Covered Calls for Premium Harvest — Quantum, AI, and the High-Beta End

This is Part 2 of two. Part 1 covers the income end — Dividend Aristocrats, 10% OTM 6-month strikes, ~7% combined annualised yield. This is the opposite spectrum: high-beta names where the option chain pays 5–10% of the underlying for a single 30-day call. The premium is real. The realised-vol path that earned it is what most retail option-sellers underestimate.

Why high-beta names pay so much

The premium on a call option is roughly proportional to implied volatility × √(time to expiry) × strike-distance scaling. On a low-vol name like KO with 17% IV, a 30-day 5% OTM call earns about 0.5–0.8% of the underlying. On a 90% IV name — common range for quantum-computing pure-plays (IONQ, RGTI, QBTS, QUBT) and recent AI IPOs in their first-year volatility window — the same 30-day 5% OTM call can earn 5–8% of the underlying. The chain pays you ten times the premium because the market judges the move-probability ten times higher.

That premium is, mathematically, fair compensation. Black-Scholes says so. The question is whether you are equipped — capital-wise, emotionally, position-management-wise — to survive the 60% drawdowns and 80% rips that earn that premium over a 12-month sample of realised paths.

Concrete math — a high-beta name at $25, 90% IV

Assume a quantum-computing pure-play trades at $25 with 90% implied volatility. The 30-day 10% OTM strike is $27.50. Premium on a chain that vol: $1.50–$2.00. Use $1.75 as a working midpoint.

Outcome	Stock at expiry	Premium kept	Capital P&L	Total return (1m)
Flat / mild drift	$25	$1.75	$0.00	+7.0%
Modest rise (within cap)	$27	$1.75	+$2.00	+15.0%
Assigned at strike	$27.50+	$1.75	+$2.50	+17.0%
Sharp rip	$35	$1.75	+$2.50 (capped)	+17.0% (vs +40% unhedged)
Sharp drawdown	$17	$1.75	-$8.00	-25.0%

Two numbers from that table deserve special attention. The +7% base case in a single month annualises to ~85% on the option overlay alone, which is why investors keep being drawn back to this trade — the math on paper is intoxicating. The -25% drawdown case is also a single month, and the $1.75 premium is a paltry 7% cushion against an 8-point downward gap. On a high-beta quantum name a -25% month is not a tail event — it’s a roughly-quarterly occurrence in their first 2 years of trading.

The premium is fairly priced for the realised vol. The 85% annualised return only happens if the underlying spends most of its days in the no-assignment, no-drawdown zone — which historically across the quantum cohort has been about 55–65% of monthly windows. The other 35–45% kill the return. The Money Temperature reading helps frame whether realised vol is likely to compress (favours the strategy) or expand (favours just owning or just selling) from here.

Rolling — the part nobody teaches

The covered-call mechanics in Part 1 assumed the call gets held to expiry. On a high-beta name, that’s often a mistake. Three rolling moves are worth knowing:

Rolling up and out

If the underlying runs into your strike with two weeks left, the call is now in the money. Closing it costs more than the premium you collected. Two paths: let assignment happen (cap captured, position closed at strike + premium) — or roll. The roll: buy back the current call (cost: intrinsic value + time value), simultaneously sell a new call at a higher strike and a later expiry. If the new sale exceeds the buy-back, it’s a credit roll: you stay long the underlying with a higher cap and another month of premium. If the new sale is less than the buy-back, it’s a debit roll: you’re paying to defer assignment, usually not worth it.

Rule of thumb: a credit roll up-and-out works when implied vol is still high enough on the higher strike. On a quantum name that just ripped 15%, the chain often pays even more on the new strike because the rip itself signals vol — credit rolls are usually available.

Rolling down and out

If the underlying drops fast and the call is now deep out of the money — say strike is $27.50, stock is at $20 — the original premium has decayed to near zero. The call is no longer earning. The roll: close the original (cheap), open a new call closer to the new spot, say $22 strike, 30 days out. The new chain at the lower spot still has elevated IV (drawdowns inflate vol), so the new premium often beats what the original would have earned.

This is the move that turns a -25% drawdown month into a -18% drawdown month — premium kept compounding even as the stock fell. It doesn’t save the trade, but it cushions it.

