Sep 8, 2026

Yemen's Houthis struck southern Saudi Arabia early Tuesday with ballistic missiles and drones. The Saudi energy ministry said fires broke out at a number of sites and forced a temporary suspension of some operations. The Saudi-led coalition put the wounded at 73 across Abha, Khamis Mushait, Jazan and Najran. The Houthis claimed Aramco facilities in Abha, Jazan and Najran plus the King Khalid airbase, and said a broader operation deep into Saudi territory would follow. Brent rose $1 to $98 on the ministry statement and reached $99.46, highest since July 24; WTI reached $94.73, highest since June 8.

Verified because the operator confirmed it: the energy ministry issued its own statement and the coalition put its name to the casualty count, carried by Reuters, Al Jazeera, AP and The National. Unverified is the size. No lost volume has been quantified, Aramco has not commented, and the Financial Times report of a separate Monday strike on Jazan remains uncorroborated by the company, which is why the desk holds that Sept 7 claim at partially verified.

The geography cuts against the loudest reading. Abha, Najran, Khamis Mushait and Jazan sit in the southwest, near Yemen and on the Red Sea side, not the eastern crude export complex at Ras Tanura, Abqaiq and Juaymah where a genuine crude shock would originate. What the southwest holds is refining and domestic product infrastructure, the leg already most impaired. This is a product story until evidence says otherwise.

Japan's labour ministry reported Tuesday that July nominal cash earnings rose 4.7% year on year to 436,401 yen a month, the fastest since January 1997 and well above the 3.8% consensus. It was the sixth consecutive month above 3%, the longest run in 34 years. Real wages rose 2.4%, a seventh straight increase and the best in about five years. Base pay rose 4.1%, fastest since April 1992, and a cleaner full-time measure stripping bonuses, overtime and sampling distortions rose 2.7%. Q2 GDP was revised up to 1.4% annualized from 1.1%.

A primary release from the Ministry of Health, Labour and Welfare, carried by Bloomberg, Reuters and the Japan Times. The one asterisk is the GDP revision, an upgrade that still missed the 1.6% to 1.8% forecast range.

The counter-evidence matters more than the beat. Wage growth this strong removes the Bank of Japan's main stated hesitation, that tightening would choke off the wage-led recovery, and swaps now price roughly 98% odds of a move to 1.25%. But household spending fell for an eighth straight month and Q2 private consumption was flat. Higher pay is not becoming demand. That gap is why the coming hike reads as an inflation and currency response rather than confirmation that Japan's domestic economy has turned.

Kyodo News reported Tuesday that the Bank of Japan plans to raise its policy rate from about 1.0% to about 1.25% on Sept 17 to 18, which would be the highest in roughly 31 years, citing the risk that rising oil and a weak yen push inflation above expectations. Reuters, citing people familiar the same day, reported officials prefer the usual 25 basis point increment while weighing a quarterly cadence. Bloomberg reported the same lean on Sept 3.

Partially verified rather than verified, and the distinction is the point. There is no Bank of Japan statement. Kyodo's language that the bank has settled on a plan is a wire lead sourced to unnamed people, and the BOJ is inside its pre-meeting window. Two independent wires plus the earlier Bloomberg report corroborate direction and size, which keeps this above unverified, but a policy decision is verified when the central bank publishes it.

The consequential detail is the size, not the hike. The Yen Carry gauge's third worsen leg requires a hike larger than 25 basis points. All three reports point to exactly 25 and swaps price that near 98%. That leg is being priced out rather than in, which argues against the aggressive carry-unwind reading, and the desk records it that way even though it points away from the more dramatic story. What stays live is the Ueda statement on cadence: a hike delivered with an explicit quarterly path is a different signal from the same hike delivered alone.

USD/JPY fell to 152.89 Tuesday, the yen's strongest since February, from Monday's 154.36 close and Friday's 156.19. Finance Minister Satsuki Katayama said Japan's stance has not shifted since the joint intervention with the US Treasury, that authorities will work to maintain orderly markets, and that she will stay in close contact with Secretary Bessent. Transmission into Japanese equities was immediate: the Nikkei fell 1.7% to 65,269 from Monday's 66,400 as a stronger yen hit exporter earnings expectations. The Kospi rose 2.4%.

