If price opens between the previous close and the trigger level, and holds there, what are the historical probabilities for selling credit spreads at the opposing ATR levels? We measured how often price does NOT reach each strike across 25 years of SPY data.
Bearish trigger box: Price opens below PDC but above the put trigger (-23.6%). You sell a call credit spread at a higher ATR level. You win if price stays down.
Bullish trigger box: Price opens above PDC but below the call trigger (+23.6%). You sell a put credit spread at a lower ATR level. You win if price stays up.
The table shows the % of days price did NOT reach that level — i.e., your win rate if your short strike was at that ATR level and expired that day.
| Condition | N | +38.2% | +50% | +61.8% | +100% |
|---|---|---|---|---|---|
| All box days | 1,483 | 66.4% · No | 76.9% | 84.5% · No | 96.2% · No |
| Held 30 min | 875 | 79.8% | 87.0% | 91.0% | 97.9% |
| Held 1 hour | 673 | 85.7% | 91.2% | 93.8% | 98.7% |
| Held 30m, top half | 342 | 77.5% | 84.5% | 90.4% | 98.0% |
| Held 30m, bottom half | 533 | 81.2% | 88.6% | 91.4% | 97.9% |
| Held 1h, top half | 242 | 86.8% | 91.7% | 94.2% | 98.3% |
| Held 1h, bottom half | 431 | 85.2% | 91.0% | 93.5% | 98.8% |
| Condition | N | -38.2% | -50% | -61.8% | -100% |
|---|---|---|---|---|---|
| All box days | 1,725 | 64.7% · No | 73.6% | 79.8% · No | 92.3% · No |
| Held 30 min | 1,004 | 76.2% | 82.6% | 87.7% | 95.9% |
| Held 1 hour | 776 | 82.9% | 88.0% | 92.1% | 97.6% |
| Held 30m, top half | 618 | 74.9% | 81.7% | 87.7% | 95.8% |
| Held 30m, bottom half | 386 | 78.2% | 83.9% | 87.8% | 96.1% |
| Held 1h, top half | 485 | 80.6% | 86.2% | 90.7% | 96.9% |
| Held 1h, bottom half | 291 | 86.6% | 91.1% | 94.5% | 98.6% |
Consider two approaches based on risk tolerance. The ±100% ATR level offers 97–99% historical win rates with minimal risk, though the credit collected is smaller. The ±61.8% level offers 88–93% rates with more premium but a worse loss-to-profit ratio when it goes wrong. A 30-minute hold tends to be the minimum filter worth waiting for.
Consider the opposing 38.2% as a potential stop. If price reaches it, the trigger box thesis may be weakening. This gives defined risk at entry. Keep in mind that the loss on a single losing trade can be several times the credit collected — size positions accordingly.
Hold time appears to matter more than box position. Top vs bottom half of the box changes historical win rates by 1–4%, while the 30-minute hold filter is associated with 10–15% higher rates. Time in the box may be the more useful signal.
This setup fires ~10% of trading days. The trigger box open + 1-hour hold occurs on roughly 1 in 10 days. It's not an every-day trade — it's a selective, high-probability setup.