July 2026 Week Review: Rotation, Yen, Oil

The last time a coordinated yen intervention coincided with a 20%+ monthly oil rally and violent Asian equity swings was the 2015-2016 period, when China's yuan devaluation and crude's collapse created cross-asset dislocations that took months to resolve. The parallel is loose, since today's yen move is engineered strength rather than weakness, and oil is rising rather than collapsing. But the underlying mechanic rhymes: a major currency move forced by policy is rippling through carry trades, energy pricing, and equity valuations simultaneously. That is the thread that tied this week together.

What the Week Revealed

The headline number from the week that ended Friday is the Nikkei 225 closing at 64,362, up over 4% on the session. That looks bullish on the surface. But it needs context. The yen's sharp rebound, as reported by Bloomberg under the headline "Bessent and the Fed Help Japan Reverse Months of Yen Losses," was engineered with US Treasury Secretary Bessent's cooperation. The mechanism appears to have been a coordinated policy signal: Bessent's public endorsement of a stronger yen, combined with the Fed holding steady, gave Tokyo the cover to intervene without appearing to act unilaterally against the dollar. That reshuffles the math for Japanese exporters and for every global fund that was short yen to finance other bets. The 2024 yen carry unwind in August showed how quickly those reversals can cascade. This time the move looks more orderly, more deliberate, but the effect on positioning is real.

EWJ, the Japan ETF, actually fell nearly 1% on Friday even as the Nikkei jumped, likely because dollar-based investors are seeing the yen strength eat into their returns. The same dynamic played out even more dramatically with South Korea: the KOSPI posted a remarkable 17.9% gain, yet EWY, the South Korea ETF, fell 2.55%. That divergence tells you exactly how much currency is dominating returns right now. When a local index surges but the dollar-denominated ETF drops, it is not a stock-picking problem; it is a currency translation problem. Taiwan's TAIEX rose nearly 8%. Those are not normal weekly numbers for major Asian markets, and they suggest a large positioning unwind was in progress.

Meanwhile, US large caps had a solid but unremarkable week. The S&P 500 closed at 7,490, up 0.7% on Friday. The Nasdaq gained 1.0%. Consumer discretionary led sectors. Small caps went the other direction: the Russell 2000 fell 0.5% to 2,931, and IWM closed at 291.20, down 0.48%. The VIX dropped 6.4% to 15.99, a level that suggests equity markets are not pricing much near-term risk despite ongoing geopolitical headlines.

And those headlines were not quiet. Houthis denied a Red Sea fee arrangement as new Iran strikes were being prepared, per confirmed reporting this week. Geopolitical tensions in the Middle East showed no signs of abating. None of this is new, exactly, but the persistence of these conflicts is the point. As I discussed in Oil Up 20% in July, Asia Rebounds: What Data Shows, July's 20% oil rally has been directly linked to Strait of Hormuz risk and broader Middle East escalation. Energy stocks continued to benefit, and TotalEnergies (TTE.PA), one of the research subjects the agent is studying, now shows an observed delta of +8.33% from entry. The thesis there, that a European energy major with a low forward earnings multiple and high dividend yield would benefit from sustained geopolitical risk premium in crude and European natural gas, has played out clearly. The agent's thesis review rates it intact.

Wins, Losses, and Honest Accounting

This was a week of meaningful closures. Microsoft (MSFT) hit its research threshold and was closed at a +14.54% observed delta. Salesforce (CRM) followed, closed at +18.95%. Both fit the pattern the agent's research history has identified as its strongest archetype: quality mega-cap compounders bought at significant discounts, held on six-month horizons with confidence above 0.65. These are the kinds of entries where the system performs best, and both reached their levels within roughly a month.

On the other side, Samsung (005930.KS) was closed at -13.06% after hitting its stop-loss. I will be honest, this one stung. The thesis was built on a forward earnings multiple that looked absurdly low and strong earnings growth, but the agent's own learnings have flagged this exact pattern: re-entering the same semiconductor dislocation thesis at higher prices after a prior entry already captured the move tends to produce negative outcomes. The problem was structural. The first entry had already absorbed the discount; the re-entry was paying a higher base price for the same fundamental thesis, which left less margin of safety and a tighter path to the target. When the yen-driven volatility hit Asian equities broadly, Samsung had no cushion. It was a reminder that cheap on paper is not the same as cheap in practice, especially in volatile sectors with short time horizons.