Verified from market data and Bloomberg's report of her comments. What the desk will not inherit is the framing that travelled with them. At least one broker note characterised Katayama and Bessent as signalling joint intervention; the underlying report says something narrower, that she restated an unchanged stance and declined to characterise the move.

The direction the language now points is what matters. Every intervention this cycle has been yen buying against a slide toward 160, and the orderly-markets framework was built for depreciation. Katayama deployed it after a rally, with the yen up roughly 4% in a week. Nothing in her remarks says Tokyo will lean against its own currency's strength. It does mean the risk this framework manages may be inverting, which is when a carry unwind turns disorderly. On the desk's test, today was not that: the single session move was 1.47 yen against a three yen marker, 49% of the trigger and below Monday's 71%, so the arrow comes off rather than escalating.

Daan Struyven, Goldman Sachs co-head of global commodities research, told the APPEC conference Tuesday that oil product flows out of Hormuz run at about 35% of pre-war levels against about 70% for crude, saying the shock is bigger for products than crude and worst for heavier products such as diesel. He gave no volumes. Kpler's pre-war baseline puts scale on it: roughly 2.2 million bpd of refined products moved out of the Middle East via Hormuz last year against 14.95 million bpd of crude. Goldman separately raised its December 2026 Brent forecast by $5 to $85 and 2027 to $80, assuming disruption persists into 2027.

Partially verified because of what it is: a spoken remark at a conference carried by Reuters, not a research note the desk has read. The 70% crude figure matches Goldman's published work since June and the direction matches shipping data independently, but the 35% product number rests on one attributed comment.

This ranks high because it explains the inflation channel, not the price. Crude has partially rerouted through Red Sea pipelines and Fujairah; refined product cannot, spare refining capacity is thin, Russia is banning exports and winter demand is ahead. It shows up where consumers meet it: US diesel hit record highs last week and Labor Day gasoline prices set records. Three days before August CPI, that is how a Gulf shipping problem becomes a Federal Reserve problem, and why strikes on Saudi southern refining matter more than their distance from the crude terminals suggests.

CLARIFICATION Sept 8, 2026: the diesel and Labor Day gasoline records cited above are September observations. They do not enter Thursday's August CPI print, which reflects average motor fuel prices across the calendar month per BLS methodology. They bear on the September print and on the forward inflation path, not on the August release. The Fed channel described here is prospective, not already in the data.

Kpler's ten day moving average of commodity vessels transiting Hormuz stands at 10 a day, the lowest since May, down from more than 15 Friday and nearly 13 Saturday. Only two vessels passed Saturday and six Sunday, mostly on the Iranian route. No VLCC has exited since Wednesday, Sept 2, and LSEG data showed a Saudi-loaded product tanker attempt an exit and turn back. UKMTO counts 27 projectile strike incidents since July 6.

Verified as ship-tracking counts reported by Reuters plus a UKMTO tally. Two things circulating alongside them are not. Araghchi told his Japanese counterpart Tuesday that Iran and Oman have made significant progress toward a temporary transit route: a diplomatic statement, with no Omani confirmation, no operational notice and no coordinates. And the President's claim of roughly 30 ships a night sits against single-digit commodity counts on two of the last three days. The two may not measure the same thing, since the Joint Maritime Information Center counts facilitated vessels of all types, but the gap is wide enough that the desk carries both.

The absent VLCCs are the number to watch. Six days without a VLCC exit is a checkable statement about whether crude can get out at all, and it sharpens the split above: crude at 70% of pre-war while no VLCC exits means the routing workaround is doing more work than the chokepoint.

A rich list alert bot posted that Coinbase gained 5,569,165,957 XRP, a 4,870.7% increase, and it circulated as an exchange inflow. XRPScan, which operates the explorer the bot reads, said Coinbase did not gain 5.5 billion XRP and that it had finished identifying and labeling 146 previously unidentified Coinbase wallets, moving the exchange to third on the rich list.

The receipt is in the bot's own output. The same alert showed the Unknown category falling by 5,569,802,824 XRP in the same hour, an almost exactly offsetting move. Balances did not change; the label on them did. A transfer that size would have a counterparty and a corresponding decline somewhere identifiable, not a matching drop in the residual bucket. That is why this publishes contradicted rather than unverified.

