MSFT Overbought Divergence Is Not an Auto-Sell; Do Not Chase NET
US Stocks · Options · News · Views
Thursday was quiet: the major US indexes slipped slightly, and volume shrank for a second day. Higher oil, firm data, and hawkish Fed chatter pushed the 10-year yield higher, while Google’s latest debt filing stayed in the background. The index tape was dull; the stock map was not. MSFT pressed into a key resistance band, NET and DDOG repriced in opposite directions after earnings, and SPCX’s first unlock day showed lighter selling than feared.
Macro and the Price of Money
July Challenger job cuts fell to 33,429, a two-year low: down 27% month over month and 46% year over year. Tech still led with 9,869 cuts. At the current pace, 2026 tech cuts could reach about 149,023, up 67% year over year. AI remained the main driver at 33% of July cuts and about 24% year to date.
Initial jobless claims rose to 199,000, still below the 201,000 estimate. The prior print was revised to 198,000, and the four-week average eased from 203,300 to 198,800. Continuing claims rose to 1.801 million versus 1.79 million expected. The labor market has not cracked, but rates are more sensitive again: if upcoming inflation runs hot, September hike pricing may have to move higher. For growth equities, the swing factor has shifted back toward the price of money.
GOOG Debt and Big-Tech Free Cash Flow
GOOG filed to issue about $25 billion across 10 tranches, maturities from 2 to 40 years — its third deal since 2025. The longest tranche was preliminarily talked around a 155 bp spread over Treasuries. Alphabet already issued more than $30 billion in February 2026 and $25 billion in November 2025.
Since 2025, Amazon, Google, and Meta have borrowed to fund capex, with aggregate issuance above $350 billion. MSFT still has not, because expected free cash flow remains positive. Big-tech FCF looks likely to bottom in 2026 (negative for several names), turn positive in 2027, and accelerate in 2028. The market bet is simple: AI monetization has to outrun debt-funded capacity build. If that path holds, equities usually front-run by 6–12 months, pointing toward mid-to-late 2027. If prices are still stuck low by then, monetization lag becomes the real question.
For GOOG, the 2027-forward valuation floor is around $297. With no position, $300 through the current $350–$360 zone is a scale-in band — not a full-size buy at the offer. Closer to $300, the payoff improves. Low-cost holders can keep waiting for the cash-flow story to catch up.
MSFT: Overbought Divergence, Still No Forced Cut
MSFT rose 2.54% into the $498–$516 core band. A real high-volume breakout opens a path toward $600+. Failure keeps it boxed between the mid-$400s and $500.
Volume rose versus the prior day but stayed light in absolute terms — more reluctance to sell than aggressive chase. That matters because $498–$516 is a heavy supply zone. Short-term money may fade a failed breakout, or wait and chase only after confirmation. RSI shows price at highs without a matching indicator high: overbought plus bearish divergence. A technical reset is likely — either a stronger volume spike that gets even more extended before a retest of $498–$516, or a cleaner pullback that repairs momentum first.
Even so, I am not planning to cut yet. Forward valuation and the second-quarter print still look healthy, and the long grind through $469–$498 left a decent mid-support shelf. A failed breakout implies roughly a 5% drawdown from here — too little swing for the hassle versus simply holding.
SPCX: First Unlock, Selling Did Not Avalanche
SPCX’s first unlock day was violent: about -1.92% at the low, roughly +6% into the close, a 7–8 point range, on rising volume. Most unlocked holders appeared to wait rather than dump; selling below the IPO print looked unattractive.
The next ~7% unlock is around August 20–21. Options are pricing August 21 monthly settlement near $115 ± $17 (about ±16%): downside toward $100, upside back near the $135 IPO area. If the rebound into that second unlock avoids heavy dumping, supply is manageable. With retail heat cooler, stabilization needs institutions. Rising volume without much downside is the tell for slow bid accumulation.
NET vs DDOG: Same Industry, Opposite Tape
NET jumped about 16% after earnings. Results and guidance both crushed, with a full-year raise — early-PLTR-style: high growth plus upward revisions. Even on 2027 forward and a generous 60x PE, fair value is only around $100 versus roughly $330 now. Momentum can keep stretching the bubble; the unwind can be brutal. Holders can ride the mood and wait for topping signals. No position means do not chase. Near support sits at $263–$283, then below $230.
DDOG fell about 9%. The quarter beat, but forward growth slowed slightly quarter to quarter, and the market immediately questioned the high-growth story. On ~18–19% 2027 growth and 40x PE, fair value is about $115; after the drop it still trades near $230 — still expensive. Same lane, opposite verdict: NET ripped on perfect guidance, DDOG sold off on a tiny growth deceleration. That is why betting earnings reactions is a negative-expectancy game long run. DDOG mid-support is $219–$236; mean-reversion support is below $166.
UBER: Not Expensive, Path Depends on Levels
UBER has spent recent quarters funding expansion at the expense of near-term profits and balance-sheet optics. The forward map barely moved: 2027 roughly $86–$130; 2026 roughly $59–$89 with a midpoint near $74. On fundamentals the stock is not rich, but 2026 still looks like negative growth, with a cleaner profit rebound into 2027.
Technically, $60–$76 is the core box. Back above $76.5 looks like stabilization worth holding into 2027. Below $60, step aside — the turnaround may slip another two years. Right-side trend confirmation needs $81.1.
In a low-volume tape, index noise fades and single-name pricing gets louder: MSFT risks a failed breakout but limited pullback, NET is for holders not chasers, DDOG is cheaper but still not cheap, SPCX hinges on unlock-two supply quality, and GOOG keeps funding AI capacity with debt while $300 remains the better entry math.