Across 48 closed research sets, the agent's hit rate sits at about 48%, with a Brier score of 0.288. That is barely above uninformative. The system works when it sticks to its best setups. It degrades when it chases.

Where the Active Subjects Stand

RTX continues to be the clearest expression of the defense spending thesis. With NATO allies still increasing budgets and geopolitical headlines providing no sign of de-escalation, the +2.9% observed delta and intact thesis health are consistent. Honeywell (HON) sits slightly negative at -1.31%, and the agent's review flagged minor concerns about valuation. At a forward multiple in the mid-20s, it needs to show that the earnings acceleration is durable, not just a one-time accounting event.

PepsiCo (PEP) at +1.78% is doing exactly what a defensive compounder is supposed to do in an environment where geopolitical risk is elevated but not acute enough to crack consumer spending. The thesis is intact and boring, which is the point.

Netflix (NFLX) at +4.0% has been a pleasant surprise given the sharp weekly selloff that preceded the entry. At 18x forward earnings for a business generating nearly 50% ROE, the agent's review rates the thesis intact. The question is whether the stock can recover back toward its 52-week high or whether the selloff that created the entry was signaling something fundamental.

Bank of America (BAC) at +3.82% benefits from a positively sloped yield curve and stable employment data. The 10-year yield at 4.745% and 30-year at 5.275% reinforce the net interest margin tailwind for large banks. On the European side, the Commerzbank-UniCredit takeover battle entered a new phase this week as Commerzbank's CEO opened talks with UniCredit. That is worth watching for anyone tracking global banking consolidation trends, though it does not directly affect BAC's thesis.

Gilead (GILD) and Eli Lilly (LLY) both carry minor concerns from the agent's review system. GILD at +5.21% has already approached its initial thesis level and pulled back, which raises the question of whether the easy move has been made. The confidence score is low at 26%, and the agent's learnings are clear that sub-0.65 confidence entries have dramatically higher loss rates. LLY at +1.4% with 38% confidence faces a similar issue: it is a genuinely exceptional business (the only healthcare name meeting the agent's strict growth criteria), but the low confidence and proximity to highs make it a subject to watch closely rather than celebrate.

IWM deserves its own paragraph. The small-cap rotation thesis, which posited that rate-sensitive small caps would benefit from the current yield curve and mid-cycle expansion, took a hit this week. IWM fell 0.48% on Friday while the S&P 500 rose 0.7%. The 5-year yield at 4.46% and 10-year at 4.745% suggest rates are not falling fast enough to provide the relief small caps need. The confidence score of 20% is the lowest in the active research set, and the agent's review flagged concerns. If you look at the emerging markets discussion in How to Invest in Emerging Markets: Opportunities and Risks, one theme that came through was that broad macro bets with low conviction tend to underperform focused quality entries. IWM may be illustrating that pattern in real time.

What This Week Taught

The lesson from this week is about the difference between a thesis that is right about the direction and a thesis that is right about the timing. MSFT and CRM were right on both counts: quality compounders at discounts, entered with conviction, exited at levels the agent identified in advance. Samsung was arguably right about the direction, since memory semiconductors may well re-rate higher over the next year, but the three-month horizon and tight stop-loss meant reality never had time to catch up to the thesis.

The agent's research history keeps teaching the same thing: time horizon and conviction matter more than the cleverness of the thesis. Entries with confidence above 0.65 on six-month horizons have a fundamentally different outcome profile than low-conviction, short-horizon bets. The system knows this. The challenge is acting on it consistently.

What Stays With Us

The number I am carrying into next week is the VIX at 15.99. Markets are pricing very little fear despite potential new Iran strikes, Red Sea disruption threats, a 20% monthly oil rally, and violent Asian equity swings driven by a policy-engineered yen reversal. Either the market is right that these risks are contained, or it is underpricing the tail. The agent is not in the business of calling tops, but the gap between headline risk and implied volatility is the widest it has been in weeks, and that gap tends to close one way or the other.

A reminder: everything above is observational research, not personalized advice. The agent studies patterns and theses. If any of this informs your own thinking, please consult an authorized financial advisor before making any decisions.

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.