Backtest Case Study · Jun 2025–Jun 2026 · QuantConnect Cloud · NVDA Weeklies

The Premium-Selling SweepNine structures on NVDA — every delta, all the data

A complete head-to-head of weekly covered-call, put-write, and wheel structures on NVDA over a +53% bull run. No single “winner” — the answer depends on which metric you optimize and whether you start from cash or already own the shares. Every number below ties to the QuantConnect end-of-algorithm log.

Most Profit
+$6,652
1 strike ITM covered call
Best Sharpe
2.10
2 strike ITM covered call
Best Sortino
1.43
30-delta / 30-delta wheel
Benchmark
+$7,432
NVDA buy & hold

The Equity Curves

Top structures vs buy-and-hold vs the real NVDA price line · 52 weeks
Cumulative P&L & NVDA Price — Jun 2025 → Jun 2026
per 100-share lot · mid fills · NVDA $137.66 → $210.69 (+53%)
$0$2k$4k$6k$140$170$200$23006/2508/2509/2510/2511/2512/2501/2602/2603/2605/2606/26 Left axis: cumulative P&L  |  Right axis: NVDA share price
1 strike ITM Covered Call (+$6,652) 30/30 Wheel (+$6,000) Buy & Hold (+$7,432) NVDA share price (right axis)
The dark-red line is the actual NVDA close each week — included so the equity curves can be read against the real move that drove them. Every option structure trailed buy-and-hold because the stock rose 53%; that is the documented behaviour of premium selling in a strong bull market, not a defect.

Every Test, Ranked

All nine structures · sorted by final P&L · ★ = most profit
StructureFinal P&LMax DDSharpeSortinoPremiumAssign
NVDA Buy & Hold+$7,432$-3,5891.131.35
1 strike ITM Covered Call ★ +$6,652 $-1,310 1.84 1.21 $20,570
2 strike ITM Covered Call +$6,553 $-1,285 2.1 1.32 $26,967
OTM Wheel (1 strike below) +$6,396 $-1,488 1.83 1.11 $13,870 0%
30-delta / 30-delta Wheel +$6,000 $-1,436 1.97 1.43 $9,458 23%
30d put / 70d call (basis floor) +$5,768 $-2,444 1.81 1.47 $8,492 21%
50/50 ATM Wheel (CBOE replica) +$5,449 $-1,729 1.62 1.31 $20,513 45%
ATM Put-Write +$4,936 $-1,736 1.38 1.12 $21,251 46%
30d put / 70d call (no floor) +$4,807 $-1,595 1.88 1.19 $16,633 26%
2 strike ITM Put-Write +$3,706 $-3,196 0.82 0.75 $33,826 71%

Sharpe & Sortino are annualized from weekly P&L changes on ~$13,200 deployed capital per lot (52-week √ scaling). Max DD and premium are in dollars. Assignment = put assignments / puts sold (or 100% for always-assigned covered calls).

Three Different Winners

The honest answer: it depends what you optimize
MOST PROFIT
1 strike ITM Covered Call — +$6,652
If you already own 100 NVDA shares and want maximum dollars, this is the answer. Deep enough to be assigned every week, capturing steady intrinsic + premium. Lowest drawdown too ($-1,310). Requires owning the stock first — you carry full downside.
BEST SHARPE
2 strike ITM Covered Call — Sharpe 2.10
Slightly less profit (+$6,553) but the steadiest week-to-week returns of any structure. Smallest drawdown of all ($-1,285). The pure volatility-adjusted winner.
BEST SORTINO + CASH START 30-delta / 30-delta Wheel — Sortino 1.43
The best downside-adjusted return, and the only top structure that starts from cash rather than requiring you to already own NVDA. Assigns just 21% of the time, entering shares below market. This is the entry strategy; the covered calls are overlays on stock you already hold.

Exact Trade Details

The precise rule set for every structure tested
StructureRule
NVDA Buy & HoldBuy 100 shares at open, hold to end
1 strike ITM Covered CallOwn 100 sh; sell weekly call 1 strike ITM; ~5 DTE; let assign
2 strike ITM Covered CallOwn 100 sh; sell weekly call 2 strikes ITM; ~5 DTE
OTM Wheel (1 strike below)Cash; sell put 1 strike OTM; CC recovery if assigned
30-delta / 30-delta WheelCash; sell 30-delta put; if assigned sell 30-delta call
30d put / 70d call (basis floor)Cash; 30-delta put; if assigned 70-delta call kept >= cost basis
50/50 ATM Wheel (CBOE replica)Cash; sell ATM put; if assigned ATM call
ATM Put-WriteCash; sell ATM (50-delta) put weekly; CC if assigned
30d put / 70d call (no floor)Cash; 30-delta put; if assigned 70-delta call, no basis floor
2 strike ITM Put-WriteCash; sell put 2 strikes ITM; CC 1-ITM if assigned
Strike spacing note: NVDA's weekly grid is $1.00 wide below $150 and $2.50 wide at/above $150. “N strikes ITM” therefore means a different dollar-distance at different prices, which silently shifted aggressiveness as NVDA rose. The delta-specified structures (30d, ATM/50d) are immune to this and are the more robust way to define strikes.

What The Sweep Proved

Findings that hold across all nine tests

1. Moneyness creates no edge

By put-call parity a put at strike K equals a covered call at strike K; premium collected is offset by the assignment haircut at every strike. That is why all nine clustered in a band — the strike dial moves risk around, not return.

2. Conservative beat aggressive

The best structures were the simplest: passive ITM covered calls and the plain 30/30 wheel. Every aggressive refinement — deep-ITM puts, 70-delta calls, dropping the basis floor — finished lower and drew down more.

3. Delta-targeting beat the CBOE ATM benchmark

The 50/50 ATM wheel (≈ CBOE WPUT methodology) returned +$5,449; the 30/30 delta wheel returned +$6,000 with a higher Sharpe. The active strike selection the passive indices lack added ~$551 and better risk-adjusted returns.

Tested & Rejected

The refinements that lost
REJECTED
70-delta call (parity reconstruction)
Floored: basis rule overrode all 23 calls (+$5,768). Unfloored: booked below-basis losses 13 times, worst drawdown of the wheels (+$4,807). Elegant theory, collides with never-sell-below-basis on a dropped stock.
REJECTED
Deep-ITM & ATM puts
2-strike-ITM put: $33k premium but pure intrinsic churn, +$3,706, Sharpe 0.82, worst drawdown ($-3,196). ATM put-write: +$4,936, Sharpe 1.38. Both assigned ~100% at bad prices.

Honest Limitations

What this test does and does not prove

Single underlying, one ~13-month window, one strongly trending regime — the exact regime where premium selling is documented to lose. Fills modeled at backtest mid; single-stock weekly spreads are wider than index options, so live results run modestly below these figures (most for the deep-ITM/high-delta legs). Only ~$13,200 of a $100k account was deployed per lot; idle-collateral yield and multi-lot sizing are the dominant unaddressed levers. The decisive untested case is a flat or range-bound underlying, where these strategies should beat buy-and-hold rather than trail it.