Silicon Valley GirlEx-Rocket Scientist: The Secret to Millionaires' Investment Portfolios
CHAPTERS
Meet Alex: engineer-turned-investor living off his portfolio
Marina introduces Alex, an ex-rocket-science/engineering type who has funded his lifestyle primarily through investing for the past decade. They set the stage for a deep dive into his AI-era strategy and the psychology that separates strong investors from reactive ones.
- •Alex has paid for his lifestyle mainly via investments for ~10 years
- •Conversation focus: AI era investing + avoiding buy-high/sell-low behavior
- •Alex’s approach is research-heavy and product-driven rather than finance-first
AI bubble or real shift? Infrastructure vs “AI software” hype
Alex argues parts of the market are bubbly, but not all AI is the same. He sees AI infrastructure (chips, servers, data centers) as supported by real demand and revenue, while some software companies may be riding the AI label without durable fundamentals.
- •AI infrastructure demand (compute, chips, data centers) looks real, not purely hype
- •Some “AI software” valuations may spike then collapse
- •He expects it may take years for the next Facebook/Google-scale AI software winner to emerge
- •Dot-com style crash could happen in pockets, not necessarily market-wide
What the P/E ratio actually tells you (and what it ignores)
They break down market cap, earnings per share, and why P/E ratios vary by company type. Alex prefers forward P/E but warns that analyst estimates are often wrong, so investors should connect valuation to real product momentum and business drivers.
- •P/E compares price (valuation) to earnings; forward P/E looks at expected next-12-month earnings
- •Alex’s rough preference: forward P/E under ~30 (context-dependent)
- •Low P/E “value” thinking can miss growth; earnings are only a snapshot
- •Analyst estimates can be materially wrong; product understanding can provide earlier signals
Sponsor segment: 2-Day AI Mastermind free seats offer
Marina pauses the interview to promote a two-day AI training event and explains what attendees will learn. She highlights limited free seats, bonuses, and a weekend schedule before returning to the conversation.
- •Two-day training covering prompts, tools (e.g., Make.com, Claude), AI in Excel, agents, and no-code apps
- •Claimed: millions attended; aims to move viewers from beginner to proficient
- •Limited free seats and time window; additional bonuses for attendance milestones
- •Returns to interview afterward
High-conviction bets: why NVIDIA dominates his portfolio
Alex reveals a concentrated position in NVIDIA built from holding since 2016. He describes how volatility and repeated drawdowns are normal in high-growth names, and how he averages in rather than trying to perfectly time dips.
- •NVIDIA is ~40% of his portfolio (previously even higher)
- •Held since 2016; has ridden multiple 20–30%+ drops
- •Uses gradual averaging-in during downturns
- •He sold part of NVIDIA for real-life expenses (wedding/honeymoon)
Palantir thesis: re-evaluating, selling, buying back, and letting winners grow
Alex walks through his multi-year Palantir journey, including trimming, changing his mind, and rebuilding the position. A core lesson: he didn’t buy these names to become huge weights—he held long enough for fundamentals and market understanding to catch up.
- •Started buying Palantir around ~$11; trimmed around ~$25; later bought back around ~$28–$32
- •Now Palantir is ~20–25% of portfolio largely due to appreciation
- •He publishes/uses thesis updates and is willing to admit being wrong
- •Theme: product transitions take time; market often misprices companies during transition periods
Portfolio construction: indexes as a base, selective stock picking on top
Alex outlines the rest of his holdings: heavy exposure to Nasdaq-linked indexes plus select mega-cap tech stocks. He uses indexes both for diversification and as a “default” place to park money when he doesn’t have a better idea.
- •Holds Nasdaq 100 index exposure plus SPMO (S&P Momentum)
- •Also owns individual mega-caps (Google, Microsoft, Amazon, Broadcom)
- •Trims overheated single-stock exposure and reallocates to indexes
- •Uses older index shares for withdrawals with long-term capital gains treatment (tax-aware funding of new buys)
When to sell vs when to do nothing: trimming, taxes, and the dead-investor lesson
Selling is part of the strategy, but mostly as trimming rather than exiting. Alex emphasizes that in most cases the best move is inactivity, citing the famous Fidelity finding that top-performing accounts often belonged to people who didn’t trade (even because they were deceased).
- •He trims when a position becomes too large or runs too hot; avoids selling 100% at once
- •Rebalances into broad indexes when uncertain
- •Belief: behavior/psychology drives the majority of returns
- •Fidelity anecdote: best performers were “dead” accounts—no panic-selling or tinkering
The “no bonds” stance: replace 60/40 with indexes + stocks (plus outside-stock diversification)
Alex challenges the traditional 60/40 stock-bond framework. He suggests diversifying outside equities via other asset classes, while inside equities blending broad indexes with a smaller sleeve of individual stocks chosen from the strongest parts of major indexes.
