Bilal Khan
SPY put credit spread bot - sister to
Jeff Stubs - auto-refreshes every 5 min
Paper trading only. n is tiny. This is NOT validated performance and proves nothing about edge yet.
GREEN d1 off
+0.38%
$5,018.85 from $5,000 baseline
| completed trades: 0
| account: PA37E0WG08IG
Latest signal
[d1 OFF] Fri trading day, VIX 15.4 (calm), 5d chg -7.5% — enter
weekday=Fri | VIX=15.45 | 5d_chg=-0.0749 | should_enter=True | 2026-05-29T19:18:41
Recent trades
| Time (UTC) | Trade | PnL | Status |
|---|
| 2026-05-29T16:42:40 | entry 1 x SPY260601P00744000 | +$0.17 credit / -$1183.0 max | filled |
| 2026-05-27T16:11:57 | entry 1 x SPY260529P00738000 | +$0.26 credit / -$474.0 max | filled |
Strategy
UnderlyingSPY (Alpaca paper)
StructurePut credit spread, 2 DTE
ScheduleMon / Wed / Fri at market open
Short legAbout 10 delta
Long leg5 points below short (defined risk)
VIX filter (d1)Disabled at MVP. Re-enables at $15k equity or after 20%+ drawdown.
ExitHold to expiry. No active management.
Sizing1 spread per trade at the $5k tier
Edge status: UNPROVEN
The original backtest priced options with a flat-volatility model that
overstated the credit collected by roughly 2x (it ignored put skew on
the long leg). On the first real fill, the actual credit was about 45%
of what the backtest assumed. The headline Sharpe numbers that used to
sit here were therefore not trustworthy and have been removed.
Honest status: this is a short-volatility premium strategy that may or
may not have a real edge after real-world friction. It is now running on
paper purely to collect real fills. We will judge it on that real data
after roughly 20 trades, not on the backtest. The bot logs modeled vs
real credit on every trade for exactly this calibration.
Generated 2026-05-29T22:46:17.853783+00:00 UTC. Read-only. No login.
Data from committed bot logs, not a live broker call.