The Last Gate Is Macro Stability: Oil, Long Yields, and the Yen Carry Trade

Corporate and market-structure evidence improved in the final week of July. Big tech continued to validate AI demand and capital spending, memory earnings estimates held up, and the transfer of a large public-equity book suggested that forced selling in crowded AI infrastructure trades had passed its peak.

The next leg of the market, however, will not be opened by earnings alone. Friday’s WTI close was $86.80 per barrel and Brent closed at $90.12. The U.S. 10-year Treasury yield ended at 4.75%, while the 30-year yield was 5.27%. The Bank of Japan held its policy rate near 1.0%, but one member preferred 1.25%. USD/JPY then fell from 163.86 on July 29 to 157.40 on August 1.¹²³

The remaining question is straightforward:

Can validated micro fundamentals overcome shocks from oil, long yields, and a potential yen carry unwind?

This report first verifies the claims circulating after Friday’s close, then maps the August 3-7 calendar and translates the outcomes into scenarios for big tech and memory stocks.

TL;DR

• Micro improved, but macro did not stabilize. Earnings, order books, and the end of forced selling improved the market floor. Oil around $90 and long yields near 5%, however, can still cap valuation multiples.
• Re-escalation in oil and Middle East risk is real. Claims that diplomacy is completely dead or that Saudi Arabia’s East-West pipeline was definitively destroyed are stronger than the available evidence. Formal ceasefire conditions have not been restored, but mediation channels remain open.⁴⁵
• The IEA’s 400 million-barrel release is real, but it was decided on March 11, not in July. Calling it a failure is not identifiable from current prices. A more defensible conclusion is that the release did not eliminate the risk premium, while the counterfactual price without the release is unknowable.⁶
• The Fed held the target range at 3.50-3.75% on July 29. The vote was 9-3, with three members preferring a 25bp increase. The hold was not a dovish all-clear.⁷
• On July 31, the official Treasury curve showed 4.75% for the 10-year, 5.28% for the 20-year, and 5.27% for the 30-year. The widely cited 5.28% belonged to the 20-year point.²
• A 63% probability of a September hike was a CME FedWatch market snapshot, not Fed guidance. BofA projected 75bp of hikes in 2026, but that was not the broad brokerage consensus.⁸⁹
• The BOJ held at 1.0% by an 8-1 vote and identified oil, semiconductor prices, and yen depreciation as upside inflation risks. Its next meeting is September 17-18, followed by October 29-30 and December 17-18.¹⁰¹¹
• The key events are the August 5 Treasury quarterly refunding and ISM Services, followed by the August 7 U.S. employment report. Strong labor data plus high oil would reopen the path to a 5% 10-year yield.

1. Verification before interpretation

The broad conclusion survives these corrections. Oil and long yields remain high, while yen carry risk has increased. The corrections simply prevent unsupported events from being used as false certainty.

2. Oil: the first macro lock

WTI moved from $82.61 on July 27 to $86.80 on July 31, after briefly falling to $79.26 on July 28. Brent moved from $88.36 to $90.12 and printed $90.74 on July 29.¹

The path reflects three observable pressures:

1. U.S.-Iran military tensions re-escalated.
2. Security risks remained around Hormuz and Red Sea shipping.
3. A formal ceasefire was not restored, leaving a fat tail around physical supply.

The correct dashboard is not rhetoric alone. It is physical cargo flow, weekly attack frequency, war-risk insurance, tanker freight, and verified mediation outcomes.

The 400 million-barrel release

On March 11, IEA members agreed to release 400 million barrels, including 172 million from the United States and 80 million from Japan. The IEA said exports through Hormuz had fallen below 10% of their pre-conflict level at the time.⁶

The release could bridge a temporary physical gap and limit an immediate price spike. It could not remove military risk, create permanent production, or normalize insurance and shipping. Current $90 Brent therefore shows that the risk premium survived, not that the release had no effect.

One-month oil scenarios

These are judgments as of August 1, not official forecasts.

The base case remains a high $80-95 range. Diplomacy is not closed, but it needs observable outputs rather than statements.