- •Controversial view: skip bonds; diversify outside stocks (real estate, crypto/metals/collectibles as desired)
- •Proposed structure: indexes as core, individual stocks as satellite
- •Prefers Nasdaq 100 for higher growth; S&P 500 for smoother ride
- •Select individual stocks from top/middle of major indexes; avoid chronic underperformers (“weed the garden”)
Dollar-cost averaging done smarter: frequency, automation, and cash posture
They discuss DCA cadence in volatile markets and why more frequent contributions can smooth timing risk. Alex also shares his approach to cash: very little inside the portfolio but a sizable emergency runway outside it due to entrepreneurial income volatility.
- •More frequent DCA (weekly/biweekly/daily) can reduce ‘one bad entry’ risk
- •Automation matters—remove emotion and timing pressure
- •He keeps <5% cash inside the investment portfolio (non-mainstream)
- •Maintains ~9 months cash runway outside the portfolio due to business income uncertainty
Using the Fear & Greed Index to add context (not to “time the market”)
Alex explains how he monitored CNN’s Fear & Greed Index during a tariff-driven selloff and used extreme fear as a signal to lean into buying. He frames indicators as additional data points—not a mechanical buy/sell trigger.
- •Follows CNN Fear & Greed Index (0–100) to gauge market sentiment
- •Example: index hit ~3 (extreme fear) near market bottom; he bought and documented moves
- •Cites Buffett principle: be greedy when others are fearful
- •Uses indicators to inform decisions, not to perfectly time tops/bottoms
Panic-selling psychology: re-framing drawdowns and “DCA out” if you must sell
Alex admits he has panic-sold before and treats it as part of learning. His practical guidance: if you believe in what you own, view downturns as discounts; if you can’t handle it, reduce exposure gradually rather than exiting all at once and regretting it.
- •Everyone makes mistakes early; panic-selling is a common learning curve
- •Re-frame drops as discounts—if you truly believe in the business
- •Panic is often driven by other people selling, not sudden business collapse
- •If selling is unavoidable: dollar-cost average out (sell 10–15% and reassess), avoid all-at-once exits
Robotics and AI infrastructure bets: liquid cooling, the “robotic stack,” and avoiding humanoid hype
The discussion shifts from mega-caps to second-order AI beneficiaries such as data-center infrastructure and cooling vendors. Alex is skeptical of near-term humanoid robots, arguing most tasks don’t require human form factors, and prefers exposure to enabling layers like compute hardware.
- •Looks for vendors and infrastructure beneficiaries (e.g., liquid cooling driven by next-gen chips)
- •Skeptical of humanoid robots today; expects specialized robots to dominate (factories, warehouses, cleaning)
- •Notes Amazon already operates massive robot fleets, mostly non-humanoid
- •Invests in exposure to the “robotic stack,” with NVIDIA as a broad beneficiary
Competition in chips: GPUs vs ASICs and building a “basket” to own the whole theme
Alex outlines how NVIDIA could be disrupted in specific workloads by specialized chips. Rather than betting on a single winner, he suggests holding a small basket (e.g., NVIDIA + AMD, or NVIDIA + Broadcom) to capture the growth of the entire segment.
- •GPUs are broad/flexible; ASICs are narrower but can outperform for specific tasks
- •Broadcom competes via ASICs; AMD is key in GPUs
- •Strategy: own 2–4 stocks to cover a whole market so the ‘winner’ matters less
- •Theme-based baskets are useful for family offices and retail alike
Crypto and Bitcoin: a missed win, an expensive Japan trip, and “invest only what you understand”
Alex distinguishes Bitcoin from other crypto, calling it more like digital gold, but he currently holds none. He shares a regretful story of selling six BTC early to fund a trip and explains he avoids Bitcoin because he lacks conviction during drawdowns.
- •Views Bitcoin as distinct from most crypto (commodity-like hedge narrative)
- •Previously held 6 BTC bought ~$1,500 and sold ~$4,500 to fund travel; regrets not holding
- •Currently holds 0 BTC due to lack of deep understanding and conviction under volatility
- •Doesn’t follow institutions blindly; aims to get ahead via domain research, accepting missed opportunities
Beginner toolkit: brokerages, research habits, and why patience compounds
They close with concrete recommendations for platforms and how to consume information without obsessing over daily price moves. Alex advises focusing on company updates and thesis validation rather than checking P&L constantly, ending on the central idea that long time horizons and patience drive outcomes.
- •Brokerages: Fidelity & Schwab (stable), Robinhood (mobile-friendly), M1 Finance (automation/portfolio pies)
- •He avoids daily portfolio checking; businesses don’t change day-to-day—market opinion does
- •Uses sources like company newsrooms, earnings calls, Yahoo News, Simply Wall St to stay updated
- •Final theme: patience and not touching positions often beats constant activity (Buffett-style compounding)
