Run a search for covered call returns and you'll find backtests promising 12%, 15%, even 20% a year. Ours come in lower. On Apple, our engine generated 6.17% annualized over 2.8 years of historical data. A backtest that ignored one specific rule would have shown 8.74% on the exact same stock, over the exact same period.
That gap isn't a flaw in our numbers. It's the single most important thing we do — and it's the reason you can trust the rest of the figures.
The One Rule That Explains the Gap
The rule is the earnings blackout: the engine refuses to sell a covered call in the two weeks before a company reports earnings. Earnings is the one scheduled event that can gap a stock 10% overnight. Sell a call into that window and you're collecting a fat premium precisely because the market knows a violent move is coming. Most of the time the move goes against the call seller — the stock jumps past your strike and your shares get called away below where they're suddenly trading, or it craters and the premium never covered the drop.
Almost every published backtest sells through earnings. That's where the inflated yields come from. They're booking the rich pre-earnings premium without pricing in the gap risk that premium exists to compensate for. We took that income off the table on purpose. (For the full reasoning, see the one week you don't sell and why it's worth it.)
What the Blackout Costs, Stock by Stock
Here's the same comparison across five profiles, with the earnings blackout on versus off:
- Apple: 6.17% with the blackout, 8.74% without — a 29% haircut to yield.
- Microsoft: 7.74% versus 8.47% — just a 9% haircut.
- The $100K Moderate portfolio: 8.10% versus 10.28% — a 21% haircut.
- The $50K energy-sector portfolio: 4.53% versus 10.50% — a brutal 57% haircut.
- NVIDIA: 3.29% with the blackout, 0.88% without.
Read that last line again. On NVIDIA, skipping earnings didn't cost yield — it nearly quadrupled it.
Why Microsoft Barely Notices and Energy Pays Dearly
The cost of the blackout isn't uniform, and the spread tells you something useful about your own holdings.
Microsoft loses almost nothing. It's a stable, liquid name with moderate implied volatility (IV) — the market's expectation of how much the stock will move. Its earnings moves are usually contained, so the premiums available outside the blackout window are nearly as good as the ones inside it. Sitting out a couple of weeks per quarter barely dents the year.
The energy portfolio is the opposite. Stack several oil and gas names together and you stack their earnings dates too. Every quarter, each holding goes dark for two weeks, and those windows overlap badly. The blackout knocks 57% off the headline yield — not because the rule is harsh, but because earnings-driven premium was doing most of the heavy lifting in a sector where the calls only pay well when a move is brewing.
The NVIDIA Case: When Safety Pays You
NVIDIA breaks the whole "cost" framing, and it's worth slowing down on.
Without the blackout, the engine sold calls into NVIDIA's pre-earnings volatility, where premiums look irresistible. Then earnings hit, the stock gapped, and the shares got called away on the spike. The engine couldn't re-enter at a sensible price afterward, so the position sat in cash — out of the game. The result over 2.8 years: 0.88% annualized, from just two trades. The fat premiums were a trap.
With the blackout on, the engine stayed out of those danger windows, kept its shares, and sold calls in the calmer stretches between reports. That discipline let it make ten trades instead of two and stay continuously covered. The result: 3.29% — a 274% improvement over the no-blackout version.
For a volatile name, avoiding earnings isn't an insurance premium you pay. It's the strategy working as designed. The "expensive" pre-earnings calls were the ones quietly destroying the position.
So Why Show You the Lower Number?
Because the lower number is the one you'd actually live. A backtest that sells through earnings is describing a strategy nobody disciplined would run — it's selling the riskiest calls of the quarter, every quarter, and reporting the average as if the gap risk never shows up. Over a long enough window, it always shows up.
When we say a $100K Moderate portfolio produced about 8% annualized, that figure already has the earnings risk subtracted. It's not the best case. It's the realistic case, which is the only case worth planning around. (See the honest answer to how much you can make for the full income picture, and the strategy comparison for how Conservative, Moderate, and Aggressive settings change it.)
Here's the honest pitch: we leave 10-30% of potential yield on the table so you never sell into an earnings gap. For most stocks, that's a modest insurance premium. For volatile names like NVIDIA, it's not even a cost — it's a benefit.
Curious what your own holdings would have done? Try it — the free estimator runs these numbers on your actual portfolio, with the earnings blackout already built in. No login required.
Frequently Asked Questions
Why are your covered call yields lower than other backtests I've seen?
Because we exclude trades in the two weeks before earnings, and most published backtests don't. Selling calls into earnings inflates the premium you collect but exposes you to overnight gap risk that the premium exists to compensate for. Our numbers already subtract that risk, so they reflect what a disciplined strategy actually returns rather than a best case that ignores the worst event on the calendar.
Does skipping earnings always reduce returns?
No. For stable names like Microsoft the cost is small (about 9%), and for highly volatile names like NVIDIA, skipping earnings actually raised the backtested yield — from 0.88% to 3.29% — because it kept the position from getting blown out on the earnings gap. The blackout's "cost" is largest for stocks and sectors where earnings-driven premium was carrying the returns in the first place.
How much yield does the earnings blackout typically cost?
Across our profiles, most portfolios gave up roughly 9% to 29% of their potential yield to the blackout. Energy-heavy portfolios paid the most — about 57% — because clustered quarterly earnings dates block many of the weeks the strategy would otherwise trade. Volatile single names can actually gain yield.
Is a higher backtested return always better?
Not if it was earned by taking risks the strategy is designed to avoid. A backtest that sells through earnings reports a higher number by booking the very trades most likely to end in a painful assignment or a missed recovery. The realistic figure — earnings excluded — is the one to plan around.
The lowest honest number beats the highest hopeful one. Every figure in this series already has the earnings risk taken out, which is exactly why you can build a plan on them.
Methodology
These results come from running the Income Factory recommendation engine against 2.8 years of historical option chain data (ORATS, 2023-2025) with all defensive features active: earnings blackouts (14 days before earnings), buy-to-close orders at a 50% profit target, a 21-day minimum days-to-expiration (DTE) floor, and per-stock rebuy thresholds (10%/12%/15% for stable, moderate, and volatile names). Strategy: Moderate, a 0.25 delta target. The "without blackout" figures come from re-running the identical engine with only the earnings blackout disabled. Results are backtested and simulated — not actual trading. Past performance does not guarantee future results.