Seven days apart: two votes, opposing incentives on Iran
Israel votes on 27 October, the United States on 3 November: seven days that govern the autumn. The Mecca Pact, the Venezuelan oil lock, the CIA's visit to Moscow and monetary fragmentation, with five quantified scenarios and their consequences for equities, commodities and currencies.
Israel votes on 27 October. The United States votes on 3 November. Between the two ballots, opposite incentives bear on the same Iranian file, and it is this calendar, far more than any doctrine, that governs the autumn.
What has changed since the last report
People often look for a grand strategy where there is only an electoral agenda. Autumn 2026 offers an almost clinical case: two elections seven days apart, drawing on two opposing electorates, exert opposite pressures on the same decision. Should the war against Iran be taken to the ground, yes or no?
This calendar constraint overlays a second, slower recomposition: the international payments system is fragmenting while the reserve system is not being replaced. The two movements are not of the same nature. Confusing them is the most common error in current geopolitical commentary.
This article defends a thesis in four steps. One: the decision window on Iran is narrow and datable. Two: the Mecca Pact is the rational response of middle powers to an order that no longer guarantees anything to anyone. Three: the American oil constraint changed in nature in January, and that change makes escalation less costly than people believe. Four: monetary fragmentation is real but it is transactional, and the only credible candidate to succeed the dollar methodically refuses to pay the price.
I — Two ballots, two opposing incentives
On 27 October, Israelis elect the twenty-seventh Knesset. On 3 November, Americans renew the House and a third of the Senate. Seven days. On the Iranian file, these two deadlines do not push in the same direction; they push in strictly opposite ones.
On the American side, public opinion is openly hostile to a ground operation. A Quinnipiac poll from March 2026 measures 74% opposition to sending ground troops to Iran; the spring Economist/YouGov surveys give 62% opposition to a land invasion and 59% opposition to the war itself, against 28% support. The decisive figure is not there: it is that Republicans themselves are split, 41% against a ground operation, 39% in favour. The fracture runs through the base, between MAGA Republicans (79% support for the war) and non-MAGA (33%).
An unpopular ground war, launched a few weeks before a midterm election, on an electorate half of which disapproves: the arithmetic needs no comment. The constraint is real, measurable, and runs until 3 November.
On the Israeli side, Benjamin Netanyahu's position is fragile. Polls from late June and July 2026 credit the Likud / ultra-Orthodox / radical-right bloc with 49 to 52 seats out of the 61 needed. Gadi Eisenkot's Yashar party oscillates between 21 and 25 seats and at times overtakes Likud. Neither has an assured coalition.
This is where precision is needed, because it is the point where analysis slips most easily. That a weakened government has an interest in a tense security climate is a classic proposition of political science, and the gap between the polls and the 61-seat threshold gives it a factual basis. But moving from objective interest to deliberate intent is a leap that nothing in the available sources lets us make. One can describe a structure of incentives. One cannot read minds.
The operational conclusion is counter-intuitive: the riskiest period is not October, it is what comes after 3 November. As long as both campaigns run, American restraint holds. Once the polls have closed on both sides, no electoral calendar disciplines the decision for a long time.
II — The Mecca Pact, without hype
On 7 August 2026, in Mecca, Crown Prince Mohammed bin Salman, Recep Tayyip Erdoğan and Shehbaz Sharif sign a joint defence agreement: an attack on any one of the three states is considered an attack on all three. The trio fields nearly 1.4 million active-duty personnel, 3,400 aircraft, 6,000 tanks, and combined defence budgets of about $124.4bn ($63.9bn for Riyadh, $51.4bn for Ankara, $9.1bn for Islamabad). The text extends the Saudi-Pakistani bilateral agreement of September 2025.
Three elements give this text its real weight, and none of them is rhetoric.
First, Pakistan is the only nuclear power in the Muslim world. A mutual-defence clause that binds it changes the nature of the adversary's calculation, regardless of any doctrine of use.
Second, the venue. Signing a military pact in Mecca is not a logistical convenience; it claims a legitimacy that extends well beyond the three signatories.
Finally, and this is the most neglected fact, Pakistan mediated the 8 April 2026 ceasefire between Washington and Tehran, and it was in Islamabad that the delegations met on 11 and 12 April. The signatory of the pact is already the recognised intermediary of both camps. This is not a protocol detail: it is what makes the whole credible.