3. The Fed and long yields: confidence plus supply

The Federal Reserve held the target range at 3.50-3.75% on July 29. Beth Hammack, Neel Kashkari, and Lorie Logan preferred a 25bp increase. The statement described inflation as elevated and highlighted uncertainty from the energy supply shock.⁷

The three dissents made another hike a live policy option. Markets at one point assigned roughly 63% odds to a September hike through CME FedWatch.⁸

But the hold alone did not mechanically cause the rise in long yields. The 10-year embeds the expected policy path, inflation expectations, real growth, Treasury supply, term premium, and foreign demand. Oil, persistent inflation, coupon supply, and changing Japanese demand all contributed.

June data were also mixed. Headline PCE fell 0.1% month over month but was 3.7% year over year. Core PCE eased to 3.3% from 3.4%, so it was not at a 12-month high. Q2 GDP rose at a 1.5% annualized rate, while consumption grew 3.2% and private domestic demand 3.9%.¹²

One-month 10-year scenarios

A September hike could paradoxically mark the peak in long yields if it restores inflation credibility and is paired with a signal that policy is sufficiently restrictive. That path requires oil stabilization as well as the hike itself.

4. BOJ and the yen: the second mechanical risk

The BOJ held the overnight call rate near 1.0% on July 31. Hajime Takata proposed 1.25%, but the motion was defeated 8-1.³

The July Outlook said Japan would be supported by policy measures, accommodative financial conditions, and global AI demand. It also warned that CPI could remain clearly above 2% in the second half of fiscal 2026 because of crude oil, semiconductor prices, and yen depreciation. Inflation risks were skewed to the upside.¹⁰

This creates an uncomfortable feedback loop for memory investors:

AI demand
→ higher semiconductor and memory prices
→ higher Japanese import inflation
→ more pressure on the BOJ to tighten
→ stronger yen and carry reduction
→ a liquidity and discount-rate shock to AI and Korean memory equities

The danger is speed, not yen appreciation itself. USD/JPY fell from 163.86 on July 29 to 160.18 on July 31 and 157.40 on August 1.¹³ A gradual move allows hedges to adjust. A 2-3% daily move can force leveraged positions to close.

The real-time warning list is:

• a 2-3% one-day fall in USD/JPY,
• a jump in yen volatility,
• a sharp rise in the JGB 10-year yield,
• simultaneous selling in high-beta tech and emerging markets,
• deterioration in cross-currency funding conditions.

The next BOJ meeting is September 17-18, not October. Later meetings are October 29-30 and December 17-18.¹¹

5. Treasury and the limits of a “Bessent put”

The idea that Treasury Secretary Scott Bessent focuses on the 10-year yield is a policy and political framing, not a transfer of monetary policy from the Fed to the Treasury.

Treasury’s actual tools are maturity composition, predictable issuance, buybacks, cash-management bills, and the demand effects of financial regulation. In May, Treasury said nominal coupon and FRN auction sizes were expected to remain steady for at least several quarters, with bills and cash-management bills absorbing near-term financing variation. It also planned up to $38 billion in liquidity-support buybacks and up to $25 billion in cash-management buybacks for the May-July quarter.¹⁴

Markets sometimes call this a “Bessent put.” It is not a guaranteed ceiling on yields. A greater bill share can postpone long-duration supply pressure, but it raises refinancing sensitivity and cannot erase the fiscal deficit.

The August 5 quarterly refunding should be read through three questions:

1. Are 10-, 20-, and 30-year auction sizes maintained?
2. How much financing variation is shifted into bills?
3. How are buyback sizes and maturity buckets changed?

Reports also said U.S. officials discussed support for the yen and viewed it as undervalued.¹⁵ No official U.S. intervention announcement was identified. Treasury’s July 23 FX report placed Japan on the Monitoring List and said its currency practices would be scrutinized.¹⁶

6. August 3-7 calendar

Korea is 13 hours ahead of U.S. Eastern Daylight Time.

The schedule uses official BLS, BEA, ISM, and Treasury calendars.^17181914 July CPI follows on August 12 and will be the first major test of how quickly the oil shock reaches inflation data.

7. Cross-asset matrix

The underappreciated case is stable oil and lower U.S. yields combined with a violent yen rally. It is positive for inflation but negative for leveraged positioning.

8. Implications for big tech and memory

Big tech

AI revenue and higher discount rates can rise at the same time.

