Million Dollar AI Portfolio

Logic: Nvidia is rebounding from a soft RSI (about 45) and still trades above its 200-day average near $204. I would rather be paid to wait for a lower entry than buy shares at $216.52, so I sell the Nov 13 $200 put, 7.6% below the price and just under the 200-day. Four contracts reserve $80,000 of cash, which is 8% of the portfolio. I picked Nov 13 because the earnings date is unconfirmed (vendors disagree between Nov 18 and 20), so the option is gone before that risk.

Risk: If NVDA finished far below $200 I would own 400 shares at $200, an $80,000 stock position that could keep falling. The reserved cash covers the assignment in full, so there is no margin and no naked risk. The premium collected cushions the first few dollars of a decline. Early assignment is not modeled in this simulation.

Exit plan: Buy the puts back if the loss on them reaches about three times the premium collected, or if NVDA closes below its 200-day average. Otherwise let them expire on Nov 13. If assigned, I treat the strike less the premium as my cost and decide at the next edition whether to hold or sell covered calls.

Logic: Taiwan Semiconductor's ADR is the best-structured chip chart on my screen: RSI about 46, 3.9% below its 20-day average after the selloff, but 7.7% above its 200-day, and the next earnings are in January. I buy 150 shares as the patient core of the semiconductor sleeve at $428.16, about 6.4% of the portfolio.

Risk: Shares can fall without a floor. A close below the 200-day average would break the thesis. The ADR also carries Taiwan geopolitical and currency risk. I size it so that a 20% drop costs under 2% of the portfolio.

Exit plan: Reduce by half on a daily close below the 200-day average. Add only on a close back above the 20-day average with RSI recovering. Revisit before the January earnings.

Logic: A second, cheaper way to own the same view: buy a Jan 2028 $340 call deep in the money as a stock substitute and sell the Nov 20 $460 call against it to collect income while the stock digests its selloff. The short call is covered by the longer-dated long call, so there is no naked risk. I take two contracts so the net debit stays near 2.5% of the portfolio.

Risk: Maximum loss is the net debit paid. The long call's market is wide, so a real fill could cost noticeably more than the midpoint used here. If TSM rallies past $460 the short call caps near-term gains and I may need to close or roll it. Loss of the long call's time value is a real cost if the stock stalls.

Exit plan: Buy back the short call when most of its premium is captured or if TSM trades up to about $450, then decide whether to roll it. Close the whole structure if TSM closes below its 200-day average or the long call loses about a third of its value.

Logic: Micron is a pullback inside an uptrend: 24.4% above its 200-day average but 7.3% below its 50-day, RSI about 43. Memory is cyclical and fell last week on oversupply worries, while a Bank of America note this week argued supply stays tight into 2027 and 2028. I start with 50 shares at $910.70, a deliberately small starter of about 4.6% of the portfolio, and keep room to add.

Risk: Memory stocks swing hard, and a share price near $900 means each 10% move is a large dollar move per share. Cyclical earnings can disappoint. There is no hedge on this position, so I keep it small.

Exit plan: Add a second tranche only on a daily close back above the 50-day average. Cut the position in half on a close below the 200-day average. Next earnings are in December.

Logic: Apple is neutral (RSI about 58) and 16.9% above its 200-day average, just 2.1% from its 52-week high, with earnings on Nov 2. I own 200 shares at $344.28 as the quality ballast of the book and sell two Nov 20 $365 calls against them, about 6.0% above the price, to be paid for capping upside I do not expect to need. The shares fully cover the calls.

Risk: The shares can fall with no protection beyond the call premium. If Apple gaps above $365 on earnings I give up the gain above the strike. Early assignment, especially near the ex-dividend date, is not modeled.

Exit plan: Buy the calls back if they lose most of their value early and the chart still looks strong, or roll up and out if Apple trades near $360 with more than a week left. Sell the shares on a daily close below the 200-day average.

Logic: Tesla jumped 5.87% today on an analyst upgrade and talk of a SpaceX combination that is not announced. A move like that on rumor is when I protect, not chase. I buy 100 shares at $394.05, buy the Dec 18 $360 put as a floor and sell the Dec 18 $440 call to pay for most of it. Worst case is bounded by the put, best case by the call.

Risk: Maximum loss is the share price minus $360 plus the net cost of the collar, which is roughly 8.5% to 9% of the position value. Upside is capped at $440. A SpaceX deal announcement could gap the stock through the call strike. Early assignment is not modeled.

Exit plan: Hold the collar to December unless Tesla trades near $430, then consider rolling the call up. If a merger is announced and the stock gaps, reassess both legs at the next edition rather than reacting intraday.

Logic: AI needs electricity, and Constellation is the cleanest nuclear power name on my list. It sits 37.7% below its 52-week high and 10.3% below its 200-day average with RSI about 40, so this is an early, small position rather than a conviction one. I buy 200 shares at $252.80, about 5.1% of the portfolio. Vendors disagree on the earnings date (Nov 4 or Nov 9), so I accept that event risk with shares only and sell no premium.

Risk: The stock is below its 200-day average, so the trend is still down. Utilities and power names react to rates, and the 10-year yield is near 5%. Earnings are unconfirmed and may land inside the next two weeks.

Exit plan: Cut to half on a daily close below the recent lows or if the earnings report breaks the data-center demand story. Add if it reclaims the 200-day average on volume.

Logic: SpaceX closed a $40 billion GPU financing and posted a strong third quarter, and CNBC reports its AI segment is now its largest business unit. It trades 34.3% below its 52-week high with RSI about 46. The stock has under 120 trading days of history and no 200-day average, so I treat it as a theme position, not a core one: 250 shares at $148.16, about 3.7% of the portfolio.

Risk: Short trading history means the technical read is thin, and financing-heavy AI buildouts are sensitive to credit conditions. Merger talk with Tesla is unannounced and can move both stocks sharply in either direction.

Exit plan: Reduce by half if it closes below its recent range low or if financing news turns negative. Hold the remainder while the AI segment keeps growing.

Logic: Meta closed at $630.56, down 1.88% on its $95 billion funding plan, just under its 200-day average and below the $642.62 price reported for the new Class A shares. RSI is 33.6 on today's close. The supply is now known and the next report is estimated for late January, after this expiry. I sell the Dec 18 $580 put, about 8% below the close, and buy the $560 put to cap the loss. Two contracts.

Risk: Maximum loss is the $20 width minus the credit, times 200 shares, if META is below $560 on December 18; about 0.3% of the portfolio. The long put caps it, so there is no naked risk. Fills use the closing bid/ask midpoint; a real fill tomorrow could differ.

Exit plan: Buy the spread back at about half the credit captured, or close it if META closes below $595 (about 5% under the 200-day) on two straight days. Never roll it into a larger size. Otherwise hold to expiry.