Several outlets, including the Jerusalem Post, report that Iran was invited to join the agreement. Neither the Iranian foreign ministry nor any of the three signatories has officially confirmed it. Turkish foreign minister Hakan Fidan, however, publicly stated that the pact was not meant to stay at three.
A Sunni axis that welcomed Iran would, by definition, cease to be a Sunni axis.
That is where the real rupture lies, and it is logical before it is strategic. Western regional policy has rested for decades on the depth of the Sunni/Shia divide. A security framework that made that divide secondary would not alter a balance of power: it would invalidate an interpretive grid. Hence the visible difficulty of American commentary in characterising the object.
Proportion must nevertheless be kept. The pact has no integrated staff, no permanent command, and no automaticity of commitment comparable to Article 5. Several analyses, including that of Foreign Policy, stress that it may remain largely declaratory. The signatories themselves have insisted on its undirected nature: collective deterrence, with no designated target. In this region, that is a message in itself.
III — Moscow: what happened, and what did not
On 25 August 2026, CIA director John Ratcliffe makes an unannounced visit of about eight hours to Moscow, the first by an Agency chief to Russia since the start of the invasion of Ukraine. He meets Sergey Naryshkin, head of the SVR, and Alexander Bortnikov, director of the FSB. According to the Washington Post, Axios and Meduza, he warns Moscow against the temptation to exploit American weakness arising from the war against Iran, raises the Baltic states, and floats the idea of a Trump–Putin–Zelensky summit. At the same time, Abp. Paul Gallagher, the Holy See's Secretary for Relations with States, leads a five-day mission to Moscow, publicly offering dialogue and humanitarian aid.
These two facts are solid. Their narrative framing is much less so. The formula in circulation, "the CIA and the Vatican travel to Moscow together", describes a coordination of which no source speaks. Religion News Service, which covered the convergence, explicitly speaks of separate approaches aimed at the same goal: ending the war in Ukraine. Simultaneity is not concertation, and the reported subject of both visits is Ukraine, not Iran.
The Baltic hypothesis, by contrast, is solid
Lithuanian president Gitanas Nausėda and Latvian president Edgars Rinkēvičs have publicly warned that Russia may be preparing "limited kinetic operations" against critical infrastructure on the eastern flank. Nausėda confirmed he has intelligence to that effect. Rinkēvičs explicitly raised a Russia that might "indirectly test Article 5". Poland has intercepted Russian aircraft probing its air defences. A simulation exercise run by Die Welt modelled a Russian incursion into Lithuania through the Suwałki gap, with 15,000 troops enough to seize Marijampolė.
Yet this is precisely what Ratcliffe went to discuss. The account of his visit mentions a warning against any attack on NATO members, the Baltic states by name. The sequence fits together: Baltic services warning of a test operation, Washington sending its intelligence chief to say so in person in Moscow rather than through diplomatic channels.
The extension of the reasoning (Moscow would open a Baltic front to saturate a NATO already committed on two theatres, and thereby relieve its Iranian partner) is strategically coherent but appears in no source. It presupposes a Russian-Iranian coordination of which nothing attests the existence. To be kept as an interpretive grid, not as fact.
IV — An empty reserve, a full lock
The US Strategic Petroleum Reserve falls to 289.7 million barrels in the week of 21 August 2026, its lowest level since November 1982. It held about 415 million barrels before the strikes of 28 February. After Iran's disruption of the Strait of Hormuz, the administration authorised in March a drawdown of 172 million barrels. The level is now approaching the threshold below which questions arise about the integrity of the salt caverns themselves.
The fact is spectacular and it is true. The reasoning usually drawn from it is much less so. From "the reserve is empty" people slide to "the United States must therefore seize Iranian oil", and that particular slide does not hold: an Iranian barrel does not become American by occupation. It takes intact facilities, terminals, insurers and an open strait, which is exactly what a ground operation destroys.
But there is a much sturdier version of the oil argument, and it does not run through Tehran. It runs through Caracas.
The Venezuelan lock
On 3 January 2026, an American military operation captures Nicolás Maduro and transfers him to New York. Seven months later, Venezuelan production has gone from an average of 941,000 barrels a day in 2025 to 1.12 million in July 2026. Exports reached 1.16 million barrels a day in April, a seven-year high. Above all, 786,000 barrels a day went to the United States in July, a record since early 2019. In August, Hunt Oil signed a production-sharing contract on two fields, and SLB a framework agreement for reservoir studies.
