Tail Profiles of Nine Leveraged ETFs: What Happened in the Worst Months
The Data First
When the backtest module went live, the first thing I did was pull the tail data for all nine major leveraged ETFs. Here's what came back:
| Symbol | Type | Historical Worst | P1 Tail | Non-Assignment Rate (30% OTM) |
|---|---|---|---|---|
| KORU | South Korea · 3× | -55.62% | -23.78% | 99.14% |
| SOXL | Semiconductors · 3× | -50.78% | -27.47% | 99.3% |
| TECL | Technology · 3× | -45.73% | -20.49% | 99.67% |
| TQQQ | Nasdaq-100 · 3× | -37.53% | -20.6% | 99.88% |
| SPXL | S&P 500 · 3× | -43.83% | -19.47% | 99.68% |
| UDOW | Dow 30 · 3× | -48.73% | -17.12% | 99.65% |
| DUSL | Industrials · 3× | -49.2% | -19.79% | 99.38% |
| TNA | Small Caps · 3× | -49.75% | -23.62% | 99.68% |
| RAM | DRAM · 2× | -32.44% | -32.41% | 66.67% |
Sell Put earns premium—small money, trade by trade. What it loses is the tail—one event can be a big loss. So look at the tail first, then talk about returns.
Three Findings Worth Remembering
1. Every historical worst is below -37%. None of the nine is spared. 3× leverage doesn't mean "3× the gains"—it means 3× the swings: if the underlying index drops 15%, the ETF prints -45%.
2. The P1 tail is gentler than the historical worst, but more realistic. The historical worst is a single extreme event; the P1 tail is the average of the worst 1% of days. KORU's P1 is -23.78%—meaning on its "bad days," it loses nearly a quarter on average.
3. Non-assignment rate doesn't track tail size. KORU has the scariest tail, yet its 1-week non-assignment rate at 30% OTM is still 99.14%. That's exactly what the OTM cushion is designed for—a 30% distance absorbs the vast majority of tail shocks.
Why the RAM Row Stands Out
RAM's non-assignment rate at 30% OTM is only 66.67%—not even in the same league as the other eight.
Two reasons: first, it's 2× leveraged rather than 3×, so its volatility structure differs; second, its sample size is likely insufficient—the backtest page flags this too. Statistics from small samples should be taken with a discount.
That's also a general principle: when you see anomalous data, ask about sample size first, then causes.
Three Takeaways for Opening Positions
- OTM width is the antidote to tail risk. The bigger an underlying's tail, the deeper your OTM should go
- Shorter periods mean less tail exposure. The 1-week non-assignment rate generally runs more than 5 points above the 4-week rate
- Don't loosen your OTM standard just because the premium is high—premium is often high precisely because the tail is fat
