Clean Slate

I rebuilt the book from scratch, as if today's $100,000 were fresh cash, then traded the gap to what I hold. The backdrop: US stocks are within half a percent of their highs, earnings season has been unusually strong, and a soft July jobs report (the economy unexpectedly shed jobs) pushed bond yields down and equities up, easing fears that the Fed will need to tighten soon. That argues for staying fully invested, but with a broad, boring core rather than a concentrated bet, and with August/September seasonality historically choppy I want a real shock absorber. What I'd buy today: a 22% total-market core (VTI) plus 15% equal-weight S&P (RSP) and 7% dividend-quality (SCHD) — 44% of the book in cheap, diversified beta that doesn't depend on any single story. 10% healthcare (XLV), which is the best-performing defensive sector over three months and still trades at a discount to the market. 6% JPMorgan, near its 52-week high with the steep-curve tailwind intact. Individual technology gets 20%: Microsoft 6%, Nvidia 5%, Amazon 5%, Alphabet 4%, Broadcom 4%. Microsoft is a new position — its quarter genuinely reset the debate about whether AI capital spending pays off (Azure grew 43%, cloud backlog jumped to roughly $678 billion), and even after a violent rally the stock is still well below its 52-week high. Broadcom is new for similar reasons: the strongest six-month trend in the universe with a real, contracted order book behind it. Rounding it out: 4% energy (XLE), 4% gold (GLD) as a cheap hedge after a brutal six months, and 8% T-bills (BIL) as dry powder. What I dropped that habit would have kept: VTV goes to zero. It's a fine fund, but between VTI, RSP and SCHD I already own its holdings two or three times over — it was adding line items, not diversification. I also cut energy roughly in half; with oil drifting lower on Iran/Hormuz negotiations, a 7.7% weight was more conviction than I actually have. And I trimmed the oversized RSP position back to a deliberate 15% rather than letting it sit at 23% simply because that's where it landed. What I kept despite the temptation to sell: Alphabet. It's down about 10% over three months and 11% off its high, and the easy move would be to cut a laggard. But the business is intact and I'd rather own it at a discount than pay up elsewhere — so it stays, at a smaller 4%. I was also tempted to drop Nvidia after a 12% week, and chose not to chase it either; 5% is where I want it. Risk notes: no leverage, no shorts, largest position 22%, and the book is essentially fully invested with 8% in bills. Everything fills at the next official close.

Rebuilt from scratch as if starting with $100,000 in cash today, then traded only the gaps versus what I hold. The book I want: a broad, diversified core (23% equal-weight S&P via RSP, 15% total-market VTI, 8% large-cap value VTV) because equal-weight and value are the calmest, closest-to-52-week-high parts of this market right now, with 30-day volatility near 10% versus 12-56% for the tech-heavy options. Around that core: 12% healthcare (XLV), the strongest sector over three months and defensively priced; 8% JPMorgan, which has strong six-month momentum and benefits from Treasury yields sitting near multi-year highs; 8% energy (XLE), up double digits over the past month with a 3.2% dividend yield; and a deliberately modest 16% in single-name AI exposure (6% Alphabet, 6% Amazon, 4% Nvidia). The remaining 8% goes into T-bills (BIL) as ballast paying roughly 3.8% with essentially no price risk. What I dropped that habit would have kept: gold (GLD) and utilities (XLU). Gold is down about 24% over six months and sits 24% below its 52-week high; with the Fed on hold and long yields elevated, I would not initiate that position today, and "it's my hedge" is not a reason to keep it. Utilities are down on the week and negative over one and three months - a bond-proxy sector fighting rising yields. Both are out entirely, replaced by T-bills, which do the same defensive job without the drawdown. What I bought new: Amazon, at a small 6%. It reported quarterly revenue of about $200.6 billion with AWS growth accelerating roughly 37% year over year - its fastest in years - and a large contracted backlog, which is genuine evidence that AI spending is converting into revenue. The stock gapped up sharply on the news and my order fills at today's close, so I am paying up; I sized it small for that reason. What I kept though I was tempted to drop: Nvidia. It is my worst performer, down over one week, one month and three months, and I nearly cut it to zero. I kept a reduced 4% because the chip complex is rebounding hard this week and because Amazon's cloud numbers argue the AI infrastructure cycle is intact - but at 4%, not 6%. Trims: RSP and Alphabet get shaved back to target after their run (Alphabet jumped over 6% today), and the crowded, high-volatility semiconductor and Microsoft trades were considered and passed over - Microsoft is up over 20% in a month with 55% volatility, which is a poor entry point for a one-week holding period.

The portfolio starts as 100% cash, so this is a clean build from scratch — nothing was inherited and nothing was kept out of habit. Backdrop: this was a jittery week. Thursday saw the worst session in a month, with the S&P 500 down about 1.2%, driven by a spike in oil after tanker attacks in the Red Sea and by fresh worries about how much the big tech companies are spending on AI. Long-term interest rates also rose, with talk of another Fed hike. Because I only revisit this book once a week and hold through whatever happens in between, I want a core-heavy, diversified portfolio rather than a handful of high-volatility bets. What I bought and why: - RSP (25%, the largest allowed) and VTI (15%): 40% in broad US equity. I deliberately tilt the core toward the equal-weight version of the S&P 500, which owns the same companies but with far less concentration in the crowded AI mega-caps that are currently the market's main source of stress; its recent price action has been steadier (lowest 30-day volatility in the universe, barely off its high). - XLV healthcare (12%): defensive, improving momentum (best 3-month return of the sector ETFs here), near its 52-week high, modest volatility. - JPM (8%): the only name in the universe sitting exactly at a 52-week high, with strong 3- and 6-month gains; higher rates and active markets are a tailwind for a large, profitable bank. - XLE energy (8%): strongest 6-month trend in the universe plus a 3.2% dividend, and it is a natural offset if Middle East tensions re-escalate. I kept it to a modest size rather than chasing, because crude has already given back part of the spike on hopes of renewed US–Iran talks — much of what I'd be buying is a geopolitical risk premium that can deflate quickly. - GOOGL (7%): sold off sharply this week even though June-quarter results beat expectations by a wide margin; the drop was about a raised AI capital-spending budget, not weak business results. At roughly 21% below its high, I'm willing to own the cash-flow engine and wait out the spending debate. - NVDA (6%): a deliberately sized, not oversized, stake in the central AI franchise. It has held up better than most AI names, but sentiment on AI spending is fragile, so it stays a single-digit weight. - XLU utilities (6%) and GLD (5%): ballast. Utilities pay 3.2% and behave defensively; gold is the cheapest insurance available here against the geopolitical and inflation tail, and it's already 25% off its high, so I'm buying it weak rather than chasing it. - About 8% left in cash as a buffer, since I can't react intraday for a week. Tempted but declined: MU was the biggest temptation — it has more than doubled in three months on an AI memory boom, but its annualized volatility is above 100% and it fell roughly 8% in the snapshot alone; that is too much single-name risk for an unattended week. MSFT is genuinely cheap-looking versus its own history (nearly 30% off its high), but it reports earnings next week while I'm locked in, so I'd rather wait and decide with the numbers in hand. CAT and GEV are strong longer-term trends but carry 50–67% volatility and have been rolling over recently. WMT and COST are quality businesses in clear downtrends with no valuation cushion, and IWM adds small-cap sensitivity to rising rates that I don't want right now.