The chronology deserves a pause. Maduro is captured on 3 January. The strikes against Iran begin on 28 February. Eight weeks separate the two events. One can see a coincidence of timing; one can also see a sequence: first secure a source of heavy crude in the Western Hemisphere, then accept the risk of a Middle East disruption. No source establishes intent. The sequence itself is factual.
It was not Iranian oil that had to be taken. It was Venezuelan, and that is done.
Why a high price serves heavy crude
Venezuela's Orinoco crude is extra-heavy and sour. It is expensive to extract, requires diluents, and can only be monetised in refineries equipped with coking units. Those refineries exist: they line the Gulf of Mexico and were designed for exactly that crude. Hence three consequences that are rarely articulated together.
- Heavy crude is structurally scarce in the United States. American production growth comes from the Permian basin, which produces light, sweet crude. Gulf Coast refineries therefore lack heavy barrels, a structural deficit that durably supports premiums on medium and heavy grades.
- Venezuela fills precisely that gap. Venezuelan crude was only 3.5% of total US imports, but about 13% of Gulf refinery imports. Its value to the American refining system is out of all proportion to its statistical weight.
- A high price turns a marginal barrel into a profitable one. A crude that is costly to produce and process only has economic interest above a certain threshold. The Hormuz disruption took Brent from $58, the EIA's initial forecast for 2026, to spikes above $107, and it was still trading around $88 to $90 at the end of August. In this configuration, the Orinoco ceases to be a political asset and becomes an industrial one.
During the Hormuz crisis, Mars (the Gulf of Mexico medium sour benchmark) traded as high as $119.3 while Brent hovered around $107 to $109: an inversion of the usual discount. This kind of premium is extremely unstable: the Mars differential lost more than $10 in a single session on 9 and 10 March 2026. Retain the direction of the phenomenon, not the level.
The resulting asymmetry
That leaves the most important point: a Gulf disruption does not hit everyone equally. The Strait of Hormuz carried about a fifth of the world's oil supply before the war. Those who depend on it are China, India, Japan, South Korea and, indirectly, Europe. The United States, for its part, now has a refining system fed by light Permian and heavy Venezuelan crude, two sources in the Western Hemisphere, out of reach of a blockade at Hormuz.
In other words: the closure of Hormuz is an asymmetric catastrophe. It costs Washington dearly in pump prices and domestic politics; it costs its systemic competitors far more in physical volumes. The disruption that weakens America weakens China more.
That this asymmetry constitutes a deliberate motive for escalation is an inference. It is economically coherent and nothing contradicts it, but no source establishes it. It must be held for what it is: a serious working hypothesis, not a fact.
Correction of a hasty conclusion — taken in isolation, the weakness of the SPR argues against escalation: it measures a vulnerability, not a capability. But placed next to the Venezuelan lock, it changes meaning. The America of 2026 is less exposed to a Gulf disruption than any of its rivals, and it has just secured the means to be so. The oil constraint no longer clearly argues against escalation. It has become ambiguous, which is something quite different.
V — Is a wounded power more dangerous?
This is the question that governs everything else, and it deserves better than an animal metaphor. It has a name in international relations: the logic of the closing window. A power that anticipates its own relative decline has a rational incentive to act while it still has the means, a thesis formalised by Stephen Van Evera on windows of opportunity and vulnerability, and by Dale Copeland on dynamic power differentials. It is not irrationality: it is a calculation, and that is what makes it worrying.
- Three pillars of power degraded simultaneously. Military, after a campaign that did not produce the announced regime change. Energy, with a strategic reserve at its lowest in forty-four years. Financial, with creditors withdrawing.
- A declared objective not achieved. The strikes of 28 February explicitly aimed at regime change. It did not happen.
- The conflict is already no longer frozen. The 14 June memorandum was supposed to end hostilities within sixty days; by July, strikes had resumed. Hormuz has still not reopened.
- The Venezuelan lock is in place. The main material objection to escalation was partly lifted in January.
- An adversary potentially tempted to open a second front. If the Baltic warnings materialised, NATO would find itself committed on three theatres.
