Five Handle: The Fed's Hawkish Hike Sends the 10-Year Above 5% While Oil Hands Back the Pipeline Premium

Scorecard: Last Week's Calls

Leading with the wins.

Fed hikes 25bp on 16 September, Medium to High: Correct, and unanimous. The FOMC lifted the target range to 3.75 to 4.00 percent on a 12-0 vote, the first hike since July 2023 (Federal Reserve, 16 September). Our flip from the hold paid.
EUR/USD small short toward 1.14, cover below 1.13, Medium: Correct, target hit. The pair fell from about 1.154 to a low of 1.1376 on 24 September and never threatened the 1.13 cover (Wise, 24 September; TradingEconomics, 24 September). Best FX call in the book.
Dollar firm in the high 90s to low 100s, Medium: Correct. DXY ran from about 99.6 to roughly 101.1, its highest since late July, as October hike odds climbed (FXStreet, 23 September).
Keep front-end duration light, buy the 2-year only on "one and done" guidance: Correct, and the condition never fired. Chair Warsh offered no forward guidance and said he was hard pressed to call financial conditions restrictive (CNN, 16 September); the 2-year rose 20bp to 4.85 percent, so staying light saved that move (US Treasury, 23 September).
India: hold, pause adds, trim above 96.5, Nifty 23,000 to 24,000, Medium: Correct. The Nifty closed 23,446.80 on 23 September and USD/INR closed 95.74, inside both ranges and clear of the trim line (Business Standard, 23 September; Reuters, 24 September).
Credit: overweight IG, underweight HY, no scale-in below 3.25 percent OAS, Medium: Correct and untriggered. HY OAS sat at 2.66 to 2.68 percent, unchanged through a 5 percent 10-year (FRED/ICE BofA, 21 September). The underweight hasn't paid yet; it hasn't cost either.
Do not fight mega-cap AI: Correct. The Nasdaq set a record close of 27,244 on 22 September and the S&P 500 finished 23 September at 7,706, up about 1.6 percent on the week (Vantage, 23 September; CNBC, 23 September). The energy hedge was a drag with Brent down about 4 percent.
Gold, contrarian add below $4,300 for a post-decision rebound, Medium: Correct, if modest. Gold bounced from about $4,263 to $4,364 on 18 September and a $4,369 close on 21 September before fading to about $4,304 (CNBC, 18 September; Rio Times, 22 September; USAGOLD, 23 September). The add is roughly one percent in the money.
BoJ hikes to 1.25 percent on 17-18 September: Correct. The BoJ raised its policy rate to 1.25 percent on 18 September on a 7-2 vote, the highest since 1995 (CNBC, 18 September).
Cautious on gilts into the Budget, Low: Correct. The 10-year benchmark yielded 5.33 percent on 23 September, up from about 5.29 percent, and every conventional gilt sold off that day (giltsyield.com, 23 September).

Now the mixed.

Oil core long, trim toward $110, re-add toward $100, High: Mixed. Brent's high was $109.45 on 15 September, so the trim missed by 55 cents; the 22 September low of $97.36 triggered the re-add; Brent then settled at $103.08 on 23 September (Investing.com, 23 September). The core long lost about 4 percent on the week, the re-add is ahead, and the discipline rule was set 55 cents too high.
Middle East base case: pipeline restarts within one to two weeks, Hormuz throttled, Brent $105 to $112: Two of three. The East-West line restarted on 22 September, eleven days after the strike, and the strait stays throttled, but Brent broke the band, trading $97 to $103 from 22 September (Reuters via Hydrocarbon Processing, 22 September; Investing.com, 23 September).

The misses, plain.

Fed base case, 60 percent: hike with a balanced-to-hawkish statement: Wrong. The 25 percent "downside" happened: 16 of 18 participants pencilled another 2026 hike, the median dot moved to 4.1 percent for end-2026, and the 2-year pushed toward our 4.8 percent marker (CNBC, 16 September; J.P. Morgan Asset Management, 16 September; US Treasury, 23 September). We had the hike right and the tone wrong.
Yen: trail the short USD/JPY toward 150, cut above 161, Medium: Wrong. The BoJ's hike read as dovish, the yen fell, and USD/JPY reached 158.28 on 23 September (Babypips, 18 September; TradingEconomics, 23 September). The 161 stop did not trigger, but the trailing leg lost, and we flag a data problem of our own: the 154 we printed last week does not square with the pair trading above 157 straight after the BoJ, so the mark was worse than we thought going in.
Constructive GBP above 1.34, add dips toward 1.34, Low: Wrong. Cable broke 1.34 on 16 September and sits near 1.333 (MTFX, 23 September). The dollar did it, not sterling, but a wrong level is a wrong level.
Neutral-to-small-long AUD, buy dips toward 0.70, trim toward 0.725, Medium: Wrong so far. The Aussie fell from about 0.721 to roughly 0.709, with a 0.7075 low, as the dollar surged (Investing.com, 24 September). The 0.70 dip we said we would buy is now a few pips away.

Too early to grade.

RBA hike the more probable next move on 28-29 September: Too early, and firming. Every big-four bank now forecasts a 25bp hike to 4.60 percent and market pricing sits near 74 percent (investingLive, 23 September; RBA Rate Watch, 24 September).
Threshold: 10-year above 5.0 percent, add long-end duration: Tripped on 23 September at a 5.11 percent close; a first add at 5.0 percent is about 11bp under water and has held one close (US Treasury, 23 September).