The pattern will repeat. Explorer relabeling produces enormous apparent balance changes with no economic content, and rich list alerts cannot distinguish the two. An exchange balance move is meaningful only when the transaction is identifiable on ledger with a known counterparty direction. This says nothing about XRP demand either way.

A live read of the XRP Ledger amendment table Tuesday shows BatchV1_1 (XLS-56) at 25 of 35 validators against a 28 threshold, three short of majority with no majority flag set, so the two week activation clock has not started. fixCleanup3_3_0 holds at 30 of 35 with its majority flag set and an on-ledger majority timestamp corresponding to Aug 28, putting earliest activation at Sept 11. Single Asset Vault (XLS-65) sits at 13, Lending Protocol (XLS-66) at 11, PermissionDelegationV1_1 (XLS-75) at 12, and ConfidentialTransfer (XLS-96) and DynamicMPT (XLS-94) at 9.

Verified because it is read from the ledger through the XRPScan API rather than from commentary. One caution: a deprecated Batch amendment also carries the XLS-56 number and sits at 12 votes. The live proposal is BatchV1_1, and conflating the two understates progress by half.

Batch enables all-or-nothing multi-account atomic transactions, the primitive underneath the RLUSD and tokenized-asset settlement work, and it is the closest non-fix amendment to threshold. None of this moves a gauge: amendment progress is engineering coverage, and an amendment is verified as live only once enabled on ledger. Three votes short is not shipped, and XLS-65 and XLS-66, which the Clearpool RLUSD credit fund depends on, remain far from it.

US spot XRP ETFs recorded zero net flow on Sept 4, ending the week with cumulative net inflows of about $1.68 billion and combined net assets near $1.48 billion. Sept 7 was Labor Day, no session and no print. Tuesday's figures post after the close. The last non-zero daily was Sept 3 at +$6.14 million, led by Franklin's XRPZ at $3.19 million and Bitwise at $2.95 million.

Verified as to what exists and, more importantly, what does not. A missing panel intraday is reporting lag and never a signal, and that applies with full force across a holiday weekend when three of the last four calendar days could not have produced a print. Reading a blank tracker this week as stalled demand is reading the calendar, not the data.

The rule has a cost worth stating, because it runs against the tape. The XRP Flows gauge carries a worsening arrow set on the completed Sept 4 week, which came in at +$18.96 million against $110.49 million prior, a decline of about 83% that moved the tracked level most of the way to the zero line without crossing it. The downgrade fires on a verified net weekly outflow and nothing else. Until a real print says so, the gauge holds at NEUTRAL regardless of how weak the flow trend looks.

A direct read of the RLUSD issuer account puts XRPL supply at $1,028,965,962, 42.5% of about $2.42 billion outstanding, leaving a non-XRPL remainder of $1,394,552,285. Ethereum leads by roughly $366 million. Independent live chain data published Sept 7 reads Ethereum near $1.38 billion against $1.03 billion on XRPL, and CoinGecko puts market capitalisation at $2.407 billion. The two reads agree closely enough that Base, Ink, Optimism, Unichain and the XRPL EVM sidechain together hold on the order of $15 million.

This contradicts the claim logged unverified Sept 7, that XRPL had retaken the lead at $810 million against $756 million. XRPL is not above 50% of supply and therefore cannot be above Ethereum. The arithmetic gives it away too: those figures imply a $1.57 billion total, about $858 million below the verified one. A tracker cannot be reading current supply and miss a third of it. Those levels match late June, when XRPL briefly did lead at roughly $801 million against $795 million. A stale snapshot circulating as a September development is contradicted, not merely unverified.

XRPL led once, in late June. By Aug 21 Ethereum was ahead at $989 million against $941 million, and the gap has widened about sevenfold since, with the 1,363,614.85 RLUSD moved off XRPL on Sept 6 one transaction inside that migration. None of this is bearish for RLUSD, whose supply is growing; it says which chain captures the activity. RLUSD settlement also does not create XRP demand beyond network fees and reserves, so this bears on the XRPL utility argument, not the token's demand case.