What this grid does not say
The slope must be resisted. The neighbouring thesis, that of the diversionary war, according to which a leader in difficulty would start a conflict to rally public opinion, is empirically one of the most contested in the discipline. And in the present case it runs into a decisive obstacle: the war is already unpopular. Escalating does not remobilise an electorate that rejects it at 59%. A diversionary war presupposes a rally effect; there is none here.
It is also worth noting what Ratcliffe went to do in Moscow. The reported message is defensive: do not exploit our weakness, do not touch the Baltics. That is the posture of a power managing its exposure, not of one preparing a major ground offensive.
An answer, as clear as the data allow
A major resumption is more likely than not within six months. But its form and date follow from the constraints, and they do not point to an American ground operation before 3 November. Before 27 October, the most plausible escalation is Israeli, aerial or covert, with a low political footprint for Washington. After 3 November, no electoral safeguard disciplines the decision any longer.
The main risk is not a deliberate decision — it is the misread incident: a tanker, a drone, a strike attributed to the wrong actor, in a system where de-escalation channels have been reduced to the point that an intelligence director has to be sent in person to get a message across. That is precisely what the 25 August visit signals.
VI — Settlement fragments, reserves do not get replaced
On the monetary side, a single distinction does all the analytical work: that between the settlement currency, the one you pay for a cargo with, and the reserve currency, the one in which a central bank stores its wealth over ten years. The first is won through convenience. The second requires foreigners to agree to hold your debt, in quantity, without being able to leave freely.
| Indicator | Level | Trend |
|---|---|---|
| Share of SWIFT payments | ≈ 3.5% | Rising — 2% in 2023 |
| Share of trade finance | ≈ 6% | Sharply rising |
| CIPS daily volume (Q1 2026) | ≈ $940bn | Record of $178.5bn in a single session |
| Share of global FX reserves | 1.93% | Falling — 2.83% in early 2022 |
That is the heart of the matter, and it fits in the last line. The yuan advances wherever the point is to pay, and retreats wherever the point is to hold. This is not a contradiction: it is the result of a choice.
In February 2026, Xi Jinping set out in the Party's theoretical journal the most explicit goal to date of a "strong currency" able to contest the dollar's dominance. At the same time, Beijing maintains a closed capital account, limited convertibility, and refuses any significant appreciation of its currency. The PBoC itself describes this caution as "a balance between development and security".
The ambition is thus displayed; it is the price that is refused. That price has a name: the Triffin dilemma. For the world to hold your currency as a reserve, you must supply it in growing quantities, and therefore accept lasting external deficits, a structurally strong currency, and free capital outflows. Each of these three terms flatly contradicts a Chinese model built on trade surpluses, export competitiveness and control of financial flows.
Beijing wants the settlement function without the reserve burden. That is not indecision: it is mercantilist consistency.
Gold, and what it does not prove
The People's Bank of China has bought gold for twenty consecutive months, the longest streak since at least 2015, including 14.93 tonnes in June 2026, its largest monthly purchase since 2023. Official reserves reach 2,346 tonnes, up about 40 tonnes since January. These purchases were made during gold's worst quarter in thirteen years.
Buying the dip, month after month, without communication: that is the behaviour of a reserve manager diversifying away from the dollar, not that of a state preparing a currency conversion. The nuance matters, because it is systematically erased.
Since 24 July 2026, the major Chinese banks (ICBC, Postal Savings Bank, Ping An, China Guangfa) have stopped offering retail clients access to spot and deferred-delivery contracts on the Shanghai Gold Exchange. Margin requirements have been raised to 120–140%.
This is probably the most misread measure of the year. It is regularly presented as the ban on "paper gold", a prelude to a gold backing of the currency. What it really is: the closure of the retail leverage channel. Chinese individuals can still buy bars and coins, subscribe to gold savings plans and hold ETFs backed by the metal. The SGE remains open; futures contracts continue to trade in Shanghai. The context makes the decision plain: gold, after a peak near $5,600 an ounce, had lost about 30%, briefly dropping below $4,000. Pulling leveraged savers out of a market in free fall is a classic prudential measure.
As for the BRICS common currency: the "Unit", a digital settlement instrument backed by 40% physical gold and 60% bloc currencies, exists as a project and pilot, launched on 31 October 2025, and is the official policy of no member. The real work streams of 2026 are the extension of BRICS Pay, the interoperability of central bank digital currencies and the NDB's local-currency lending. That is considerable. It is not a currency.