Hit rate: ten clean wins, four misses, two mixed, two too early. The stars were the EUR/USD short and the front-end discipline. The weak spot was the dollar: we underestimated how far a hawkish Fed would carry it, and the yen and sterling calls paid for that.

We also correct two calendar errors from the last edition: the FOMC meets 27-28 October and 8-9 December, not 28-29 October and 17-18 December (Federal Reserve calendar).

Key Levels Dashboard

Levels are the 23 September close or 24 September intraday. Changes are versus the levels printed in the 15 September report.

Brent crude, ~$103/bbl, down ~4% on the week. Settled $103.08 on 23 September after a $97.36 low on 22 September, when the Saudi pipeline restarted and Iran floated a Hormuz reopening; rebounded on Pezeshkian's UN speech (Investing.com, 23 September; TradingEconomics, 24 September).
WTI crude, ~$93/bbl, down ~6% on the week. Tracked Brent lower; the EIA reported a 3.0 million barrel crude build (TradingEconomics, 23 September; EIA via Reuters, 23 September).
Gold spot, ~$4,304/oz, up ~1% on the week. Bounced to $4,369 after the FOMC, then faded as the dollar and yields rose (USAGOLD, 23 September).
Silver spot, ~$65.1/oz, up ~4% on the week. Fell 3 percent on 23 September; gold/silver ratio near 66 (USAGOLD, 23 September).
Federal Reserve funds target, 3.75-4.00%, up 25bp on the week. Hiked 16 September on a 12-0 vote; a further hike on 28 October is about 70 to 73 percent priced (Federal Reserve, 16 September; CNBC, 23 September).
ECB deposit rate, 2.50%, unchanged. Pricing for a 29 October hike slipped to about 60 percent from roughly 73 percent (ECB Watch, 24 September).
BoE Bank Rate, 3.75%, unchanged. Held 6-3 on 17 September with three votes for 4.00 percent; long-dated gilt sales halted (Bank of England, 17 September).
RBA cash rate, 4.35%, unchanged. A 29 September hike to 4.60 percent is about 74 percent priced and forecast by all four majors (RBA Rate Watch, 24 September; investingLive, 23 September).
RBI repo rate, 5.25%, unchanged. Next decision 7 October (5paisa MPC schedule).
Latest US jobs print, August +162k, unemployment 4.1%, unchanged. Initial claims fell to 196k in the week to 12 September; next payrolls 2 October (Department of Labor, 17 September).
S&P 500, ~7,706, up ~1.6% on the week. Recovered its post-FOMC losses on AI strength, then fell 0.75 percent on 23 September as the 10-year touched 5.135 percent (CNBC, 23 September).
US 2-year Treasury, ~4.85%, up ~20bp on the week. Repriced for a second hike (US Treasury, 23 September).
US 10-year Treasury, ~5.11%, up ~12bp on the week. First close above 5 percent since July 2007; 5.135 percent intraday (US Treasury, 23 September; CNBC, 23 September).
US 30-year Treasury, ~5.40%, up ~5bp on the week. The curve bear-flattened; the front end led (US Treasury, 23 September).
DXY, ~101.1, up ~1.5 on the week. Highest since late July (FXStreet, 23 September).
EUR/USD, ~1.138, down ~1.4% on the week. Hit our 1.14 target; low 1.1376 on 24 September (TradingEconomics, 24 September).
GBP/USD, ~1.333, down ~1.4% on the week. Broke 1.34 on 16 September and stayed below (MTFX, 23 September).
AUD/USD, ~0.709, down ~1.7% on the week. Month low 0.7075 despite rising RBA hike odds (Investing.com, 24 September).
USD/INR, ~95.7 to 96.0, roughly flat on the week. RBI intervention capped the move; a deputy governor flagged the case for rupee gains (Reuters, 24 September; TradingEconomics, 24 September).
USD/JPY, ~158.3, yen down ~2.7% versus our prior print. The BoJ's dovish hike sank the yen; a reported MoF rate check puts intervention on the table (TradingEconomics, 23 September; Babypips, 18 September).
Nifty 50, ~23,447, up ~0.2% on the week. Metals led, IT lagged; the 23,400 floor held (Business Standard, 23 September).
US HY OAS, ~2.66 to 2.68%, flat on the week. No widening through a 5 percent 10-year (FRED/ICE BofA, 21 September).
UK 10-year gilt, ~5.33%, up ~4bp on the week; 30-year ~5.82%, down from near 6 percent after the BoE halted long-dated sales. Added this week because the BoE's QT change is material (giltsyield.com, 23 September).

Executive Summary

The shock changed address this week. Last edition the story was a severed Saudi bypass and Brent at $108; this edition it is a Federal Reserve that hiked and kept the door open, and a Treasury market that priced it. The 10-year closed at 5.11 percent on 23 September, its first close above 5 percent since July 2007, after touching 5.135 percent intraday; the 5-year crossed 5 percent for the first time since 2007 the same day (US Treasury, 23 September; CNBC, 23 September; TheStreet, 23 September). The trigger was not oil. S&P Global's September flash PMI showed US activity accelerating at the fastest rate since July 2021 with input costs surging, a run of Fed speakers leaned hawkish, and CME pricing for a second 25bp hike on 28 October jumped to about 73 percent (CNN, 23 September; CNBC, 23 September).

