Research — 5 September 2026

Verdict

A 25bp hike on September 16 is the base case for most desks (~2/3 priced); the disagreement is about what follows — BNY argues one hike may bring the effective rate close to a ~4% neutral, while others see additional hikes into 2027.

Analysis

The causal chain runs through two data prints — August NFP (JPM's Feroli: 50k vs Street 55k, unemployment 4.1%) and CPI on Sept 11 — into the Sept 16 decision. The counterargument matters: UBS's base case is still a hold on cooling underlying inflation, and dissents from dovish Governors are plausible. The critical nuance from BNY is the terminal: OIS implies a peak of ~4.2% and 10y-horizon expectations ~4%, while a hike takes the effective rate to 3.88% — so if inflation keeps easing toward 2%, one hike may exhaust the cycle, which would cap front-end and dollar upside. Because NFP now operates in a 'good news is bad news' frame (JPM), a strong print lifts yields and pressures equities, while a large downside miss reignites stagflation concerns — an asymmetric setup for both rates and equity beta into mid-September.

Tension

BNY/MUFG see the hike bar effectively cleared barring weak data; UBS holds that core PCE near 0.2% m/m still permits a hold; JPM still officially forecasts December.

What changed

Warsh walked back two concerns from his July press conference (target credibility, tool choice), sharpening the message that higher inflation meets a higher funds rate, and markets moved from ~36% to ~2/3 September-hike pricing.

Desk views

BNY expects a September hike unless CPI slows significantly, then argues the terminal may be near: OIS peak ~4.2% vs effective rate of 3.88% post-hike means 'we may not need much more than one hike.'

Evidence IDs · P003

JPM (Feroli) notes Warsh walked back two July concerns, sharpening the higher-rates message; JPM still forecasts a December hike but sees September risks as having risen, with August employment and CPI as the deciders.

Evidence IDs · P024

UBS is the outlier: base case remains a hold, as core PCE ran close to 0.2% m/m in June and July and employment growth has moderated; risk of a hike has increased but disinflation in 2H should support lower rates in 2027.

Evidence IDs · P038

JPM desk frames NFP as 'good news is bad news': a stronger print pushes yields up and equities down; Feroli forecasts 50k jobs vs Street 55k, with Goldilocks in the 30k–70k range.

Evidence IDs · P029

MUFG: 'sufficient speed' language shifts the balance of risks to a hike, but the Board's dovish tendency means NFP and the Sept 11 CPI could shift risks back to a hold; USD gains from the speech were modest (0.3% net).

Evidence IDs · P032

StanChart: market September-hike probability rose to c.60% from 36% post-speech; USD broke its 200DMA, gold fell 3.1%; Warsh also abandoned forward guidance, raising reaction-function uncertainty; STC sees inflation reaching 2% by mid-2027, limiting hikes to one this year.

Evidence IDs · P037

Investor implication

Position for a hawkish September hike as base case but do not overpay for a long hiking cycle; the CPI print on September 11 is the highest-information event for rates, dollar and equity beta.

Trade expression

No forced trade; keep front-end rate exposure hedged and treat the Sept 11 CPI as the position-sizing event.

Invalidation

A materially soft CPI print or weak NFP that unwinds September-hike pricing.

Verdict

Bearish long-end pressure persists (persistent inflation, fiscal concerns, energy risk), but Treasury buybacks and Fed credibility signals cap term premium near-term; the debasement trade (gold, anti-dollar) has structural staying power.

Analysis

August delivered multi-decade long-end highs: US 30y at 5.31% (post-2007 high, Aug 17), German 30y at 3.81%, Japanese 30y at 4.14%; 10y Treasury at 4.75% (highest since Jan 2025). BNY frames this as a 'battle of attrition' between bond markets and policymakers, with rates — not equities — the key transmission channel, and intervention/policy signaling increasingly unable to reverse adjustment without a fiscal or inflation-outlook change. The Treasury's long-end buyback 'twist' (announced Aug 19, running Sept 9–Nov 4, bills issued to fund long-end purchases) is the interventionist counterweight: JPM sees no near-term catalyst to drive term premium meaningfully higher, and ING argues higher long-end yields are now read through the lens of potential Treasury interventionism, feeding the debasement trade that even a hawkish Warsh could not fully unwind. UniCredit adds the collision risk: a Fed hiking into a Treasury that is trying to compress long-end costs into the Nov 3 midterms could become a medium-term USD drag and revive 'sell America'/de-dollarization. The counterargument: JPM technicals flag a weekly momentum-divergence buy signal on 30y bonds that historically precedes multi-month mean reversion to lower yields, and UniCredit's Weekly Supply Preview shows heavy but well-absorbed EGB supply. Investor consequence: duration is no longer reliable portfolio ballast (BlackRock: >80% of the global bond universe yields above 4%), so income is attractive but duration risk is not.

