Digital-Asset Perpetuals · gTrade

abstract

This paper describes the methodology of the Q7 S1 algorithm as it trades crypto perpetual futures on gTrade (Gains Network), an on-chain perpetuals exchange. The algorithm trades Bitcoin (BTC), Ether (ETH), Solana (SOL) and XRP (XRP) on four timeframes — one, three, five and fifteen minutes — as sixteen independent configurations, each with its own calibrated parameters and its own capital. Four independent entry engines propose trades; an additive confluence score decides whether a proposal is taken and, through Q7's ELU system, the leverage at which it is posted; a liquidation guard computed from that leverage and a risk governor bound how large it may be; and a four-leg exit ladder, with every leg calibrated, manages it to its close. Built into the engines is a set of research-based enhancements, each drawn from the published literature or stated plainly as a Q7 construct, that add evidence to the entry decision, stand the engines down in abnormal conditions, scale position size down when risk is concentrated, and shape profit targets to the market's current behaviour. Every parameter that is not a risk rule is calibrated per configuration on historical data and validated on data the calibration never saw. Risk rules are fixed and are never calibrated.