Our Fed call worked and our Fed scenario did not. We got the 16 September hike right, unanimous, to 3.75 to 4.00 percent (Federal Reserve, 16 September). We gave 25 percent odds to a hawkish hike with dots signalling more; that is what arrived. Sixteen of 18 participants see another move this year, the median end-2026 dot sits at 4.1 percent, and Chair Kevin Warsh ran the shortest press conference on record, about half an hour, without a word of forward guidance (CNBC, 16 September; J.P. Morgan Asset Management, 16 September; Bloomberg, 16 September). The 2-year rose 20bp to 4.85 percent. We never bought it, because our condition, one and done, never appeared.

Oil went the other way. Saudi Arabia restarted the East-West pipeline on 22 September at a low pumping rate, with Aramco targeting a return to about 4 million barrels a day; three of the line's 11 pumping stations were damaged and a security source put a full restart at six to eight weeks (Reuters via Hydrocarbon Processing and The Express Tribune, 22 September). Iran floated reopening Hormuz within seven days if Washington eased military pressure and its port blockade, and Trump called his meeting with Iranian envoys "very productive", even as President Pezeshkian told the UN that Iran would not let the strait be used against it (OilPrice, 22 September; TradingEconomics, 23 September; Bloomberg, 23 September). Brent traded $109.45 to $97.36 in eight sessions and settled at $103.08 on 23 September (Investing.com, 23 September). Our core long lost about 4 percent; the $100 re-add rule caught the low.

The dollar was the collateral damage. DXY reached 101.1, its highest since late July, and took the yen, sterling and the Aussie with it (FXStreet, 23 September). The EUR/USD short hit its 1.14 target, the one clean FX win. The yen short lost as the BoJ's 1.25 percent hike was read as dovish and USD/JPY ran to 158 (CNBC, 18 September; TradingEconomics, 23 September). Sterling broke 1.34. Gold, our contrarian add, bounced and then held about $4,304 through a 5.1 percent 10-year and a 101 dollar, which is more resilience than the consensus expected (USAGOLD, 23 September).

The new book: half-size the crude long and expect Brent below $100 within a month; keep the 2-year purchase on hold and expect a second Fed hike in October; take the contrarian side on the long end in small size; close the yen short; flip sterling to neutral; run half the euro short toward 1.125; resume staggered India adds; stay long gold; stay underweight high yield at spreads that make no sense against a 5 percent risk-free curve.

Signal-Filter Takeaways for Clients

The binding constraint on risk assets is now the discount rate, not the barrel. A 10-year at 5.1 percent with the Fed still hiking means every asset that repriced this week did so for one reason; oil's retreat was the only offset. Watch the 28 October FOMC pricing, not the Hormuz headlines, for the next big move.
Oil's premium has a repair schedule attached to it. The pipeline ramp toward 4 million barrels a day is mechanical and dated, and the diplomatic track now has a concrete Iranian proposal on the table. That skews Brent lower over a month even though the strait stays throttled; hedge the tail, don't chase the spike.
The dollar is doing the tightening for the rest of the world. Sterling, the yen and the Aussie fell on US rates, not on domestic news, so the case for each currency is intact but the timing is hostage to the Fed. We size those positions smaller and set tighter stops until the October meeting is priced.

01 Middle East: The Bypass Comes Back, the Strait Stays Shut

What happened. The biggest change since the last edition is mechanical. Saudi Arabia restarted the East-West pipeline on 22 September at a low pumping rate, with a China-bound cargo scheduled to load at Yanbu and Aramco aiming to return flows to about 4 million barrels a day against a 7 million barrel nameplate (Reuters via Hydrocarbon Processing, 22 September; The Express Tribune, 22 September). Three of the 11 pumping stations were damaged; a security source said 40 percent of capacity takes a couple of days and a full restart six to eight weeks, and Aramco had already told European term customers that October allocations would be zero (The Express Tribune, 22 September; OilPrice, 22 September). Riyadh blamed an Iraqi militia for the 11 September strike (Reuters via The Express Tribune, 22 September). On the diplomatic track, Iran proposed reopening Hormuz within seven days if the US eases military pressure and lifts its blockade of Iranian ports (OilPrice, 22 September). Trump described a meeting with Iranian envoys as "very productive" and said more talks are planned, while President Pezeshkian told the UN General Assembly on 23 September that Iran cannot let others have free access to the waterway while it is used against Iran (TradingEconomics, 23 September; Al Arabiya, 23 September). The war is now nearly seven months old (AP, 24 September). We found no verifiable new development on the Salalah talks or around Bab al-Mandeb this week; both stay on the watch list.

Why it matters. The two-front supply story we described last week is collapsing back to one front. The bypass is returning on a known timetable, and the market has begun to price the repair rather than the outage: Brent's 22 September low of $97.36 came on the restart and the Iranian offer, not on any ceasefire (Investing.com, 23 September). What remains is the strait, and on the strait there is now a written proposal from Tehran and a presidential "very productive" from Washington. That isn't peace. It is enough to cap the premium while the pipeline ramps.

Scenarios

Base case, 50%: the pipeline reaches 40 percent within days and climbs toward 4 million barrels a day over six weeks, talks continue without a deal, Hormuz stays throttled, and Brent drifts $95 to $108.
Downside for prices, 30%: a Hormuz reopening deal on the Iranian seven-day framework, Brent toward $85 to $90.
Upside for prices, 20%: talks collapse or a fresh strike on Saudi or Gulf infrastructure, Brent back through $112.

