Oil Weekly Gain and Asia Tech Selloff Explained

The opening of August 2015 offers a loose but instructive parallel to this week. Back then, a surprise yuan devaluation combined with collapsing crude prices created a messy cross-asset selloff that took months to sort out. Today's drivers are almost the inverse: oil is spiking up on supply fears rather than collapsing, and the tech weakness is sector-specific rather than macro-driven. But the pattern of two dominant stories pulling capital in opposite directions at the same time? That part rhymes.

Let me walk through what the data is showing this Friday morning.

Two Stories, One Week

The first story is Iran. U.S. forces launched new strikes overnight, with Iranian state media reporting hits on civilian infrastructure including bridges and a railway junction. American forces also boarded a ship in the Strait of Hormuz. This is a meaningful escalation from where things stood even a few days ago. As I discussed in US-Iran Strikes and Korea Selloff: Market Impact, the market had been remarkably calm in the face of military exchanges earlier this week. That calm is fraying now. Headlines describe oil as set for a significant weekly gain, driven directly by the abrupt halt in Strait of Hormuz traffic recovery. Energy names are catching a bid as a result.

The second story is a sharp selloff in Asian semiconductor and technology stocks. Taiwan's TWII index fell 6.47% on Friday. South Korea's KOSPI dropped 6.37%. Japan's Nikkei declined 4.03%. These are not small numbers. Headlines point to a deepening chip rout, with anxiety over AI-related semiconductor spending spilling over into the broader tech complex. In Japan, the decline may have been compounded by the ruling party warning about suspected collusion between activist investors and private equity, adding a layer of regulatory uncertainty on top of the sector-specific pressure. The Nasdaq fell 1.47% on Thursday, and the tech-heavy QQQ ETF dropped 1.64%.

Meanwhile, the defensive rotation is visible in the data. The S&P 500 was down 0.51%, but the damage was concentrated in technology (QQQ down 1.64%) while midcaps (MDY) actually gained 0.51% and small caps (IWM) were essentially flat at negative 0.06%. That divergence, where mega-cap tech sells off while the rest of the market holds up, is the hallmark of a sector rotation rather than a broad risk-off event. Reports from Thursday's session indicated healthcare and consumer staples sectors were catching bids, consistent with money moving from growth into defensives and commodity beneficiaries.

The VIX rose 6.76% to 16.73, which is elevated but not extreme. Bond yields ticked modestly higher across the curve, with the 10-year at 4.569% (up 0.53%) and the 30-year at 5.098% (up 0.30%). That combination of gently rising yields and rising volatility usually signals that the market is repricing risk rather than panicking. If investors were truly frightened, you would expect yields to fall as money floods into Treasuries. Instead, the bond market is pricing in the inflationary implications of an oil supply disruption.

It is also worth noting that the broader growth and tech sentiment took another hit when Bernstein SocGen cut its Netflix price target on revenue concerns, adding to the weight on growth names.

What This Means for the Research Subjects

This is a reminder that what follows is observational research output, not personalized financial guidance. Anyone making investment decisions should consult an authorized financial advisor.

Let me work through all eleven active subjects the agent is tracking.

TTE.PA (TotalEnergies) is the most direct beneficiary of the week's dominant story. The thesis was built around Strait of Hormuz escalation providing a tailwind to European energy majors, and that is exactly what has happened. TTE is showing a positive observed delta of 1.26% from entry. With oil set for its biggest weekly gain in months on the back of supply disruption fears, the fundamental backdrop for this subject has strengthened. The thesis remains intact per the agent's latest review.

META (Meta Platforms) is down 0.7% from entry, which is modest in context. The broader Nasdaq selloff is dragging on all mega-cap tech, but META's thesis was built on growth-adjusted valuation rather than momentum. The agent actually closed a prior META research entry on July 10 with a positive observed outcome of 14.76% after the subject hit its base case level. This current entry is a fresh study at a higher price point. Given the agent's own learning that re-entries at higher prices often produce worse outcomes, the 0.7% dip barely registers as meaningful yet, but it is worth watching. The thesis review still rates this as intact.

005930.KS (Samsung Electronics) sits at a tiny positive delta of 0.2% from entry, but the environment around it just got much rougher. The KOSPI fell 6.37% on Friday as the chip rout deepened across Asia, and the broader selloff in semiconductor names is putting pressure on memory chip valuations. Samsung's thesis was built on extreme valuation dislocation in memory semiconductors, and I will be honest: the agent's own research history has flagged a pattern where re-entering semiconductor deep-value theses at progressively higher prices tends to produce worse results. The sector-wide selling pressure makes this one worth watching closely. The thesis review still calls it intact, though the weekly market reflection noted this as a potential blind spot.

BAC (Bank of America) is one of the week's quiet winners with a 3.05% positive delta from entry. Financials have been benefiting from the rotation out of tech, and the yield curve remains positively sloped (the 10-year at 4.569% versus the 3-month at 3.697% is a healthy spread). The thesis here was straightforward: quality financials with rate stability near cycle highs. This is the kind of setup the agent's research history says works best, mega-cap names with strong margins bought during periods of broader uncertainty. Thesis intact.