VII — What all this sketches
The thread is not a plot, it is a convergence of constraints, which is both less romantic and more worrying, because constraints are verifiable.
A dominant power weakened on all three pillars simultaneously: military, energy, financial (Chinese holdings of US debt have fallen from more than $1,300bn in 2013 to $633bn in June 2026). But a weakened power that, in the same movement, secures control of the planet's second-largest heavy-crude deposit and redirects 786,000 barrels a day to its own refineries. The weakness is real; it is not passive.
Facing this, middle powers that stop waiting for a guarantee and take out their own insurance: that is the Mecca Pact. And a systemic competitor that accumulates metal and builds payment pipes, while stubbornly refusing to shoulder the burden that would make it the next monetary hegemon.
In this picture there is no successor forming. There is a disintegration without relief: an order that loses its capacity for constraint faster than another acquires its own. This is the least stable configuration of all, because it has no arbiter. And that is why the calendar matters so much: in a system without an arbiter, the remaining safeguards are domestic. They have an expiry date, set for 3 November.
VIII — Five scenarios, and what they do to prices
Before the scenarios, a prerequisite without which all the directions below would be wrong: the transmission channel has changed.
For fifteen years, a geopolitical shock traded one way: equities down, bonds up, the central bank coming to the rescue. That mechanism assumes a central bank free to respond to growth. That is no longer the case. With CPI inflation at 3.4%, PCE at 3.7%, a Fed led by Kevin Warsh who has set his bar on inflation, and three dissenters who voted for a hike in July, an oil shock no longer produces easing: it produces imported inflation facing a Fed that cannot respond.
An energy shock in an inflation-first regime pushes equities and bonds down. There is no bond hedge.
A second prerequisite, specific to FX: in an oil shock, what separates currencies is not safe-haven status, it is the terms of trade. Net producers collect (USD, CAD, NOK); net importers pay (JPY, EUR, and most emerging markets). That is why the yen, a classic haven, behaves badly in this kind of shock: Japan imports almost all of its crude. The Swiss franc remains a clean haven.
| Scenario | Prob. | Brent | Gold | US equities | EU equities | Long yields | DXY | EUR/USD | USD/JPY |
|---|---|---|---|---|---|---|---|---|---|
| Tense status quo | 40% | → | → | → | → | → | ↑ | ↓ | ↑ |
| Israeli escalation (before 27 Oct) | 20% | ↑↑ | ↑ | ↓ | ↓ | ↑ | ↑ | ↓ | ↑ |
| US escalation (after 3 Nov) | 15% | ↑↑ | ↑↑ | ↓↓ | ↓↓ | ↑↑ | ↑↑ | ↓↓ | ↑↑ |
| Baltic shock | 10% | ↑ | ↑↑ | ↓ | ↓↓ | ↑ | ↑↑ | ↓↓ | ↑ |
| De-escalation — Hormuz reopens | 15% | ↓↓ | ↓ | ↑↑ | ↑↑ | ↓ | ↓ | ↑ | ↓↓ |
The weightings are subjective and are revised with each data point. The directions follow from the mechanisms described above.
Hormuz remains partly obstructed, strikes continue at low intensity, neither electoral calendar is disrupted. Geopolitics stops being the driver; inflation becomes it again.
- Energy
- Brent in an 85–95 range. The Mars–Brent differential stays high; that is where the carry is.
- Equities
- Trendless indices, wide dispersion. Energy, defence and Gulf refiners outperform; airlines and chemicals suffer.
- Metals
- Gold supported without momentum after the 30% correction.
- FX
- Dollar firm on rate differentials: a Fed that cannot cut is a currency that cannot fall much.
Low political footprint for Washington, real but bounded market effect. The market tests a closure of Hormuz without believing in it.
- Energy
- Brent +8 to +15% towards 100–105, then partial retracement. The cleanest trade is not Brent but the widening of the differential on heavy sour crude.
- Equities
- SPX −3 to −6%, Europe more. Marine insurers and airlines on the front line.
- Metals
- Gold +5 to +8%. Silver amplifies in both directions.
- FX
- USD, CAD, NOK and CHF rise; EUR and JPY fall, the two big importers. Do not buy the yen on this shock.