What to watch. The pipeline's flow rate against the 4 million target, any US response to the Iranian seven-day offer, and whether the Salalah track revives.

House view. The geopolitical premium is decaying on a schedule. We cut the crude core long to half size and keep it as the tail hedge, not the trade. High conviction on the direction of the premium, Medium on the level.

02 Energy Markets: Brent Round-Trips $109 to $97, and Inventories Build

What happened. Brent settled $103.08 on 23 September after a $97.36 intraday low on 22 September and a $109.45 high on 15 September; WTI sat near $93 (Investing.com, 23 September; TradingEconomics, 23 September). The EIA's weekly report for the week to 18 September showed a 2.97 million barrel commercial crude build to 426.4 million against expectations of a 641,000 barrel draw, with gasoline down 1.69 million barrels, distillates down 0.43 million and refinery utilisation falling 2.8 points to 94.0 percent (EIA via Reuters and investingLive, 23 September). The SPR fell a further 405,000 barrels to about 284.6 million, still the lowest since 1982 (EIA via Reuters, 23 September). Diesel is the pressure point: US diesel prices hit records on 16 September and Energy Secretary Chris Wright is now working with refiners on a voluntary cut in diesel exports rather than a formal ban (CNBC, 16 September; TradingEconomics, 24 September).

Why it matters. The market was tight on crude and tighter on products. It is now loosening on crude, with an unexpected build, while products stay stretched. That combination argues for a lower flat price and a wide diesel crack, which is a refiner's market rather than a producer's. Our oil long was a bet on the disruption premium; the premium is now being repaired at a known pace, and demand faces a 5 percent 10-year. The trim rule we set at $110 was 55 cents too high; we own that. The re-add at $100 worked.

Scenarios

Base case, 55%: Brent drifts to $95 to $100 within four weeks as the pipeline ramps and inventories build.
Downside for prices, 25%: a Hormuz deal, Brent $85 to $90.
Upside for prices, 20%: a new attack on Gulf infrastructure or a collapse of talks, Brent above $112.

Market implications

Asset class: Cut the core crude long to half; take profit on the $100 re-add above $105; refiners over producers while the diesel crack stays wide.
Currency and flows: A lower oil price relieves the rupee and the yen at the margin; the dollar's rates bid now matters more than the petro bid.
Sector rotation: Energy loses its hedge value as Brent falls; keep refining exposure, trim upstream.
Entry and exit: Re-add the core only on a break above $108 with a new supply hit; below $95, the half-size core is the floor position, not a sale.

What to watch. The pipeline flow rate, next week's EIA build or draw, the diesel export decision, and Chinese import data.

House view. Brent below $100 within four weeks. We hold a half-size core long as the tail hedge and stop adding. Medium conviction.

03 United States: A Hawkish Hike, a 5 Percent 10-Year and October in Play

What happened. The newest material development is the bond market, not the Fed meeting. The 10-year closed at 5.11 percent on 23 September, its first close above 5 percent since July 2007, after touching 5.135 percent; the 5-year crossed 5 percent for the first time since 2007; the 2-year reached 4.85 percent and the 30-year 5.40 percent (US Treasury, 23 September; CNBC, 23 September; TheStreet, 23 September). The move followed a September flash PMI that showed activity accelerating at the fastest rate since July 2021 with input costs surging, a 5-year auction that tailed 3.1bp, and hawkish remarks from Barr, Musalem, Goolsbee, Barkin and Collins; October hike odds moved to about 73 percent (S&P Global via CNN, 23 September; TradingEconomics, 23 September; CNBC, 23 September). Initial claims fell to 196,000 in the week to 12 September (Department of Labor, 17 September). The 30-year mortgage rate is 7.19 percent (CNBC, 23 September). Behind all of this sits the 16 September decision: a unanimous 25bp hike to 3.75 to 4.00 percent, a statement calling activity "solid", dots showing 16 of 18 participants expecting another 2026 hike with a 4.1 percent median for end-2026 and 2027, and a 30-minute press conference in which Warsh submitted no dot, offered no guidance, and pointed to hyperscalers' "competition for capital" as a driver of yields (Federal Reserve, 16 September; CNBC, 16 September; J.P. Morgan Asset Management, 16 September; Bloomberg, 16 September; Fortune, 16 September). Two catalysts changed shape. The 30 September shutdown risk is off the table, with a continuing resolution running to 11 December (NPR, 1 September; NBC News). And Xi Jinping is in Washington, with both sides agreeing on 23 September to extend the Busan tariff truce, due to expire 10 November, by two months (AP, 24 September).

Why it matters. Flip flagged on the scenario, not the call. Our base case was a hike with balanced guidance; we got the hawkish version. A Fed that won't call conditions restrictive at 4 percent, with a hot PMI and a 5 percent 5-year, is a Fed that hikes again unless October's data breaks. The 2-year at 4.85 percent already prices most of that; we still don't buy it, because the asymmetry now runs to 5 percent if the dots are honoured. The long end is a different question. At 5.11 percent the 10-year offers the highest carry in a generation while the front end does the tightening, oil is rolling over and the curve is bear-flattening. That is where we take the other side.