The wheel

If the call gets assigned, you’re now in cash. The wheel: immediately sell a cash-secured put at a strike below where you just sold (say the stock is back at $26 after assignment at $27.50, sell the $24 put). If the put gets assigned, you’re back in the underlying at $24, with the put premium in hand. From there, sell another covered call. Rinse, repeat. The wheel is the strategy’s natural full cycle: collect premium, sell at strike, buy back lower, repeat. On a high-beta name that swings widely, the wheel can produce sustained income from sideways-to-volatile underlying. See the cash-secured put entry for the put-side mechanics.

What can go wrong

Five failure modes, listed in roughly the order they’ll bite you:

• Sharp upside breakout. A quantum name announces a partnership with a hyperscaler, gaps 50%. Your $27.50 call is now worth $7.50 of intrinsic value alone, vs the $1.75 you collected. You either accept assignment (capped at $29.25 effective) or roll at a heavy debit. Either way you miss the run from $27.50 to wherever it stabilises ($35? $50?). The math says the premium pre-priced this — that’s cold comfort when you watch IONQ or QBTS run 80% on a positive press release.
• Sharp downside. Same name announces an SEC investigation, drops 35%. Your $25 stock is now $16.25. You kept the $1.75 premium (cushion: 7%) — net position is down 28%. The cap-loss-with-premium math protects nothing on a leg like that. This is the failure mode that wipes out a year of premium harvest in a single month.
• Earnings vol crush. Implied vol on these names goes parabolic into earnings — sometimes 150%+ IV the week before. Writing a covered call right then earns enormous premium. Earnings prints, the gap happens (either direction), implied vol collapses 40 points in an hour. If you sold the call AFTER the vol crush, you didn’t harvest the rich premium. If you sold BEFORE, the gap usually overwhelmed the premium. The earnings window is the worst-EV section of the cycle for this trade.
• Liquidity. Many quantum and small-cap AI names have thin option chains. Bid/ask spreads of $0.30 on a $1.75 premium mean a 17% transaction cost on entry-and-exit round-trip. Rolling becomes prohibitively expensive. Only write covered calls on names with at least 100 contracts of open interest at the strike you’re targeting.
• Tax friction. Premium received is short-term ordinary income in most jurisdictions — top-bracket federal + state in the US can chop the headline 85% annualised down to ~50% net. In Germany this lands at ~26% Abgeltungsteuer + solidarity surcharge. Don’t model the gross premium; model the net.

When the strategy works / when it doesn’t

Honest decision framework, four checks:

1. You already wanted to own the stock. If you wouldn’t buy IONQ at $25 today as a thesis-driven long, don’t buy it just to write the call. The premium isn’t enough to make a bad thesis good.
2. Position size is set by the downside, not the premium. If a 50% drawdown in this position would damage the overall portfolio more than a normal stock-picking mistake would, the position is too big. Premium income should be a rounding error in absolute terms — single-digit percent of portfolio NAV, not double-digit. See the Kelly criterion entry for the sizing math on asymmetric-payoff positions.
3. You have rolling discipline and a plan. Don’t enter without a written exit / roll-up / roll-down / wheel decision tree. The math punishes hesitation more than it punishes the wrong choice.
4. Implied vol is in the top quartile of its 12-month range. Selling premium at the 25th percentile of IV is selling cheap insurance against a known volatile name — bad EV. Selling at the 75th+ percentile is selling expensive insurance — that’s the trade.

If three of those four are true, the harvest variant can earn its keep. If two or fewer, the income-end strategy from Part 1 is the safer place for the same capital.

How Closelook’s Derivatives portfolio treats this end

The Derivatives reference portfolio writes covered calls predominantly on the income end — large-caps with stable cash flows. The high-beta harvest variant appears in the portfolio only as a deliberate, sized-down sleeve, usually capped at 10–15% of portfolio NAV in total premium exposure. The reasoning: the math works on paper across hundreds of trades, but most retail investors can’t execute hundreds of trades. They execute five, hit one bad earnings gap, and abandon the strategy at the worst possible point in the realised-vol cycle. Defensive sizing is the only thing that lets you stay in long enough to capture the cross-sectional average.

One question worth sitting with

The harvest variant is, structurally, selling insurance against catastrophe on names that occasionally have catastrophes. Every premium you collect is the market handing you cash in exchange for accepting tail risk. Over many trials the EV is positive. Over any individual trial the path can be brutal. The right question isn’t “is the premium enough” (the chain says yes, almost always). The right question is whether you can survive the path required to earn the premium over a 24-month horizon, including the 3–4 months in that window when the strategy will give back most of what it earned. If the answer is no, the income-end Part 1 strategy is where the capital belongs.