameter 1 clears out ineffective positions, while Parameters 2 and 3 "take the money and run," so that avoiding a loss counts as a win.

Mean method (recommended): Determines direction from the mean trend of the last 5 minutes of data — steady and moderate.
Extreme method: Determines direction from the high/low extremes of the last 5 minutes — more responsive (higher risk).
Signal method: Determines direction from the high/low values of the last 8 hours — more lagging (lower risk). 💡 Note: All methods make decisions based on the previous 8 hours of data. They are incompatible with traditional systems such as MT and use non-standard sampling.

TP multiplier: Uses 1% of the current price as the base unit, multiplies it by the TP multiplier, and adds Black Swan Index weighting. 📘 Example: long opened at 2000, TP multiplier 2, short-side Black Swan Index 3 → TP price = 2000 + 2000×1%×2 + 2000×1%×3 = 2100. Adding distance: When the long/short Black Swan Index exceeds the set value, adding to the position is triggered, with up to 10 consecutive orders added each time. 📘 Example: long average price 2000, current price 1940 (index 1); when the price falls to 1880 and the reference point is long → adding to the position is triggered, lowering the average price. Density: Controls the positive/negative price difference from the previous order when opening in the same direction. 💡 Tip: The SL multiplier and grid spacing are being deprecated, so no special attention is needed. Adding to positions is governed by [Position Control] and can be enabled as needed.

Configuration: When using proxy software such as Clash, enter the local port; clear this field after enabling TUN mode. 💡 Recommended: Use a Clash subscription and configure the port. This avoids a global connection and does not affect normal computer use.

Purpose: Manually specify the current major trend direction (long/short), which affects order weighting. 💡 Strongly recommended: Set this manually every time you restart the software. If you don't, the first order may take more than 8 hours to appear (because the system needs to accumulate data). Experienced users can use this to make decisions more efficiently.

Recommended ratio: Each order is 5% of capital; with 60 cumulative orders the effective leverage is 3x, which can withstand at least a 33% move. 💡 Core logic: Orders are dispersed and their PnL is independent, using a "voting principle" to follow the trend and reduce position risk.

Parameter 2 (Amplitude Threshold): An alert is triggered when the price amplitude over the last 8 hours exceeds this value.
Parameter 3 (Alert Trigger): A "buy the dip / short the top" alert is issued when the price retraces from its high to this value.
Parameter 4 (Alert Cancellation): The alert is cancelled when the price retraces further to this value. 💡 Intended for: Experienced manual traders looking for a reference. Beginners are advised not to trade manually based on this signal.

Order switches: Control opening, take profit, and stop loss for the long and short sides separately.
Log switches: We recommend enabling "Order Log" to record trading data; "Trend Log" produces a large volume of data, so we suggest leaving it off.
Simulation switch: Always enable it when testing and always disable it for live trading (when enabled, no real orders are placed). 💡 Recommended setup for beginners: Check Long Open, Long TP, Long SL, Short Open, Short TP, Short SL, and Order Log.