Issue No. XXI  · 

Friday's rally was bought on a misreading

Equities treated July's 23,000-job contraction as insurance against a September hike. The softer wage data points to weakening demand, and with inflation at 3.5% the Fed cannot answer it quickly.

12 min read·Free to read·All issues
Vol. I  ·  No. 21 — Established MMXXVI — Price: by subscription
The Research Brief

A Weekly Survey of Markets, Policy, & Power

Issue No. 21 · Monday, the 10th of August, MMXXVI · London

This Week, in Brief
Four notes for the hurried reader
  1. The DAX gained 2.69% and closed at a record 26,319, its best week since June, while the FTSE 100 managed 0.30%. That gap comes down to what sits inside each index.
  2. The Bank of England held at 3.75% on a 6–3 vote, with three members wanting 4.0%. Its own projection has CPI peaking near 3.2% in the fourth quarter, which means the live argument on Threadneedle Street is about hikes.
  3. The US economy shed 23,000 jobs in July and prior months were revised down by 103,000 — and the S&P 500 closed at a record anyway, on the theory that a broken labour market restrains a hawkish Fed.
  4. Rheinmetall's order backlog swelled to roughly €80bn, and the shares fell about 15%. Berlin cancelled a frigate programme, and the market re-learned that a defence pledge is not a signed contract.
§ I. from the editor

The Editor's View

A note from the desk

Friday's Rally Was Bought on a Misreading

The Research Brief takes the view that Friday's equity rally was bought on a misreading. The market treated a 23,000-job contraction as insurance against a September hike, when the more informative number in the release was average hourly earnings at 3.2% year-on-year — the softest since May 2021 — alongside participation falling to 61.4%. Wage growth that weak, with people leaving the workforce rather than being absorbed by it, is a demand signal, and demand signals do not stay confined to the labour data. Headline inflation at 3.5% means the Fed cannot respond quickly even if it wanted to. Equities are therefore priced for a central bank that is both dovish and unconstrained, and it is neither. The strongest counterargument is the earnings line: FactSet has S&P 500 second-quarter profit growth at 47.4%, and companies growing like that can absorb a soft quarter of consumption.

§ II. approx. four minutes

Macro & Policy Backdrop

Macro & Policy Backdrop

The UK is in better shape than the commentary suggests, but a Bank forecasting rising inflation does not cut.

The UK is in better shape than the commentary suggests. First-quarter GDP came in at 0.6%, services leading at 0.8%, and July's services PMI jumped to 51.7 from 49.4 — a five-month high and a return to expansion in the sector that carries the economy. The IMF is reported to have nudged its 2026 growth forecast up to 1.0% from 0.8%. Set against that, headline CPI at 2.6% in June — with core easing to around the same level — looks tolerable until you read the Bank's own August projection, which has inflation climbing to roughly 3.2% by the fourth quarter. A central bank forecasting a rise in inflation from here does not cut. The 6–3 hold on 30 July, with three members pushing for 4.0%, tells you the committee knows it.

The ECB is in a stranger position. It raised by 25bp in June because of Middle East energy pass-through, then held the deposit rate at 2.25% on 23 July, and does not meet again until 10 September. That is a five-week gap during which the Strait of Hormuz will either move toward reopening or back toward escalation, with no scheduled opportunity to respond. Frankfurt has effectively pre-committed to sitting still through the most consequential variable in the European inflation outlook.

In the US, the Fed held at 3.50%–3.75% on a 9–3 vote, three dissenters wanting a rise. June's disinflation was real — headline down to 3.5% from 4.2%, core to 2.6% from 2.9% — but it came almost entirely from energy prices falling after the April ceasefire, which makes it a borrowed improvement rather than an earned one. Wednesday's July CPI is the pivot. Consensus is 0.1% month-on-month after June's 0.4% decline; anything at 0.2% or above says the energy relief was a one-off and puts September back in play. Meanwhile the 10-year Treasury sits at 4.60%, well above the funds rate, because the OBBBA's projected $2.8–3.8trn addition to the deficit over the coming decade has to be funded by somebody.

"A central bank forecasting a rise in inflation from here does not cut."

§ III. to Friday's close

UK & European Markets

UK & European Markets

The STOXX 600 rose 1.70% in local currency to a fourth consecutive record close. Beneath that, the dispersion was wide: Italy's FTSE MIB up 2.96%, the DAX up 2.69% to 26,319, the CAC 40 up 2.41% to 8,714.93, and the FTSE 100 up 0.30% to roughly 10,901. London's problem was Monday. When Trump called off the planned Iran strike, WTI fell about 5% to $80.34 and Brent to $83.77, and an index weighted as heavily toward BP and Shell as the FTSE 100 cannot rally on cheaper oil the way Frankfurt can. Friday's DAX gains were led by SAP at 4.30%, with Scout24 and Infineon behind it.

