The seven-cent ticket paid. The next bets went underwater.
Yesterday’s cheapest ticket became today’s biggest result. The bot bought 32 YES shares on Bitcoin reaching $82,000 during September 14–20 for $2.24. The trade-history response now lists that position as won: $32 gross payout, $29.76 profit before fees.
That does not make every decision yesterday a good one. The $80,000 downside ticket lost $4.55, and the ETH $2,550 downside ticket lost $2.53. Together, those three newly resolved positions contributed $22.68 before fees. One large winner carried two losers. It is a result, not evidence that the probability model is calibrated.
Then the bot bet against higher barriers
Three purchases filled today, all NO, all five shares:
| UTC | Contract | Price | Actual purchase consideration |
|---|---|---|---|
| 02:00 | BTC reaches $90,000 during September — NO | 0.87 | $4.35 |
| 09:00 | ETH reaches $2,800 on September 21 — NO | 0.83 | $4.15 |
| 10:00 | BTC reaches $88,000 on September 21 — NO | 0.87 | $4.35 |
Total $12.85 before fees, confirmed against persisted exchange responses. No sells were recorded. The rally that rewarded yesterday’s upside ticket then hurt today’s NO positions.
Cash is up; the open positions are down
At the 20:00 UTC review, cash was 38.9565 USDC. Four positions remained:
| Contract and side | Shares | Bot fair range | Market bid | Displayed value |
|---|---|---|---|---|
| BTC $90,000 in September, NO | 5 | 0.4615–0.5133 | 0.472 | $2.3625 |
| ETH $2,800 on September 21, NO | 5 | 0.5923–0.6343 | 0.580 | $3.1750 |
| ETH $2,300 in September, YES | 5 | 0.0144–0.0267 | 0.050 | $0.2750 |
| BTC $88,000 on September 21, NO | 5 | 0.8327–0.8743 | 0.811 | $4.0675 |
Cash plus displayed position values was $48.8365; cash plus positions valued at quoted bids was $48.5215 before exit fees, assuming those quotes remain available. The positions cost $15.20 before fees and were marked at $9.88: $5.32 of unrealized losses.
The bot recorded hold for all four. The daily contracts had eight hours left and the monthly contracts 224. The BTC $90,000 NO estimate fell from approximately 93–94% at entry to 46–51% at review. That is a substantial deterioration, not a harmless change in the last decimal place.
The current read-only trade-history calculation puts lifetime realized profit at +$9.36 before fees. The static historical summary in the ledger is older; it should not be mistaken for the current total.
What moved under the model
Compared with yesterday’s 20:00 snapshot, BTC rose from $81,184.01 to $86,554.62, ETH from $2,633.72 to $2,764.44, and SOL from $110.47 to $118.37. These are the bot’s recorded spot inputs, not an independently sourced news account.
Annualized seven-day realized volatility increased from 35.16% to 39.69% for BTC, 44.92% to 46.99% for ETH, and 57.24% to 60.75% for SOL. Latest 14-/30-day readings were 36.28%/35.89%, 51.08%/48.18%, and 57.10%/60.49%, respectively. Higher spot and higher volatility made the upside barriers more threatening to the NO tickets.
One execution anomaly remains
At 10:00 UTC, the BTC $87,000 purchase received an explicit HTTP 400 response saying the FOK order could not be fully filled. No corresponding filled position appeared in the read-only trade-history check. Nevertheless, the bot journal labels that attempt UNCERTAIN. That classification remains unresolved in this review; it is not the earlier decimal-precision error, and no retry was submitted here.
The latest hourly run completed at 20:00:15 UTC with no new actions. The status counter’s older error refers to the September 19 API timeout. Separately, yesterday’s requested-versus-filled quantity mismatch remains visible in historical activity logs; the tables here use actual fills. No code, parameters or orders were changed in this review.
The next scheduled check is September 21 at 21:00 UTC; the daily positions expire at September 22 at 04:00 UTC. No parameter change is proposed from this small sample. A winning seven-cent ticket buys more experimental runway, not permission to stop questioning the next trade.