Scenarios

Base case, 60%: a second 25bp hike on 28 October to 4.00 to 4.25 percent, the 2-year 4.9 to 5.0 percent, the 10-year 5.0 to 5.3 percent.
Downside, 25%: a weak 2 October payroll and a soft September CPI pull October odds below 50 percent; the 2-year back to 4.6 percent, the whole curve rallies.
Upside for yields, 15%: hot payrolls and CPI push talk of 50bp; the 10-year through 5.35 percent and equities take a 5 percent hit.

Market implications

Asset class: Stay light on the 2-year until the October meeting is priced; add long-end duration in small size at a 5.10 to 5.20 percent 10-year, stop on a 5.35 percent close. This is the contrarian call of the week.
Currency and flows: A hiking Fed against a doubting ECB keeps the dollar 100 to 103 into October.
Sector rotation: Mega-cap AI still has the earnings; rate-sensitive small caps and housing don't. Hold the leaders, avoid duration in equity form.
Entry and exit: Buy the 2-year only at 5.0 percent or on an October statement that reads as the last hike; add the 10-year at 5.10 to 5.20 percent and the 30-year at 5.40 percent, in tranches.

What to watch. The 2 October payrolls, 6 October PCE, 14 October CPI, and Micron's 30 September results as the AI capex read.

House view. We expect a second 25bp hike on 28 October (Medium), keep the 2-year purchase on hold, and make a contrarian small add to the long end at a 5.10 to 5.20 percent 10-year, stop 5.35 (Medium).

04 Europe: Strong PMIs, Falling Hike Odds, a Euro at the Target

What happened. EUR/USD fell to 1.1376 on 24 September, through our 1.14 target, as the dollar surged (TradingEconomics, 24 September). The euro area's own news was firm: the September flash PMI showed the fastest private-sector growth in nearly three and a half years (S&P Global via TradingEconomics, 23 September). Yet market pricing for a 29 October ECB hike slipped to about 60 percent from roughly 73 percent a week ago as oil fell (ECB Watch, 24 September). The deposit rate stays at 2.50 percent after the 10 September move.

Why it matters. The euro is being sold on the rate gap, not on European weakness, and that gap widened this week from both ends: the Fed's October odds rose while the ECB's fell. A strong PMI keeps the October hike live, which caps how far the euro falls; a falling ECB probability keeps the short alive. We take profit on half and let the rest run.

Scenarios

Base case, 55%: the ECB hikes to 2.75 percent on 29 October and signals the end; EUR/USD 1.125 to 1.14.
Downside for the euro, 30%: the ECB holds while the Fed hikes, EUR/USD toward 1.11.
Upside for the euro, 15%: soft US data pulls the Fed's October odds down, EUR/USD back above 1.15.

Market implications

Asset class: Front-end euro rates are fully delivered; hold, don't add. European financials stay overweight while the curve is positive.
Currency and flows: Take profit on half the EUR/USD short at 1.14; run the rest with a target of 1.125 and a stop at 1.152.
Sector rotation: Financials over exporters; the weak euro helps exporters only if US demand holds through 5 percent yields.
Entry and exit: Re-sell EUR/USD on rallies toward 1.15; cover the remainder below 1.125.

What to watch. The 29 October decision, the early-October euro area flash inflation print, and Governing Council speakers on second-round effects.

House view. We expect a 25bp ECB hike to 2.75 percent on 29 October (Low) and keep half the EUR/USD short toward 1.125 (Medium).

05 United Kingdom: The BoE Holds 6-3, Stops Selling Long Gilts, and Cable Breaks 1.34

What happened. The Bank held Bank Rate at 3.75 percent on 17 September on a 6-3 vote, with Greene, Mann and Pill voting for 4.00 percent, and said that if the Middle East conflict persists, "as appears to be the case", and second-round risks grow, policy may have to tighten (Bank of England, 17 September). The same day it halted active sales of long-dated gilts, paused all gilt sales until April while it consults the DMO on selling directly to the Treasury, and cut planned sales to £20 billion a year within a £46 billion annual unwind; 30-year yields had their biggest one-day rally since April, to a three-week low (Reuters, 17 September). August CPI printed 3.1 percent, the first reading above 3 percent since March, and the Bank sees a peak above 4 percent in early 2027 (Bank of England, 17 September). By 23 September the rally had reversed at the long end: the 10-year benchmark yielded 5.33 percent and the 30-year about 5.82 percent, with all 71 conventional gilts selling off on the day (giltsyield.com, 23 September). Cable broke 1.34 on 16 September and trades near 1.333 (MTFX, 23 September). The 28 October Budget stays the next fiscal test.

Why it matters. Flip flagged on sterling. Our constructive GBP call assumed the pound would keep outperforming a firm dollar as it had for two weeks; a 101 dollar broke that. Nothing sterling-specific drove the move, but the 1.34 floor is gone and a three-dissent MPC that hasn't hiked is not the catalyst that brings it back. We move cable to neutral with a sell-rally bias. On gilts, the BoE's retreat from long-dated sales removes one structural seller and is the first support the long end has had in months; it is not enough to buy ahead of a Budget with halved headroom and a 5.1 percent Treasury curve.

Scenarios

Base case, 55%: the BoE hikes 25bp on 5 November after a run of 3 percent-plus CPI prints; cable 1.31 to 1.34; the 10-year gilt 5.2 to 5.5 percent.
Downside for sterling, 30%: the Budget disappoints on headroom, the 30-year retests 6 percent, cable through 1.30.
Upside for sterling, 15%: the Fed's October hike fades and the dollar rolls over, cable back above 1.35.