This is now the defining feature of European equity performance, and it cuts both ways. The FTSE 100's energy weighting is a drag while diplomacy is progressing and a hedge if it collapses — which is why the index has become, in practice, a geared bet on the Strait of Hormuz staying shut. The German index has the opposite exposure: Siemens raised its outlook on 6 August after a record third quarter, SAP is compounding cloud backlog at 26%, and Rheinmetall's order book keeps growing regardless of the oil price. Sterling has held firm on the Bank's hawkish posture, which helps FTSE 100 constituents earning at home and hurts the dollar earners that dominate the index by weight. The week's losers were instructive too: Zalando fell 16% on a growth downgrade blamed on weak sneaker demand, which is a small but real data point about the European consumer.

DAXDAX26,319+2.69%
CAC 40CAC8,714.93+2.41%
FTSE MIBMIB—+2.96%
FTSE 100UKX10,901+0.30%
STOXX 600SXXP—+1.70%

§ IV. to Friday's close

US Markets

US Markets

The headline weekly moves were tiny: SPY closed at 773.26, up 0.29%, QQQ at 723.03 up 0.40%, DIA up 0.17%, IWM up 0.37%. SPY's entire weekly range was just over half a percent. That flatness conceals a violent middle: Monday delivered a 1.48% S&P gain to 7,600.50 and a 2.1% Nasdaq surge on the Iran de-escalation, a Dow record, and then Friday's record close on a payrolls print that was, by any conventional reading, awful.

The internals are consistent with the rate story rather than the growth story. Nasdaq over Dow, and small caps outpacing the Dow as well, is what you get when the market decides the hiking risk has been removed — the Russell's constituents carry floating-rate debt and are the most direct beneficiaries of a rate ceiling. What you do not see is cyclical leadership, which is what a rally built on confidence in growth would produce. AppLovin's 19% single-day collapse on a 1.2% revenue miss, against 53% growth, is the honest read on how much perfection is embedded in the high-multiple end of this market.

SPDR S&P 500 ETFSPY773.26+0.29%
Invesco QQQ TrustQQQ723.03+0.40%
SPDR Dow Jones ETFDIA—+0.17%
iShares Russell 2000IWM—+0.37%

§ V. a longer view

Geopolitics & Policy Implications

An ink silhouette will be set here weekly: a skyline, a row of figures, an instrument of state.

Geopolitics & Policy Implications

Hormuz has been effectively closed since the February strikes on Iran, and the IEA calls it the largest supply disruption in history. It is a defensible description: the strait carried around 25% of seaborne crude and petroleum products and 19% of LNG in 2025. This week produced the most substantive movement yet. Iran and Oman agreed the geographical coordinates of a proposed safe shipping corridor, and Tehran immediately clarified that a completed framework would not mean reopening, and that US concessions come first. That pattern already has form — the two sides signed an MOU in June, and Iranian forces attacked vessels deemed noncompliant in July.

Lufthansa's results are the cleanest measurement of how this transmits into corporate earnings. Record revenue of €11.1bn, and adjusted EBIT down €490m to €383m, because fuel cost an extra €750m in a single quarter and the full-year bill is now guided at €8.7bn. The stock fell around 10%. The second-order point is that IAG, Ryanair, easyJet and Air France-KLM are buying the same kerosene into the second half, and Lufthansa's guidance range of €1.7–2.2bn — half a billion euros wide — tells you management has no confidence in forecasting its largest variable cost. Carsten Spohr added the detail that matters more for 2027: Gulf carriers have returned at near-full capacity and are redeploying it into transfer traffic, so European long-haul is absorbing a fuel shock and a competitive pricing shock at the same time.

The defence story fractured in an interesting way. Rheinmetall grew first-half sales 39% to €5.2bn, lifted second-quarter operating profit 115% to €562m, and expanded its backlog to roughly €80bn, most of it in fixed orders. It then cut full-year sales guidance by about €300m to €13.7–14.2bn because Germany cancelled the F126 frigate programme, and the shares dropped 15%. Nothing about European rearmament demand changed this week; what changed was the market's assumption that political commitment converts mechanically into revenue. BAE Systems and Leonardo are untouched by F126 specifically, and their broader spread of government programmes across multiple sovereign customers becomes more valuable if Berlin's procurement discipline is a preview rather than an exception. The other live geopolitical thread is the Houthi threat to Bab al-Mandeb, which is the route Saudi Arabia has been using to work around Hormuz — close the workaround and the diplomatic timeline stops mattering.


§ VI. three names worth following

Signals & Implications

Signals & Implications

The energy trade is asymmetric and the market is only halfway there.

Brent at $83–84 sits well below April's WTI peak near $115 but comfortably above the post-ceasefire lows, which implies the market is partially discounting a reopening. If one arrives, the earnings BP, Shell and TotalEnergies just reported on elevated realisations erode fast, and the FTSE 100's composition turns from headwind to something worse. The mirror image is that BASF and LyondellBasell get input relief they have not had since February, and the airlines get a second-half guidance upgrade rather than the downgrade currently embedded. Anyone sitting long energy producers and short energy-intensive industrials is holding the same directional bet twice.