Higher AI revenue and backlog
→ higher future cash flow

Higher long and credit yields
→ higher discount rate
→ lower present value of the same cash flow

August should therefore be judged by whether AI revenue growth exceeds CAPEX growth, backlog converts to cash, power constraints allow utilization, financing costs remain contained, and existing capacity earns enough pricing power to absorb depreciation.

Memory

Oil and rates do not immediately erase HBM, server DRAM, or NAND orders. They hit equities first through discount rates and positioning.

The pullback is more likely an opportunity when HBM and server-memory volume assumptions hold, hyperscaler CAPEX remains intact, foreign and program selling slows, currencies move in an orderly manner, and earnings estimates hold or rise.

It becomes a warning when $100 oil and a 5% 10-year persist together, hyperscalers cancel server deliveries for power or financing reasons, 2027-2028 memory volume estimates fall, or a yen surge triggers broad deleveraging.

Korea-specific transmission

Korea combines heavy semiconductor weight with high foreign-flow sensitivity.

• High oil worsens import prices and terms of trade.
• High U.S. yields weaken the won and raise foreign required returns.
• A rapid yen rally can unwind yen-funded positions.
• A gradual yen rally can improve pricing competition for autos and machinery.
• Continued AI demand supports semiconductor exports and earnings.

One-direction labels such as “yen strength is good for Korea” are therefore insufficient. Speed and sector-level earnings attribution matter.

9. Operational checklist

Before Monday’s open:

1. Check whether Brent gaps above $90.
2. Verify weekend attacks and actual Hormuz cargo flow.
3. Watch for USD/JPY breaking below 157 abruptly.
4. Check whether Treasury futures reprice the 10-year above 4.75%.
5. Monitor whether Korean foreign futures and cash selling resume together.

Macro stabilization begins when at least three of four conditions hold:

• Brent stays below $85 for three sessions.
• The U.S. 10-year returns below 4.60%.
• Daily USD/JPY volatility falls below 1%.
• Big-tech and high-yield credit spreads narrow.

If two of Brent above $100, the 10-year above 5%, and a 3% daily USD/JPY decline occur together, deleveraging should be treated as the immediate risk even when micro fundamentals remain constructive.

10. Red team

This framework can fail in four ways.

1. Oil can rise while long yields fall if the shock destroys demand quickly and recession risk dominates inflation.
2. The yen can strengthen in an orderly, well-signaled process without a 2024-style carry unwind.
3. Treasury maturity management and buybacks can improve liquidity without controlling the structural level of long yields.
4. AI revenue and memory earnings can grow fast enough to overwhelm a 4.8-5.0% discount rate, making macro stability a condition for multiple expansion rather than a condition for price gains.

11. Final judgment

The end-July rebound had better quality than a simple oversold bounce. Earnings estimates held, and a major forced seller appears to have exited its public-equity book. The micro floor improved.

The macro ceiling, however, became heavier. Oil returned to around $90, U.S. long yields approached 5%, three Fed voters preferred a hike, the BOJ explicitly identified upside inflation risks, the yen began a rapid reversal, and Treasury’s long-end supply strategy faces an August 5 test.

The correct early-August question is not whether the earnings evidence was real. It was. The question is whether discount rates and funding markets stabilize enough to let that evidence be capitalized into prices.

Four gauges will answer it:

Brent crude
U.S. 10-year yield
USD/JPY volatility
big-tech and high-yield credit spreads

When all four ease together, the July rebound can become a trend. If one accelerates again, the rebound may remain a high-volatility repair rather than a durable rerating.

12. Evidence status

Facts: Oil prices use Yahoo Finance daily futures data through July 31. Treasury yields use the official U.S. Treasury curve. Fed and BOJ decisions use official statements. The coming calendar uses BLS, BEA, ISM, and Treasury schedules.

Judgments: One-month probabilities and ranges are scenario estimates as of August 1. “Macro locks” and “the second mechanical risk” are analytical transmission frameworks.

Blocked: The exact size of global yen carry positions, official U.S. intervention activity, the final Middle East mediation outcome, Monday’s September Fed pricing, and the counterfactual oil price without the IEA release cannot be observed from public data.

This report is market analysis, not personalized investment advice. Conflict, oil, and rate pricing can change quickly and should be rechecked before Monday’s open.