Market implications

Asset class: Stay off 10-year gilts into the Budget; start a small 30-year gilt long only above 6.0 percent, where the BoE's sales halt and the yield both argue for it.
Currency and flows: Neutral GBP; fade rallies toward 1.345, target 1.315, cover above 1.355.
Sector rotation: No fresh call.
Entry and exit: Sell cable at 1.345; buy the 30-year gilt at 6.0 percent, add at 6.2.

What to watch. The 28 October Budget, the 5 November MPC with a new Monetary Policy Report, and whether the DMO consultation ends BoE gilt auctions for good.

House view. Flip flagged. Cable moves from constructive above 1.34 to neutral with a sell-rally bias toward 1.315 (Low); we expect a 25bp BoE hike to 4.00 percent on 5 November (Low); gilts stay cautious until the Budget (Medium).

06 Australia: All Four Majors Now Call the Hike, and the Aussie Ignores Them

What happened. CBA brought its hike call forward from November and now expects the RBA to lift the cash rate 25bp to 4.60 percent on 29 September, joining Westpac, NAB and ANZ, which sees a second move to 4.85 percent in November (investingLive, 23 September). Market pricing for the 29 September move sits near 74 percent (RBA Rate Watch, 24 September). Governor Bullock told a parliamentary committee that some of the upside risks to inflation appear to be materialising (investingLive, 23 September). The latest monthly CPI showed 3.5 percent headline and 3.6 percent trimmed mean for July, and the August print lands 30 September, the day after the decision (OrbitRemit, September 2026). AUD/USD fell to about 0.709 with a 0.7075 low despite the hawkish shift (Investing.com, 24 September).

Why it matters. The rates leg of our Australia view is about to be graded and looks right; the currency leg has been wrong because the dollar overwhelmed it. A hike that every major bank and three-quarters of the market expects is mostly priced, so the Aussie's bounce depends on the statement pointing to November, which ANZ already does. The dip toward 0.70 we said we would buy has arrived. We buy it, small, ahead of the meeting.

Market implications

Asset class: Australian rates biased restrictive; a second hike in November is the more probable path if August CPI stays above 3.5 percent.
Currency and flows: Add a small AUD/USD long at 0.700 to 0.705, stop 0.695, target 0.72 within two weeks of the hike.
Sector rotation: No fresh call.
Entry and exit: Buy 0.700 to 0.705; trim at 0.72; cut below 0.695.

What to watch. The 29 September decision and statement at 2:30pm AEST, the 30 September August CPI, and the front end of the AUD curve.

House view. The RBA hikes to 4.60 percent on 29 September (High). We add a small AUD/USD long at 0.70 to 0.705 and expect 0.72 within two weeks of the move (Medium).

07 India: The Rupee Holds the Line, So We Resume Adds

What happened. USD/INR closed 95.74 on 23 September and opened near 95.85 on 24 September, inside our 95.5 to 96.5 band, with the RBI intervening to cap losses and a deputy governor flagging the case for rupee gains as oil eased (Reuters via Business Recorder, 24 September; TradingEconomics, 24 September). The Nifty closed 23,446.80 on 23 September, up 0.5 percent on the day, with metals up 2.4 percent and IT down 1 percent (Business Standard, 23 September). The RBI's next decision is 7 October; the August meeting held at 5.25 percent with a neutral stance and expected inflation to peak in the October to December quarter (5paisa MPC schedule; Forbes India, 5 August, background).

Why it matters. Flip flagged from hold to modest adds. Last week the importer was taking the oil shock through the index and the currency at once. This week the shock reversed: Brent is $5 lower, the rupee didn't break, and the central bank is talking the currency up. The 96.5 trim line was never tested. With the Nifty sitting on the floor of our range and the largest external headwind fading, we go back to staggered adds, in small size, in the two sectors we already own.

Scenarios

Base case, 55%: Brent stays below $105, USD/INR drifts to 95.0 to 95.5 within a month, the Nifty climbs toward 24,000.
Downside, 30%: a US 10-year above 5.3 percent drives EM outflows, USD/INR through 96.5, Nifty below 23,000.
Upside, 15%: a Hormuz deal takes Brent under $90 and the Nifty reclaims 24,200 with the rupee near 94.5.

Market implications

Asset class: Resume staggered adds to large-cap IT and financials below 23,500; stop the adds if USD/INR breaks 96.5.
Currency and flows: USD/INR to 95.0 to 95.5 within a month; RBI intervention plus lower oil make a break of 96.5 the less likely path.
Sector rotation: Financials benefit most from a steadier rupee; IT gives back some of its weak-rupee tailwind.
Entry and exit: Add on Nifty closes below 23,500; trim the tilt above 24,200 or on a USD/INR break of 96.5.

What to watch. The 7 October RBI decision, USD/INR against 96.5, and the 2 October US payrolls as the EM-flow signal.

House view. Flip flagged. Hold the structural long and resume small staggered adds; the RBI holds at 5.25 percent on 7 October (Medium); USD/INR to 95.0 to 95.5 within a month (Medium).

