What a c8alpha pick looks like.

Five real picks from the system, fully detailed: the signal that fired, the LLM-generated thesis, the conviction score, the validated exit rule, and the actual outcome compared against the S&P 500 over the same holding window. Four winners and one loss — chosen deliberately, because single-side curation isn't honest.

AEVA
Unknown · Automotive
Conviction
MED
Uoa Slow Build · signal fired 2026-05-11

AEVA uoa_slow_build: 4d LEAP streak (62% conv, 4.6x rvol), drift PF 4.37, 5d post-ER

AEVA shows a 4-day LEAP-tier accumulation streak with 62% conviction and 4.6x avg rvol, supported by drift PF 4.37 (medium_validated cell, OOS n unknown). Signal is 5 days post-earnings with 80 days to next print. Conviction capped at MED due to medium_validated cell strength (unknown OOS sample size).

Entry2026-05-11 @ —
Exit2026-05-25 @ $25.32
Return+63.04%
S&P 500 same window+1.53%
Days held14
Alpha+61.51%
Exit rule: fade_window
PSNL
Unknown · Health Insurance
Conviction
MED
Uoa Slow Build · signal fired 2026-05-12

LEAP-tier 3d streak (37% conv, 12.4x rvol, 85% calls) on drift PF 4.37; 5d post-ER, 85d to next print

uoa_slow_build shows drift PF 4.37 with medium_validated cell strength (OOS n unknown). PSNL exhibits LEAP-tier signal: 3d streak at 37% conviction, 12.4x avg rvol, 85% calls with 57% LEAP concentration, and trend acceleration from 1.6x → 3.7x → 31.8x. Signal fires 5 days post-earnings with 85 days to next print, providing clean runway for thesis development.

Entry2026-05-12 @ —
Exit2026-05-25 @ $8.47
Return+40.23%
S&P 500 same window+1.68%
Days held13
Alpha+38.55%
Exit rule: fade_window
ORBS
Unknown · Unknown
Conviction
MED
Institutional Oi Build · signal fired 2026-05-19

Institutional OI build (3d streak, 59% LEAPS, 1.47% OTM weighted) on medium-validated cell (PF 2.18, n=160, p=0.003)

The institutional_oi_build play shows medium-validated evidence with PF 2.18 across 160 OOS observations (perm_p=0.003). Signal features include a 3-day streak with 59% LEAPS concentration and 1.47% OTM-weighted positioning, scoring 11/17 on the feature checklist. Conviction capped at MED due to playground tier and medium_validated cell strength (n_oos=160, PF=2.18).

Entry2026-05-19 @ —
Exit2026-05-25 @ $0.94
Return+29.41%
S&P 500 same window+2.30%
Days held6
Alpha+27.11%
Exit rule: fade_window
BLDP
Unknown · Unknown
Conviction
MED
Uoa Smart Divergence · signal fired 2026-05-15

BLDP uoa_smart_divergence: 4-day accum streak, 96% calls, 34x vol—medium-validated cell (PF 1.80, n=1303)

BLDP shows unusual options activity with a 4-day accumulation streak, 96% call bias, and 34x volume ratio on the uoa_smart_divergence play (PF 1.80, OOS n=1303, perm_p=0.002). The cell is medium-validated with 1303 out-of-sample observations, supporting moderate conviction. Playground tier caps conviction at MED; 10 days post-earnings with next print 88 days out provides a clean runway for the thesis to develop.

Entry2026-05-15 @ —
Exit2026-05-25 @ $5.54
Return+24.49%
S&P 500 same window+1.54%
Days held10
Alpha+22.95%
Exit rule: fade_window
ASTE
Unknown · Industrial Machinery
Conviction
MED
Uoa Slow Build · signal fired 2026-05-11

LEAP-tier 3d streak, 93% calls, 21.8x avg rvol; drift PF 4.37 (medium_validated cell)

ASTE shows LEAP-tier unusual options activity with a 3-day streak at 21.8x average relative volume and 93% call-weighted flow, including 33% LEAPS. The uoa_slow_build play shows drift PF 4.37 in a medium_validated cell (OOS n unknown, capping conviction at MED). Signal 5 days post-earnings with 86 days to next print provides clean runway for the thesis to develop.

Entry2026-05-11 @ —
Exit2026-05-25 @ $48.58
Return-8.01%
S&P 500 same window+1.53%
Days held14
Alpha-9.54%
Exit rule: fade_window
In retrospect
The mechanical signal was clean — 21.8x average RVOL with 93% call-weighted flow and 33% LEAP concentration is exactly what uoa_slow_build is designed to detect. The market didn't validate the institutional thesis within the 14-day hold window; the position closed at -8.01%. We publish the loss because not every signal works, and single outcomes — winners or losers — don't validate or invalidate the play's mechanical edge. The validated drift reference holds across all 130 observations, not any one of them.