AI is being marked to earnings for the first time.

Airbnb's AI assistant now resolves nearly 45% of support issues without a human and cut support cost per booking by roughly 16% year-on-year — a measurable margin line, which is why full-year EBITDA margin guidance went to at least 35.5%. SAP put AI into more than 90% of its fifty largest deals and grew cloud backlog 26%. AppLovin, whose model is the product, missed by 1.2% because a model upgrade landed after quarter-end and lost a fifth of its value. Where AI is a cost lever the market pays; where AI is the revenue engine, the timing risk is now priced brutally. Booking Holdings and Expedia should show the same support-cost effect; Trade Desk carries the same model-cycle fragility.

The deficit is doing the tightening the Fed isn't.

The OBBBA adds roughly 0.9 percentage points to US GDP this year and $2.8–3.8trn to the deficit over a decade, and the 10-year at 4.60% is where that arithmetic is being settled. A term premium that stays elevated regardless of what the funds rate does hits anything financed with long-dated debt: UK housebuilders such as Persimmon and Taylor Wimpey, whose land banks are carried against a cost of capital that has not normalised, and regulated utilities like National Grid and United Utilities running large capital programmes on borrowed money.

China is softening quietly.

The Caixin manufacturing PMI eased to 50.9 in July from 51.7, a four-month low, with new orders and production both growing more slowly. Still expansion, but the direction matters for LVMH, Richemont and Hermès, whose Chinese demand recovery is the load-bearing assumption in every European luxury forecast, and for Glencore, Rio Tinto and Anglo American on the industrial side. Export orders returned to growth for the first time in three months, which complicates the picture rather than resolving it.


§ VII. briefly

Earnings Notes

Airbnb

Revenue up 17% to $3.61bn, EPS $1.37 against $1.25 expected, shares up 9% after hours. Nights and seats booked grew 10%, accelerating from the first quarter, and first-time booker growth of 11% was the best in four years. The CFO's line that conflict impacts were "less than anticipated" is the read-across: discretionary travel demand has not broken.

Rheinmetall

Covered above, but the number that will still be true in three years is the backlog, reported at roughly €80bn and heavily weighted toward fixed orders. The guidance cut is a one-programme event; the multiple compression it triggered is a repricing of political risk across the whole sector.

AppLovin

$1.92bn revenue against $1.94–1.95bn expected, adjusted EBITDA margins of 84%, and a 19–21% drawdown. A business earning those margins is not in trouble. A business priced for never missing is a different proposition.

Bayer

Up more than 4% on a surprise 1.9% rise in quarterly operating profit. Modest in isolation, and a reminder that the bar for German industrial recovery stories is currently on the floor.


§ VIII. four threads, gathering

Narratives to Watch

  1. 1. Hormuz. Coordinates agreed, reopening explicitly withheld pending US concessions. Confirmation would be a publicly communicated sanctions waiver or an Omani-brokered framework with an actual timeline.
  2. 2. Warsh's framework review. Five task forces, one of which is weighing whether to scrap the dot plot — the primary instrument through which global fixed income prices forward expectations. Any statement from Warsh or FOMC members on the review's timeline will be the next marker of the new regime's reaction function.
  3. 3. The Burnham cabinet's first spending statement. Reeves's departure has not been priced into UK domestic equities. Balfour Beatty, Centrica, QinetiQ and Chemring all sit downstream of decisions nobody has announced yet.
  4. 4. Earnings breadth. That 47.4% aggregate growth rate leans heavily on technology. NVIDIA on 26 August is where the gap between announced hyperscaler capex and actual chip orders gets measured.

§ IX. to be watched

The Week Ahead

Wednesday's July CPI is the only scheduled release that can reprice the whole market. But any unscheduled Hormuz announcement will matter more than all of it.

13 AugUS July CPI (13:30 UK time) — consensus 0.1% month-on-month after June's 0.4% fall; a 0.2% print revives the September hike conversation and unwinds Friday's rally
13 AugCisco reports — enterprise IT spending intent and a real-time reading on whether the AI buildout is still accelerating ahead of NVIDIA
14 AugApplied Materials reports — semiconductor capital equipment orders, as close as we get to a real-time reading on AI infrastructure momentum
14 AugONS first estimate of UK second-quarter GDP — after Q1's 0.6% and a services PMI back in expansion, a soft print would sit oddly with the rest of the domestic data
✦   ❦   ✦
Notice to the reader This newsletter is published for informational and educational purposes only. It does not constitute investment advice, a solicitation, or a recommendation to buy or sell any financial instrument. Past performance is not indicative of future results. Financial markets involve risk. Always conduct your own due diligence and consult a qualified financial adviser before making any investment decisions. This publication is not authorised or regulated by the Financial Conduct Authority (FCA).
· ✦ ·

The Research Brief · theresearchbrief.co.uk · London · MMXXVI

Have the next one arrive on Sunday.

One issue a week: macro and policy, markets, geopolitics, and the week ahead. No advertising.