08 Precious Metals and Safe-Haven Assets

What happened. Gold did what we said it would, then stopped. It rose from about $4,263 to $4,364 on 18 September and a $4,369 close on 21 September once the hike was out of the way, then slipped to about $4,304 on 23 September as the dollar hit 101 and Musalem and Goolsbee both signalled more tightening (CNBC, 18 September; Rio Times, 22 September; Kitco, 23 September; USAGOLD, 23 September). Silver ran harder both ways, up to about $65 and then down 3 percent on 23 September; the gold/silver ratio is near 66 (USAGOLD, 23 September). No new central-bank purchase data landed this week.

Why it matters. The contrarian add is about one percent ahead, a win on the letter and a draw on the spirit. What matters more is what didn't happen: gold held $4,300 through a 5.11 percent 10-year and a two-month high in the dollar, the two forces that broke it a fortnight ago. That is a market being absorbed by the official-sector bid at the level we chose. With a second Fed hike now the base case, the near-term upside is capped and the next add belongs lower.

Market implications

Asset class: Hold the core gold long and the $4,300 add; the next add is at $4,200, not before. Silver stays smaller.
Currency and flows: A 101 dollar and rising real yields cap gold into 28 October; a Fed pause after October is the release valve.
Sector rotation: No fresh call.
Entry and exit: Add gold at $4,200; trim above $4,700 unchanged; add silver below $62.

What to watch. The 28 October FOMC pricing, the dollar against 102, and any monthly central-bank purchase releases.

House view. Gold holds $4,200 through an October hike and trades $4,200 to $4,450 into the meeting; we hold and don't add above $4,300. Medium conviction.

09 Sovereign and Credit Conditions

What happened. HY OAS held at 2.66 to 2.68 percent while the 10-year closed above 5 percent and the 5-year crossed 5 percent for the first time since 2007 (FRED/ICE BofA, 21 September; US Treasury, 23 September; TheStreet, 23 September). The IIF's Global Debt Monitor reported that global debt rose by more than $10 trillion in the first half of 2026 to a record above $365 trillion, with emerging-market debt up $6.5 trillion to more than $110 trillion, led by China (Reuters, 23 September). Warsh named hyperscalers' competition for capital as a factor pushing yields higher (Fortune, 16 September). Japan's 10-year JGB yield fell after the BoJ hike, the mirror image of the Treasury move (Babypips, 18 September). The BoE stepped back from long-dated gilt sales (Reuters, 17 September).

Why it matters. Spreads haven't widened, and that is now the risk rather than the reassurance. Every basis point of this week's move landed on the risk-free leg, so all-in high-yield borrowing costs rose without any compensation for the credit itself. A 2.66 percent spread over a 5.11 percent Treasury is a market pricing no recession, no refinancing wall and no AI capex disappointment, while the Fed is telling it to expect another hike. The supply side is the other half: record global debt and hyperscaler issuance are what a term-premium spike looks like from the inside. We stay up in quality and, against the tape, expect spreads to widen.

Market implications

Asset class: Overweight IG, underweight HY, unchanged; the HY scale-in stays at 3.25 to 3.5 percent OAS.
Currency and flows: No fresh call.
Sector rotation: Short-spread-duration IG and energy issuers with hedged production; avoid single-B AI infrastructure paper.
Entry and exit: Begin the HY scale-in only past 3.25 percent; the first widening to 3.0 percent is a signal, not an entry.

What to watch. HY OAS against 3.0 percent, the 5-year and 7-year auction tails, and the November refunding announcement.

House view. Overweight IG, underweight HY, and, against the tape, HY OAS to 3.0 percent by the 28 October FOMC. Medium on the positioning, Low on the timing.

10 Other Flashpoints

Japan and the yen. Flip flagged. The BoJ hiked to 1.25 percent on 18 September on a 7-2 vote, with Asada and Sato dissenting, and the yen fell on the dovish framing; USD/JPY sits near 158.3 and the pair is on intervention alert after a reported rate check (CNBC, 18 September; Babypips, 18 September; TradingEconomics, 23 September). We close the trailing short. A fresh short only on a confirmed MoF intervention or above 160, with a cut above 162. The summary of opinions is due 1 October and the next meeting 29-30 October. Medium conviction on the exit.
US-China. Xi Jinping is in Washington. The two sides agreed on 23 September to extend the Busan truce, due to expire 10 November, by two months, while Trump's new 12.5 percent forced-labour tariffs on 60 partners, including China, remain and Bessent said the Busan deliverables are not yet fully met (AP, 24 September; Al Jazeera, 24 September). We expect no chips breakthrough and a modest risk-positive outcome; the truce now runs into January.
AI concentration. The Nasdaq closed at a record 27,244 on 22 September and chips rose 2 percent that day; Micron reports 30 September as the next capex read (Vantage, 23 September). Warsh's "competition for capital" line is the first time a Fed chair has tied hyperscaler funding to the level of yields (Fortune, 16 September); we don't fight the leaders, and we do watch their bond supply.
US fiscal. The 30 September shutdown risk is gone; the continuing resolution runs to 11 December, which becomes the next fiscal date (NPR, 1 September; NBC News).
Ukraine/Russia. No material new development we can verify beyond Zelenskyy's UNGA-week call for a trilateral meeting with Trump and Putin (TradingEconomics, 23 September).
Red Sea security. No verifiable new development around Bab al-Mandeb this week; the Houthi hold on Mocha and Perim stays a second chokepoint risk.
US midterms. 3 November, no new development; the extended China truce and a shutdown-free autumn remove two pre-election risks.