Tension

JPM sees both fiscal and monetary policy 'effectively capping term premium' near term, while BNY sees attritional upside pressure on yields that intervention cannot reverse — a genuine disagreement on the long-end trajectory.

What changed

Treasury's Aug 19 buyback announcement plus Warsh's speech produced the largest weekly curve move of the past year (2s/10s flattening), and the dollar failed to hold post-speech gains.

Desk views

BNY: geopolitics is now an attritional backdrop; persistent inflation, fiscal concerns and energy risk keep investors demanding greater duration compensation, and intervention struggles to reverse it without a change in fiscal/inflation outlook.

Evidence IDs · P002

ING: the dollar gave back half its post-speech gains despite front-end support; markets read higher long-end yields as Treasury-interventionism risk, feeding the debasement trade — but ING expects hawkish Fed expectations to cement this week and DXY to reclaim 100.0 barring fresh intervention news.

Evidence IDs · P022

A numerical detail from this desk view was omitted because it could not be matched exactly.

Evidence IDs · P040

JPM rates: curve twisted flatter; with the Fed and Treasury effectively capping term premium, JPM sees no near-term catalyst for higher term premium and unwinds 2s/10s steepeners; still forecasts a 25bp December hike with risen September risk.

Evidence IDs · P026

JPM technicals: 30y retains bearish medium-term trend but a weekly momentum-divergence buy signal has triggered, historically preceding multi-month mean reversion to lower yields — a tactical long-end counterpoint.

Evidence IDs · P027

Investor implication

Avoid treating long-duration government bonds as ballast; own income selectively at the front/middle of curves, and keep gold/anti-fiat exposure as the structural hedge the debasement narrative demands.

Trade expression

Relative-value over outright: curve flatteners per JPM; gold as the debasement hedge rather than short-duration bets alone.

Invalidation

A credible fiscal consolidation signal or a rapid disinflation path that removes the duration-compensation argument.

Verdict

Structurally bullish bias on debasement and official-sector buying; tactically vulnerable to a hawkish Fed (gold fell 3.1% on Warsh's speech) — buy drawdowns, don't chase.

Analysis

RBC's Gold Strategy argues macro drivers alone cannot explain current prices, but gold's haven/store-of-value/non-debaseable framing fits the environment: investor inflows returned 'in size' (north of 200 tons this year) and central banks resume at over 700 tons this year and next. The mechanism is the debasement trade documented by ING and UniCredit: rising long-end yields read as fiscal/monetary collision risk push capital to non-debaseable assets. The tension: a confirmed September Fed hike and firmer real front-end rates are a direct near-term headwind — gold is ~6% off its August high (JPM market intel) — and RBC's own middle-scenario forecasts (Q3 4,558; Q4 4,370 $/oz) sit near spot rather than implying a large upside gap, meaning much is priced.

What changed

Central-bank flows resumed after an earlier pause, and the Treasury buyback announcement gave the debasement narrative a concrete catalyst.

Desk views

RBC (high conviction, unchanged): investor inflows north of 200 tons this year; central banks over 700 tons this year and next; also cites a negative correlation between gold and Trump approval, eyeing midterms.

Evidence IDs · P036

ING: the debasement trade is the lens through which markets read rising long-end yields; it has 'still not been fully unwind' even after the hawkish Warsh repricing — a persistent bid under gold.

Evidence IDs · P022

Investor implication

Treat gold as the portfolio hedge against the fiscal-monetary collision scenario rather than a momentum trade; expect chop into the September FOMC.

Trade expression

Buy tactical weakness around hawkish Fed repricings as the structural flow thesis remains intact.

Invalidation

A durable Fed-tightening cycle with rising real yields AND a halt in central-bank purchases.

Verdict

Structurally higher Japanese yields and a weak-but-policy-capped yen (USD/JPY 155–165 range base case); GPIF allocation shifts, not FX intervention, are the realistic incremental support tool.

