03 — Three-Gate Live Audit Standard (CLUE/1.0)
Before any candidate invariant discovered on historical ticks can be promoted into the live execution book, it must pass all Three Strict Audit Gates simultaneously.
---
Gate 1 — Out-of-Sample Walk-Forward Stability
An invariant cannot merely fit the full sample window; it must exhibit positive expectancy in both temporal halves and achieve a statistically significant Sharpe ratio:
$$\text{First-Half Expectancy } \mu_{1/2}^{(1)} > 0, \qquad \text{Second-Half Expectancy } \mu_{1/2}^{(2)} > 0, \qquad \text{Sharpe} \ge 1.0$$
---
Gate 2 — Execution Friction Stress Test (2.0x Slippage + Commission)
Every candidate invariant is re-simulated after doubling all bid/ask half-spread costs and brokerage/clearing fees:
$$\text{Net PnL}_{2\times\text{Friction}} = \text{Gross Edge} - 2.0 \times \left(\text{Half-Spread Slippage} + \text{Commission}\right) > \$0.00$$
| Venue Track | Standard 1x Friction | Stress-Tested 2x Friction | Audit Requirement |
| :--- | :--- | :--- | :--- |
| US Equity Options (SPY/QQQ/IWM) | 1x Option Spread + $0.65/contract | 2x Option Spread + $1.30/contract | Net PnL $> \$0.00$ |
| NYSE Arca FX ETF Options (FXE/FXY/FXB/UUP) | 1x ETF Option Spread + $0.65/contract | 2x ETF Option Spread + $1.30/contract | Net PnL $> \$0.00$ |
| Spot G10 Interbank FX (EUR/USD, USD/JPY, GBP/USD) | 0.8–1.2 pips round-trip | 1.6–2.4 pips round-trip | Net PnL $> \$0.00$ |
---
Gate 3 — Exact Permutation Null Destruction ($p < 0.05$)
To prove that the edge originates from the causal temporal ordering of the 4-symbol orbit rather than unconditional market drift, we shuffle the timestamps across $N = 200$ Monte-Carlo permutations while preserving the marginal return distribution:
$$p = \frac{1 + \#\left\{k \in \{1,\dots,200\} : \text{PnL}_{\text{perm}}^{(k)} \ge \text{PnL}_{\text{actual}}\right\}}{201} < 0.05$$
---