Calibrate Your Entries with Historical Backtests: The Cycle × OTM Matrix
Is 30% OTM Safe Enough?
This is the question I get asked most. People asking it want a number, but the honest answer is: it depends on how long you plan to hold.
The same 30% OTM cushion means completely different assignment odds when held for 1 week versus 4 weeks. People used to go by feel. Not anymore—the Historical Backtest module turns this question into a lookup table.
How to Read the Matrix
Open "Historical Backtest" inside "My Portfolio," and the first table is the cycle × OTM matrix:
- Horizontal axis: OTM width (20% / 25% / 30% / 35% / 40%)
- Vertical axis: holding period (1 week / 2 weeks / 4 weeks)
- Each cell: the historical non-assignment rate—the share of similar past positions that expired without assignment
Take SOXL as an example (equal-weighted mean):
| Period \ OTM | 20% | 25% | 30% | 35% | 40% |
|---|---|---|---|---|---|
| 1 week | 96.5% | 98.25% | 99.3% | 99.77% | 99.77% |
| 2 weeks | 92.16% | 95.09% | 97.43% | 98.36% | 99.53% |
| 4 weeks | 88.16% | 91.68% | 94.37% | 96.37% | 97.66% |
Reading it is straightforward: SOXL at 30% OTM held for 1 week has a historical non-assignment rate of 99.3%; held for 4 weeks, it drops to 94.37%. Every extra stretch of time you hold, you pay for it in risk.
A Complete Pre-Trade Calibration Routine
My current routine before opening a position:
- Step 1: Spot a contract on the Opportunities page; note its OTM width and days to expiry
- Step 2: Open the backtest matrix, find the matching cell, check the historical non-assignment rate
- Step 3: If the rate falls below my personal line (mine is 95%), either go deeper OTM or shorten the period
- Step 4: Switch to the "untouched during period" view and look again—expiring unassigned doesn't mean the ride was smooth
Step 4 matters. The non-assignment rate only tells you where price stood on expiration day; the untouched rate tells you how scary the journey was. The wider the gap between the two numbers, the bumpier the ride.
Nine-Underlying Tail Comparison: Know "How Bad Can It Get"
Below the matrix is a nine-underlying tail comparison table with three numbers: historical worst, the P1 tail (the average of the worst 1%), and the non-assignment rate at 30% OTM.
This table isn't for comparing returns—it's for understanding risk. All nine are leveraged ETFs, yet their tails differ wildly. For a detailed breakdown, see "Tail Profiles of Nine Leveraged ETFs" in the Data & insights series. For now, remember one line: different underlyings, same OTM width, completely different safety margins.
Three Things Not to Misread
- Non-assignment rate ≠ win rate. Avoiding assignment doesn't mean you made money—a mid-period premium drawdown may have stopped you out early
- Historical non-assignment rate ≠ future probability. It describes the past; it promises nothing about the future
- Untouched at close ≠ untouched intraday. Wick-throughs that recovered by the close are invisible to the matrix
The same disclaimer sits at the top of the page. Tools give you a frame of reference; the decision is always yours.
