From backtest curves to live execution: slippage and capacity limits

The gap between theoretical optimum and execution friction

Under frictionless assumptions, high-turnover mean-reversion signals produce deceptively smooth equity curves. In live order matching, bid-ask spreads, exchange fees, and latency drag systematically erode capital.

Many strategies with attractive theoretical returns merely transfer expected profits to liquidity providers once execution impact is accounted for, decaying toward zero or negative net yield.

Turnover penalties and dynamic execution models

Ziju Quant embeds execution drag directly as an endogenous constraint in portfolio optimization, applying non-linear impact penalties to incremental turnover to avoid uneconomic trading.

Signals are evaluated not just on directional accuracy, but on whether their information ratio decisively surpasses the marginal transaction cost of rebalancing.

Capacity ceilings and circuit breakers

A signal viable at modest scale often deteriorates as assets grow. As orders consume a larger share of average daily volume (ADV), market impact scales non-linearly.

Capacity ceilings and slippage thresholds must be specified during research. When realized slippage deviates beyond expected tolerance, the model triggers an automatic review rather than waiting for drawdown.

Testing hypotheses against real-world friction

Only strategies that withstand strict liquidity limits and harsh execution friction hold genuine empirical and investment value.