Analysis

Three forces converge in Japan. First, the bond repricing: 10y JGB yields hit 3% for the first time since 1996 (BNY/DB), part of the global long-end sell-off but amplified by fiscal supply — Mizuho reports FY2027 general account budget requests expected to exceed JPY130tn (vs JPY122.4tn prior year, an increase of at least ~JPY7-8tn, with METI seeking JPY7.7tn and Defense JPY8.9tn). BNY's attrition frame applies: intervention and signaling 'will increasingly struggle to reverse it without a change in the underlying fiscal or inflation outlook.' Second, the yen: JPM estimates net JPY selling of ~¥16.4tn ($103bn) since the Takaichi administration took office (October 2025), meaning large short positions remain outstanding — a source of sharp-cover risk in stronger-yen scenarios. JPM's base case is USD/JPY 155–165 with targets of 160 (3Q26) and 164 (4Q26), expecting BOJ hikes roughly once per quarter; above 160, the most realistic additional measure is a GPIF allocation shift that limits JPY downside gradually rather than spiking it. Third, spillover: rising JGB yields reprice the global duration floor and pressure Japan-heavy equity duration hedges. The investor consequence: Japanese duration is a supply-and-policy short, while yen shorts are crowded enough that any BOJ acceleration or coordinated intervention (Treasury 'defended its interest in intervening jointly with Japan' per Natixis) creates asymmetric squeeze risk.

Tension

BNY sees the yield rise as near-irreversible; JPM sees policy (BOJ hikes, possible GPIF shift) capable of stabilizing — but both agree the adjustment pressure is upward.

What changed

10y JGB crossed 3% for the first time since 1996, coinciding with outsized FY2027 budget requests and a Treasury statement on joint intervention.

Desk views

BNY: Japan is the clearest additional pressure point; 10y JGB at three-decade highs; intervention slows but cannot reverse the adjustment absent fiscal/inflation change.

Evidence IDs · P002

JPM: base case USD/JPY 155–165 (targets 160 3Q26, 164 4Q26) with BOJ hikes roughly once per quarter; net JPY selling of ~¥16.4tn since Oct 2025 means large shorts outstanding; GPIF allocation shift is the most realistic correction mechanism above 160.

Evidence IDs · P030

Mizuho: FY2027 general account requests expected above JPY130tn, at least ~JPY7-8tn above the prior year's JPY122.4tn, growth/crisis-investment led — a supply overhang for JGBs.

Evidence IDs · P033

Investor implication

Underweight JGB duration on supply/fiscal grounds; hedge rather than stack yen carry, given the documented short-position overhang.

Trade expression

Avoid adding yen carry at the top of the range; JGB curve steepeners express the fiscal-supply view.

Invalidation

A large GPIF allocation shift or BOJ acceleration that breaks USD/JPY below 155.

Verdict

Fundamentals are exceptional (11.3% 2026 GDP growth, current account 23.6% of GDP, TWD appreciation path to 31.6), yet positioning is stretched the other way — a classic disconnect between macro momentum and fund flows.

Analysis

The CBC is expected to deliver a 12.5bp hike in December 'given strong growth and rising inflation concerns,' and a record-high basic balance plus renewed foreign equity inflows underpin a USD/TWD forecast of 31.6 by end-2026. The mechanism that matters for global investors: Taiwan is the bellwether of the AI hardware cycle, so the GDP upgrade corroborates the capacity-expansion phase of the buildout (also visible in GS's Nvidia revenue-doubling note), while the flow divergence means expectations for AI hardware earnings may still have room to re-rate — or that investors are pricing a 2027 deceleration (CACIB forecasts GDP slowing to 3.8% in 2027).

Tension

Strongest-in-class macro revision vs the largest EM underweight increase — either positioning or fundamentals must give.

What changed

CACIB raised 2026 GDP from 8.2% to 11.3% and the current account to 23.6% of GDP; TSMC raised 2026 capex in mid-July.

Desk views

CACIB: growth broadening beyond exports; CBC 12.5bp hike expected in December; TWD appreciation to 31.6 by end-2026 on record basic balance and renewed foreign equity inflows.

Evidence IDs · P008

A numerical detail from this desk view was omitted because it could not be matched exactly.

Evidence IDs · P023

GS: AI infrastructure demand 'extraordinarily robust' with Nvidia revenue doubling y/y and ~70% growth forecast next year, supply constrained — corroborating the capex boom driving Taiwan's upgrade.