Low probability, maximum impact. A fifth of global supply disappears. The only scenario where equities and bonds fall together and durably.
- Energy
- Brent 130–150 and beyond depending on duration. European gas dragged along by substitution. Refined products tighter than crude.
- Equities
- −12 to −20%. Long-duration stocks suffer twice over, from rates and from growth.
- Rates
- The central trap: long yields rise while growth collapses. The 60/40 protects against nothing.
- Metals
- Gold above its previous peak, the only hedge that works here.
- FX
- DXY sharply higher. EUR/USD broken, USD/JPY very high. Importing emerging markets (INR, TRY, ZAR) most exposed.
The scenario least covered by portfolios. Oil is not the relevant instrument: gas is.
- Energy
- Brent moderately bullish. The violent move is in European gas and electricity.
- Equities
- Europe −8 to −15%, far beyond the United States. European defence sharply higher.
- Rates
- Flight to the Bund, tempered by the expectation of defence issuance. Peripheral spreads widen; that is where the stress shows.
- FX
- EUR sharply lower, USD and CHF sharply higher. PLN, HUF and CZK are the weak link.
The scenario nobody is positioned for. Its effect is twofold: the fall in crude is disinflationary, which finally gives the Fed back its room for manoeuvre, with support from both earnings and rates.
- Energy
- Brent −15 to −25% towards 65–72. The Venezuelan lock loses most of its value.
- Equities
- +5 to +8%, led by cyclicals. Energy underperforms.
- Metals
- Gold pulls back on profit-taking, but central bank buying does not stop.
- FX
- Dollar lower, and above all the yen is the big winner, the best terms-of-trade improvement in the G10.
Three reflexes that no longer work
- Buying bonds on geopolitical risk. Valid when the central bank can ease. Facing a supply shock with PCE at 3.7% and an inflation-first Fed, duration amplifies the loss instead of offsetting it.
- Buying the yen in risk-off. The reflex ignores that Japan imports almost all of its energy. In an oil shock, terms of trade outweigh haven status.
- Treating gold as a fear asset. The 30% correction from the peak near $5,600 an ounce showed it: gold responds first to real yields and the dollar. It hedges reliably in only one of the five scenarios.
What would invalidate this reading
- A major escalation before 27 October would indicate that the electoral calendar weighs less than this article claims.
- An official Iranian confirmation of joining the Mecca Pact would move the agreement from insurance policy to regional tipping point.
- An opening of China's capital account is the only signal that Beijing finally accepts the price of reserve status.
- A lasting reopening of Hormuz would strip the Venezuelan lock of most of its strategic value.
- A Russian operation against a Baltic state would shift the crisis's centre of gravity towards Europe.
Four series can be read almost in real time and are worth more than commentary: the Mars–Brent differential (heavy-crude tightness), marine insurance premiums in the Gulf (they anticipate closures), the Dutch TTF (thermometer of European risk), and the pace of Venezuelan exports to US refineries (solidity of the lock). A divergence between these series and the dominant narrative generally precedes the repricing.
Methodology note and sources
Each factual block in this article carries its verification status: Established for a fact corroborated by reference media, Reported for single-source or unconfirmed information, Speculative for a claim or projection with no identifiable source. The scenario weightings and asset directions in the final section are explicit conditional reasoning, not forecasts: they describe transmission mechanisms, not price paths. This article is an analysis for information purposes; it is neither investment advice nor a recommendation to take a position.
- Mecca Joint Defence Agreement
- Al Jazeera — contents of the 7 August agreement
- Jerusalem Post — invitation extended to Iran
- Foreign Policy — the thesis of the pact without reach
- Washington Post — Ratcliffe's visit to Moscow
- Axios — the trilateral summit hypothesis
- Religion News Service — CIA / Vatican convergence
- Timeline of the 2026 war
- Quinnipiac — 74% oppose ground troops
- Israeli general election of 27 October 2026
- CNBC — SPR below 300 million barrels
- NPR — lowest level in forty years
- NPR — Venezuelan exports after Maduro's capture
- Columbia CGEP — US actions in Venezuela
- Semafor — Xi and the reserve-currency goal
- CEPR — the renminbi's unconventional route
- Kitco — PBoC gold purchases
- BullionStar — exit from retail paper gold
- CNBC — Chinese holdings of US debt
- Warnings from the Baltic presidents