Performance terminal

📊 Weekly NQ Trade Desk Self-Grade

A compact self-audit of this week’s automated NQ calls. The system handled the early-week upside structure well, but was too slow to downgrade bullish bias as Friday’s selloff accelerated. Raw positioning details remain internal; positioning was checked as part of the process.

Journal coverage: 84 reports seen lifetime; 47 unique NQ reports evaluated in this weekly window; 175 graded horizon outcomes.

Setups graded

Unique weekly NQ reports; 175 graded horizons

A/B outcome rate

74.3%

130 of 175 graded horizon outcomes

Target-hit rate

47.4%

Directional target level reached before horizon close

Invalidation / stop-hit

23.4%

Opposing level touched across graded horizons

Directional accuracy

49.7%

Move direction aligned with call bias

Avg MFE

+60.6 pts

Average favorable excursion where available

Avg MAE

96.7 pts

Average adverse excursion where available

Avg R multiple

Missing

Risk unit / trade sizing not present in journal

Horizon Performance

Directional accuracy, target hits, invalidations, and excursions by forecast horizon. Labels include status text and do not rely on color alone.
Horizon	Graded	Directional Accuracy	Target Hit	Invalidation Hit	Avg MFE	Avg MAE	Assessment
30m	44	47.7%	40.9%	22.7%	+34.2 pts	48.4 pts	Mixed / noisy
60m	44	50.0%	45.5%	22.7%	+43.8 pts	69.0 pts	Acceptable but weak edge
120m	44	50.0%	50.0%	22.7%	+64.8 pts	117.5 pts	Best balanced horizon
240m	43	51.2%	53.5%	25.6%	+100.4 pts	153.0 pts	Highest target capture

Setup Types

Best and worst setup buckets from available fields.
Type	Outcomes	Target Hit	Avg Score
Best: Bull / stretched	132	58.3%	90.9
Bear / balanced	7	85.7%	88.4
Worst: Bull / balanced	36	0.0%	72.4

No-Trade Condition Quality

No explicit no-trade field exists in the journal. Using “balanced + weak edge / high reversal risk” as a proxy, the system did not abstain enough on Friday. Quality: Needs improvement.

Missing: formal no-trade labels, confidence score, stop distance, and R-multiple fields.

Data Completeness

Available: grade, bias, stretched flag, support/resistance hits, MFE/MAE, horizon moves. Missing: explicit setup taxonomy, R multiple, sizing, slippage, and trade execution outcomes. Pending: final Friday 240m windows.

What Worked

Longer horizons captured more of the weekly structure.
Early-week bullish/stretched calls were repeatedly rewarded.
The late Friday bearish flip graded well on available 30m–120m windows.

What Failed

Directional accuracy near coin-flip is not acceptable as a standalone edge.
Friday bullish calls remained active through a fast regime break.
Bull/balanced bucket produced zero target hits this week.

Adjustments for Next Week

Add a volatility/regime-break circuit breaker after large VWAP deviations.
Downgrade or no-trade bull/balanced setups until target-hit rate recovers.
Require explicit stop distance and R-multiple logging for every setup.
Score 120m/240m higher than 30m when horizons conflict.