Throughput Stats

Throughput Stats measures how a strategy spends time and capital — frequency, holding time, exposure, turnover, workable size — rather than whether it wins. A reference system generates the trade list the throughput stats family is measured on: a moving average cross or Donchian breakout, simulated single-position with fills at signal closes. It reads beside the rest of the validation toolkit.

How to Trade the Throughput Stats?

• Rolling Exposure %: the share of the recent window spent holding a position, the background tinted while one is open; above the alert level, capital is near continuously deployed.
• Turnover: notional traded over average equity, total and per year — the number behind commission drag; High Turnover marks cost-sensitive territory.
• Capacity / Order: the Participation Limit times typical traded value sampled at entries, versus the average fill; a breach flips the verdict and fires Capacity Exceeded. The verdict and the turnover and capacity alerts stay silent below Minimum Closed Trades.

Throughput Stats Settings

• Entry Rule (default Moving Average Cross; also Donchian Breakout) with Average Type (default EMA), Fast Length (default 20), Slow (default 50), Breakout Length (default 20).
• Trade Direction (default Long & Short): Long Only exits to flat instead of reversing.
• Initial Capital (default 10000) and Position Size (% Of Equity) (default 100): the account model behind turnover.
• Rolling Exposure Window (Bars) (default 200), Exposure Alert Level (%) (default 80), Turnover Alert Level (x / Year) (default 20), Minimum Closed Trades (default 10).
• Participation Limit (%) (default 5) and Liquidity Window (Bars) (default 100): the capacity estimate.
• Show Dashboard (default on) with Position (Top Right) and Size (Small), plus in-market tint and trade-marker toggles.

Frequently Asked Questions

Why measure throughput separately from expectancy?

Because they answer different questions. Expectancy prices the average trade; the throughput family reports how often that trade arrives, how long capital stays committed, and what it costs to run.

Does the choice of reference system matter?

Only as a generator. The measurements are rule-agnostic, so pick the rule closest to your own cadence and read the stats, not the simulated equity.

How seriously should I take the capacity figure?

As a first-pass estimate, not a promise. Capacity has no closed-form expression, so the build multiplies a participation ceiling by the typical traded value where entries occurred. Symbols without volume data report no capacity.