Upcoming Events and Catalyst Calendar

24 September, Trump-Xi White House summit and US weekly claims. The truce extension is agreed; chips and soybeans are the open items.
29 September, RBA decision, 2:30pm AEST. Our High-conviction hike call and the new AUD long are graded here.
30 September, US Q2 GDP third estimate (BEA), Australian August CPI, Micron results, and the US fiscal year end with a continuing resolution in place.
1 October, BoJ summary of opinions. The read on how dovish the 1.25 percent hike really was.
2 October, US September payrolls. First test of the October hike pricing.
6 October, US August PCE (BEA). The Fed's own inflation gauge, two weeks after the hot PMI.
7 October, RBI decision. We expect a hold at 5.25 percent.
14 October, US September CPI. The last big print before the October FOMC.
27-28 October, FOMC, no projections. Our second-hike call is graded.
28 October, UK autumn Budget. The fiscal test that keeps us off 10-year gilts.
29 October, ECB decision, US Q3 GDP advance estimate and September PCE (BEA). Our Low-conviction ECB hike call is graded.
29-30 October, BoJ meeting with Outlook Report.
3 November, US midterms.
5 November, BoE decision with Monetary Policy Report. Our Low-conviction hike call is graded.
8-9 December, FOMC with projections. The year-end dot plot.
11 December, US continuing resolution expires.
17 December, ECB and BoE decisions. The year-end policy checkpoint.
Around 10 January 2027, extended Busan truce expiry.

Recommendations (Staged)

Immediate, oil: Cut the core Brent long to half size at $103; take profit on the $100 re-add above $105; re-add the core only above $108 on a new supply hit. Horizon: four weeks, target below $100.
Immediate, rates: Keep the 2-year purchase on hold until the October meeting is priced or the 2-year reaches 5.0 percent. Add long-end duration in small size, contrarian: the 10-year at 5.10 to 5.20 percent, the 30-year at 5.40 percent, stop on a 5.35 percent 10-year close. Trigger: the 2 October payrolls.
Immediate, Japan: Close the trailing short USD/JPY at market; re-short only on confirmed MoF intervention or above 160, cut above 162.
Immediate, Europe: Take profit on half the EUR/USD short at 1.14; run the rest toward 1.125 with a 1.152 stop.
Immediate, Australia: Add a small AUD/USD long at 0.700 to 0.705, stop 0.695, target 0.72. Trigger: the 29 September RBA hike.
Near-term, UK: Cable to neutral; fade rallies toward 1.345 for 1.315, cover above 1.355. Stay off 10-year gilts into 28 October; buy the 30-year gilt only above 6.0 percent.
Near-term, India: Resume small staggered adds to large-cap IT and financials below Nifty 23,500; stop the adds if USD/INR breaks 96.5; trim above 24,200.
Near-term, gold: Hold the core and the $4,300 add; next add at $4,200; trim above $4,700; add silver below $62.
Structural, credit: Overweight IG, underweight HY; scale into HY only past 3.25 to 3.5 percent OAS; treat a move to 3.0 percent as confirmation of the widening call, not an entry.
Structural, equities: Hold the mega-cap AI leaders; buy a 3 percent S&P 500 pullback toward 7,450 before the October FOMC; avoid rate-sensitive small caps and housing until the 10-year peaks.

Thresholds That Change the Calls

Brent above $108 on a new supply hit: rebuild the oil core long to full size.
Brent below $95: hold the half-size core as the floor; no further sales.
Payrolls on 2 October below 100k with rising unemployment: October hike odds fall; buy the 2-year at 4.75 percent and add to the long end.
10-year closes above 5.35 percent: cut the contrarian long-end add.
2-year at 5.0 percent: buy it regardless of guidance.
DXY above 102.5: cut the AUD long and lower the gold add to $4,150.
Confirmed MoF yen intervention or USD/JPY above 160: re-short USD/JPY; above 162, cut it.
EUR/USD above 1.152: cover the remaining euro short; below 1.125, cover it in profit.
GBP/USD above 1.355: cover the cable short bias; 30-year gilt above 6.0 percent, start the gilt long.
RBA holds on 29 September: cut the AUD long at 0.695 and move the Australian rates view to neutral.
USD/INR breaks 96.5: stop the India adds and trim the tilt.
HY OAS past 3.0 percent: the widening call is confirmed; past 3.25 to 3.5 percent, begin the HY scale-in.
Gold below $4,200: add to the core long; above $4,700, trim.

Markers of Concern

Levels are a dated snapshot, the 23 September close or 24 September intraday.
Scenarios are probabilities, not forecasts.
Several levels this week come from market data aggregators rather than exchange settlements; Brent quotes on 24 September ranged about $3 between futures and CFD feeds.
The contrarian long-end add is against a market that has sold the 10-year to a 19-year high with a second hike priced; a hot October CPI takes it out at the 5.35 percent stop.
Cutting the crude long to half exposes us to a fresh attack on Saudi or Gulf infrastructure, which would send Brent back through $112 in days.
Our yen mark last week looks to have been wrong by several yen; the closed short carries a larger loss than the 154 print implied.
The RBA call is High conviction into a decision that three-quarters of the market and every major bank expect; a hold would hit the AUD long and the rates view at once.
The Trump-Xi summit outcome was not available at the time of writing; the calendar entries for the truce assume the 23 September extension holds.
Credit spreads near 2.66 percent still offer no cushion; the widening call is against the tape and may take longer than the October horizon we set.