Market News · U.S. stocks and cross-asset risk
Software Retreat Outlasts Nvidia’s Earnings Relief
edition · Completed trading session: November 17, 2026
Wall Street’s problem is broader than one chipmaker. Tuesday’s selling reached software and consumer shares, while investors paid up for defensive earnings and near-term protection. Nvidia’s evening results offered an AI-demand counterweight—not a reversal of the regular-session losses.
The session in four numbers
The Dow fell 1.65%, according to the closing-session figures in Zacks’ November 17 report. Its larger decline than the Nasdaq Composite argues against describing Tuesday as only an AI correction. Nike closed at $35.85, down 6.3%; ServiceNow finished at $92.90, down $6.47, or about 6.51%. These are individual examples, not a complete market-breadth measure. [1] [2]
Three forces behind the retreat
1. Software shares face a separate test
ServiceNow’s decline shows that earnings risk is not confined to semiconductor suppliers. Its near-7% loss substantially exceeded the broad indexes. Seeking Alpha’s 4:24 p.m. report describes a sector pullback; it does not establish a single company-specific cause. [2]
Interpretation: A chip earnings beat alone need not repair software valuations. Watch whether application providers retain customers and pricing power as AI products compete for enterprise budgets.
2. Defensive does not mean inexpensive
Yahoo Finance’s midday report, citing Duality Research, put consumer staples’ forward price/earnings ratio at 24.2×, against 17.2× for technology. The same report described staples as a beneficiary of rotation and highlighted Costco, Procter & Gamble and Coca-Cola. These are reported valuation snapshots, not closing sector returns. [3]
Interpretation: Moving into steadier cash flows can reduce business risk while increasing valuation risk. The 7.0-turn reported multiple premium is not proof that every staple is expensive or every technology stock cheap.
3. Bond demand is insurance, not easy money
Bloomberg reported intraday Treasury haven buying, with yields down as much as four to five basis points after earlier gains. The dated 10-year closing series ultimately fell only 1.8 basis points from Monday, to 5.120%. Intraday extremes and the close describe different windows. [4] [8]
Interpretation: Risk aversion can lift bond prices even when inflation and expectations of tighter monetary policy keep the level of yields restrictive. A daily bond rally is not evidence of a new easing cycle.
Consumer resilience keeps the rate question open
October retail sales rose 0.2%, following a 0.8% September gain, according to CNN’s report published November 17 at 8:54 a.m. and updated at 11:29 a.m. The control category, which excludes gasoline stations, autos, building materials and restaurants, rose 0.6%, versus 0.5% in September. The headline slowdown therefore coexisted with stronger underlying spending. [5]
These are nominal, seasonally adjusted sales—not inflation-adjusted household consumption. CNN also notes October CPI increased 0.4%; that is context from an earlier inflation release, not a fresh November 17 CPI reading. Subtracting headline CPI from total retail sales would not produce a reliable real-consumption estimate because the baskets differ. [5]
The portfolio implication: Spending strength can support revenue while making rapid rate relief less likely. That combination favors attention to margins, financing needs and customer mix over a simple “strong economy means higher stocks” rule.
Hedging demand rose across three completed sessions
VIX index points, not percentages or recession probabilities. Non-trading dates are omitted. Source: dated StockApp closing series. [8]
What the market assessment does—and does not—say
StockApp’s November 17 market analysis describes a cautious environment with backwardated volatility futures, consistent with near-term protection demand. Its 3/10 risk score is coverage-constrained: undated flows and breadth were excluded, and missing inputs earned zero points. It should not be read as ten independently observed bearish signals. [9]
The analysis also uses revised macro observations rather than verified historical release vintages. Consequently, this dashboard does not use those unemployment, GDP or recession-model readings to assert what investors knew on November 17. The dated closes above and the contemporaneous retail-sales article provide the firmer time anchors.
Corporate developments: separate the clock from the catalyst
November 17 reporting only; after-hours observations do not explain earlier index losses.
Nvidia
AFP, 6:25 p.m.; after hours
AFP reported quarterly revenue above $115 billion, net income of $65.3 billion and EPS of $2.77 versus $2.48 expected. Shares rose just over 2% in the after-hours window described by the article. [6]
Investor test: Strong AI demand is an operating positive. The article also describes interim leadership and an ongoing CEO search; business momentum and governance uncertainty are distinct risks. The after-hours gain is not the November 17 closing return.
Home Depot
GuruFocus, 2:09 p.m.
Reported revenue of $43.4 billion and adjusted EPS of $3.86 exceeded cited estimates of $42.8 billion and $3.84. Full-year sales growth was guided to about 4.5%, but adjusted operating margin to about 12.8%, the low end of the earlier range. [7]
Investor test: A sales beat is not a margin beat. Watch contractor demand and conversion of sales into profit. The article’s roughly 2% premarket decline is an earlier observation, not a closing figure.
Nike
Zacks, 4:45 p.m.; close
Shares closed at $35.85, down 6.3%. The article’s earnings estimates are forecasts, not reported results. [1]
Investor test: Price weakness in a consumer brand reinforces the need to distinguish resilient aggregate spending from the fortunes of an individual retailer or manufacturer.
The dated StockApp news digest also connects software weakness, defensive rotation and Nvidia’s evening earnings. Its summaries are corroborating context, not substitutes for the linked publisher accounts. [10]
Cross-asset risk: the dollar changes the diversification story
Yahoo Finance’s November 17 article, updated at 1:21 p.m., reported that the dollar index had closed Monday, November 16, at its highest level since April 2, 2025. From the dollar’s September low through that Monday, it described silver losing nearly a third, bitcoin about 17% and gold about 14%. Those are multiweek comparisons through November 16, not November 17 daily returns. [11]
Interpretation: A portfolio can hold AI equities, gold and bitcoin yet retain common exposure to expensive funding and dollar strength. These reported drawdowns do not establish a permanent correlation, but they caution against treating different asset labels as automatic diversification. Tuesday’s Treasury rally offers a different response to stress, without removing duration or inflation risk.
Next dated catalysts—not known outcomes
Wednesday, November 18: Fed minutes
Yahoo Finance’s November 17 account flags the October meeting minutes for Wednesday. Watch whether the language supports further tightening; no minutes content is assumed here. [11]
Wednesday, November 18: Target earnings
CNN’s November 17 account identifies Target’s report as the next consumer check. Compare discretionary demand, household trade-down and margin guidance with October’s firmer control sales. No reported outcome is included. [5]
Thursday, November 19: Walmart earnings
CNN lists Walmart for Thursday. The useful contrast is customer mix and purchasing power, rather than whether one company alone validates the entire consumer economy. [5]
A conditional decision map
- Stabilization: If earnings strength broadens beyond chips and volatility retreats, the case for a narrow panic becomes more credible. One after-hours rally is insufficient confirmation.
- Persistent valuation pressure: If nominal demand remains healthy but rate expectations stay elevated, distinguish self-funded growth from businesses dependent on outside capital.
- Broader stress: If software weakness spreads while defensive multiples remain rich, neither a technology label nor a staples label is enough. Reassess position size, liquidity and valuation together.
These are scenarios, not assigned probabilities or trading instructions. The central distinction is between evidence of durable cash generation and temporary relief in sentiment.