RTX Corporation is showing a 2.45% negative delta from entry, which puts it among the weaker active subjects. The thesis was built on defense spending tailwinds, and geopolitical tension should logically support that. But the agent's review flagged minor concerns, specifically the risk of government spending adjustments. RTX's 26x forward earnings multiple is richer than most defense peers, so it has less valuation cushion. The agent is watching this one closely.

CRM (Salesforce) has been one of the strongest research subjects, with a 9.04% positive observed delta. The thesis pointed to an extreme valuation dislocation for a profitable enterprise software company, and the market has been closing that gap. Even with the broader tech selloff, Salesforce's enterprise focus insulates it somewhat from the pure AI-chip anxiety driving semiconductor names lower. Thesis intact.

ADBE (Adobe) is the standout in the research set at a 15.34% positive delta from entry. The original thesis noted Adobe was nearly 50% off its highs with strong profitability metrics, essentially arguing the market had over-discounted AI disruption fears. So far, that thesis has played out. With the current tech selloff adding noise, the key question is whether Adobe can hold its gains. The thesis review rates it as intact, and at a 15% observed delta, this confirms the pattern the agent has observed: mega-cap compounders with strong margins bought during broad weakness tend to produce the best outcomes.

GILD (Gilead Sciences) is up 10.13% from entry, which is notable because healthcare has historically been one of the agent's weaker sectors. The agent learned from prior misses that healthcare positions only work when they are driven by genuine growth catalysts rather than just low valuations. Gilead's earnings growth exceeds the threshold the agent identified as critical. However, the thesis review flagged a minor concern: the subject recently approached its base case level around $136 and has since stalled. With the healthcare sector catching a bid as part of the defensive rotation this week, the near-term backdrop is supportive, but this one needs to either push through or the review system will reassess.

LLY (Eli Lilly) shows a 3.19% positive delta. The thesis was built on GLP-1 drug demand as a secular growth driver. Eli Lilly represents the hypergrowth healthcare exception the agent's learnings specifically carved out. The review flagged minor concerns about a recent pullback from peak levels, which could signal profit-taking. But with healthcare rotating in this week, the near-term backdrop is supportive.

EWG (Germany ETF) is the weakest active subject at a negative 2.53% delta. The thesis anticipated European outperformance driven by ECB easing and fiscal expansion, and earlier in the week German equities were performing well. But the Iran escalation is hitting European sentiment. The DAX fell 0.34% on Friday, and EWG dropped 0.63%. Headlines about Burberry's strong U.S. sales offsetting Iran war impact in Europe illustrate the divergence: American consumer demand remains solid, but European-facing businesses are absorbing geopolitical headwinds. The thesis review flagged minor concerns about renewed trade friction risks, which is exactly what is materializing. This subject bears watching.

Finally, IWM (Russell 2000 ETF) is up 3.67% from entry but carries the lowest confidence score in the research set at just 20%. The agent's own calibration data shows that positions below 60% confidence have a very high loss rate, and 20% is well below that threshold. On Thursday, IWM was essentially flat (down 0.06%), which is notable relative strength compared to the Nasdaq's 1.47% decline. Small caps are less exposed to the semiconductor selloff and benefit from domestic revenue orientation when geopolitical risk rises. But the low confidence means this subject is on a short leash.

Two Exits Worth Noting

The agent closed MU (Micron) on Friday at a 12.88% negative observed outcome after the subject hit its stop-loss. This is the pattern the agent's research history has been warning about: re-entering semiconductor memory theses at progressively higher prices after prior wins tends to fail. The first two entries on similar memory chip theses produced strong positive outcomes. The later entries have not. As noted in How to Invest 100k in 2026: 5 Strategies Ranked by Risk, mechanical discipline matters more than conviction.

PEP (PepsiCo) was also closed earlier this week at a 4.20% negative observed outcome. The confidence gate triggered, which means the agent's own system recognized the thesis was too weak to hold through a drawdown. This fits the broader learning: defensive consumer staples entered at low conviction tend to underperform.

What I'm Watching Next

The divergence between energy and tech is the clearest signal in the data right now. Oil's weekly gain is pricing in a sustained Hormuz disruption. The Asia tech selloff is pricing in a chip spending slowdown. Both of those things could reverse quickly, or they could deepen. What interests me is how the rotation into healthcare and staples plays out next week. If that continues, it suggests the market is genuinely repositioning for a world where oil-driven inflation makes growth stocks less attractive on a relative basis, rather than just reacting to a single day of headlines.

The agent's hit rate across closed research sets sits at about 49% overall, with better outcomes in the 0.65+ confidence band. Seven of the eleven active subjects carry thesis-intact ratings. The two with the deepest drawdowns, RTX and EWG, both had minor concerns flagged before today's data. That is the review system working as intended, even if the outcomes are not always comfortable.

I will be curious to see if Strait of Hormuz developments over the weekend change the energy calculus by Monday.

Research output, not investment advice. The material above is observational and educational. The operator of Observed Markets may hold personal positions in subjects the agent studies (disclosed at observedmarkets.com/conflicts-of-interest). Always consult an authorized financial advisor before any investment decision. Past observed outcomes do not predict future results.