US Yields Refuse to Fall: S&P 500's Third Down Day, Meta Rally and the 2,000-Dollar Foldable iPhone

The bulletin opens with a clear statement: the S&P 500 has fallen three days in a row and the other indexes are following it. Yesterday brought a 0.47% drop in the S&P 500, 0.2% in the Nasdaq and 0.77% in the Dow; US bond yields rose 3 basis points to 4.83% on the 10-year and 4.43% on the 2-year. A slightly softer dollar helped gold and silver, the fear gauge rose, and oil closed near $100: Brent at 101.6, WTI at 96.78. The author's portfolio finished the day up 0.018%; a positive surprise in Meta and strength in Micron kept it afloat despite red almost everywhere else.

Two technical signals stand out: the options market flipped into negative gamma, so any slide could accelerate fast; and breadth narrowed, with 10 of 11 sectors falling and only energy rising. SPY is rushing toward its 50-day moving average, QQQ still holds above its 10- and 21-day lines, and semiconductors — beaten up in July — are the one area holding the market up in September. The author's band is explicit: below 760 on SPY the drops get violent, above 770 comes relief; the max-pain level for Friday's options expiry sits at 758. Fresh earnings have dried up, corporate buybacks will pause into the new reporting season, upside catalysts are scarce; hope is pinned on Anthropic's listing.

The headache has the same name: Iran-US tension. Tehran said it will escalate, Trump argued the standoff lasts until the elections and softens after; the author reads it as Tehran playing to make Trump lose while Trump shows no intent to back down. The bill reaches the pump: a gallon of US diesel hit 473.40 cents, squeezing household budgets and Trump's popularity. The second front runs through Russia and Europe: hits on sea transport push wheat prices toward their peaks, and the whole commodity complex including copper is climbing.

The second mood-killer was Treasury Secretary Bessent's announcement: long-bond buybacks raised from $2B to $6B. The market had expected $8B, so bonds sold off; yields spiked intraday then calmed into the close. The author expects the purchases, once they actually start, to press yields down a bit; but his prescription for a lasting fall is frank: America must shrink its deficit. The logic is to buy back the long end while issuing short, betting on AI-driven productivity gains over the medium term; independent coverage confirms both the announcement and the selloff reaction.

The author finds the rate panic overdone: history shows US stocks have climbed before with 10-year yields at 4.8%; with nominal growth near 6%, a long yield around 5% looks normal to him. He sees no inflation problem either, while admitting the market disagrees with him. The calendar is packed: PPI today, CPI tomorrow, labor data alongside Oracle and Adobe earnings as the day's key reports. The picture points the same way as expectations of a September hold from the Fed even though inflation looks past its peak.

Unable to end the Iran war or bring oil down, Trump seems to have chosen cash handouts: $5,000 per head if the Republicans win the election. The author's math is simple: times 300 million people that is $1.5T, it needs Congress approval, and it piles deficit on deficit; mass impact looks limited. His more interesting thesis is political: losing the House or the Senate would bring balance, block hasty ill-considered moves, and the market would welcome it. He admits he backed Trump early and that stocks are up on net under him, but files the $5,000 promise under senseless ideas.

The fun chapter belongs to Michael Burry: the hero of the famous crisis film, the diligent analyst who read mortgage files one by one and saw 2008 early. Since then the record is poor: he had to shut his fund, his $200-a-month letter collapsed, his Nvidia and Palantir shorts bled, the Lululemon short blew up; now he looks to be moving to cash and giving up the crash call. The author's lesson is general: judge every investor inside their own era, those who fight the market end up embarrassed; he frames his own 34-year record of tripling and quadrupling the index as a tech-bull player whose 2022 bear-market exam was weak. He is equally cold on bank targets: Barclays lifting the S&P target from 7,800 to 7,950 reads to him as the never-get-caught-out reflex of cutting on falls and hiking on rallies; his own year-end target stays above 8,000.

Qualcomm arrives with an honest confession: the author bought twice and lost twice, and his rule is firm — a stock, fund or even country that burned him twice gets no third try. Still he finds the company interesting: it is moving into data centers, building its own accelerators, and announced a major cooperation with Amazon; the stock sits at $176 versus a $250 peak, the cheapest on multiples among AI chip names. A few good headlines could lift it further, but he is not buying himself; it is served to readers with that transparency.

The day's real position is Meta: the author bought options a month ago, added the stock heavily early this week, and yesterday's rally rescued the portfolio. The story swung wildly: first lagging AI work plus management fights, then recovery; the new MES models are no frontier breakthrough but they are good. The core thesis is consumer AI: nobody wants to train models at home, people want daily life made easier; against pricey rivals that need a fresh app install, an assistant living inside Instagram, WhatsApp and Facebook arrives with a ready-made distribution network. The demos are everyday life: shopping, cake plans, finding outfits, calendars, health data from the watch, a workout camera. At 18-19x forward earnings the valuation is cheap; the metaverse bill's bad aftertaste plus the will-AI-spending-pay-back question explain the discount. The comparison bites: while loss-making giants are talked about at $2T listings, cash-generating Meta with 2 billion monthly users stands at $1.5T; the stock fell from $750 to $653, and yesterday's newsflow turned positive.

Apple excited him for the first time in years: the foldable iPhone won him over, the iPhone 18 line looked boring. The price is steep: starting at $2,000, reaching $3,000 with its special case; he jokes he is afraid to check Turkish prices. His thesis is a price ladder: the average iPhone price climbed from $600-700 in 2016 to $1,100-1,200, and this product walks it toward $2,000; distancing from Korean and Chinese pressure while sharpening the premium image, with ecosystem lock-in doing its work — as an Apple loyalist with an aging battery, he will buy one. He treats launch-day dips as a classic and sees the benefit over the long term. Two closing notes: Starbucks and its inviting seating areas, plus a calendar focused on tomorrow's CPI and today's PPI; his remark is blunt: AI isolates us, time spent with three monitors and four or five AIs exceeds family time.