Evidence IDs · P017

Investor implication

Taiwan/EWT exposure is a leveraged AI-capex play with positive macro momentum but crowded-out positioning; entry on flow-driven weakness is better supported than chasing strength, with the 2027 deceleration as the terminal risk.

Trade expression

Prefer direct Taiwan/semi exposure over broad EM, sized to the 2027 deceleration risk.

Invalidation

A capex-guidance cut from TSMC or renewed Chinese export-control escalation.

Verdict

Crowded EM reallocation into Brazil/China against near-term outflows and a hawkish Fed is a fragile setup; relative value (CACIB's long BRL/MXN) beats outright beta into Brazil's October election.

Analysis

The mechanism is a rates-driven search for real-yield and debasement hedges (BNY notes Latin America and safe havens performed best in August, suggesting 'expectations of lower U.S. real rates are driving up interest in the region'), but the hawkish Warsh repricing cuts against the LatAm carry thesis — BNY explicitly warns Warsh's comments 'may slow the momentum.' CACIB's trade desk resolves this tactically: long BRL/MXN 3-month forward (ref 3.284, target 3.3770, stop 3.245) because MXN is more exposed to a Fed-hiking series while BRL volatility is concentrated around the October election; and short PEN/CLP purely on valuations. The investor consequence: EM beta into a September Fed hike is a bad-carry setup with crowded positioning; idiosyncratic, election-aware relative-value expressions carry better risk/reward.

Tension

Consensus allocation says overweight Brazil; realized flows and JPM's own downgrade say de-risk — and the Fed's path is a direct headwind to the LatAm real-rate story.

What changed

July consensus data shows the largest month-over-month Brazil/China overweight increases, colliding with the post-Jackson Hole hawkish repricing.

Desk views

A numerical detail from this desk view was omitted because it could not be matched exactly.

Evidence IDs · P023

BNY: Latin America and safe havens led August flows (not carry); expectations of lower US real rates drove regional interest, but Warsh's comments 'may slow the momentum.'

Evidence IDs · P002

CACIB: tactical relative value over direction — long BRL/MXN 3M forward (ref 3.284, TP 3.3770, SL 3.245) on MXN's Fed sensitivity vs BRL's election-date volatility; short PEN/CLP on valuations.

Evidence IDs · P005

Investor implication

Do not chase EM beta into a hawkish September FOMC; if expressing the LatAm real-rate thesis, use relative value with election-aware stop-losses.

Trade expression

CACIB-style relative value (long BRL/MXN) over outright EM longs into September.

Invalidation

A dovish Fed surprise or Brazilian polls breaking decisively would undermine the relative-value logic.

Verdict

A European-specific energy squeeze (Qatar LNG barely flowing, Strait of Hormuz constraints) keeps energy-driven inflation alive in the eurozone independent of the Fed path, complicating the ECB and pressuring energy-importer equities.

Analysis

UniCredit documents that TTF has surged to a 50% premium over Brent in comparable terms — versus an average 6% discount from 2023 to present — with only minimal amounts of Qatar's ~9bcm monthly LNG production being exported, in a market where ~20% of global oil and LNG flows transit the Strait of Hormuz. The counterargument: energy markets have 'continued to adapt to the disruption' (UBS), and the premium is partially seasonal/supply-idiosyncratic rather than demand-driven. For investors, this is a differentiated channel from US rates: it hits European industrials and importers' margins while supporting energy exporters and inflation-linked assets, and it independently constrains the ECB even as the Fed dominates headlines.

What changed

TTF decoupled from Brent since end-June on collapsed Qatari LNG exports, reaching a 50% oil-equivalent premium not seen since early 2023.

Desk views

A numerical detail from this desk view was omitted because it could not be matched exactly.

Evidence IDs · P039

A numerical detail from this desk view was omitted because it could not be matched exactly.

Evidence IDs · P035

A numerical detail from this desk view was omitted because it could not be matched exactly.

Evidence IDs · P017

Investor implication

European energy inflation is an independent variable from the Fed: own energy exposure and inflation links against European importer margins, and expect the ECB to stay hawkish regardless of US developments.

Trade expression

Overweight European energy and inflation-linked exposure; underweight gas-intensive European industrials until the TTF-Brent premium normalizes.

Invalidation

A Hormuz reopening plus restored Qatari LNG flows that collapses the TTF premium.