Market Daily: Weak US hiring sharpens the rate debate as Nike and Wetherspoon confront tougher conditions

The day ahead

Friday’s US employment report provides the central evidence for this weekend’s briefing. Non-farm payroll employment increased by 29,000 in September, unemployment was 4.2%, and the estimated gains for July and August were revised down by a combined 60,000. This shows slower hiring, but it does not determine the Federal Reserve’s next interest-rate decision.

Company results add a mixed picture. Nike reported lower quarterly revenue and expects a high-single-digit revenue decline for its full financial year. Wetherspoon’s annual profit fell as costs increased, although its recent like-for-like sales growth was stronger. Investors therefore need to distinguish completed results from management forecasts and restructuring targets.

Reuters reported gains in global shares and support for government bonds during Friday’s session, alongside lower oil prices. These are dated observations, not proof that the employment report caused every move. Crypto developments included new Chainalysis findings about the Bitget theft, an Arbitrum restriction on new Stylus activations and NEAR Intents’ reported recovery of exploited funds.

News selection closed at 06:10 British Summer Time, or 05:10 UTC, on Saturday 3 October 2026. Traditional markets were closed. The fixed cryptocurrency observations were recorded separately at approximately 05:07 UTC while crypto markets were trading continuously.

UK Market Update

Wetherspoon separates completed cost pressure from recent sales

Daily Business reported that J D Wetherspoon’s sales for the year to 26 July increased by 5.2% to £2.24 billion. Profit before tax and separately disclosed items nevertheless fell by 28% to £58.6 million from £81.4 million.

The report identified £46 million added to the wage bill, £31 million of additional repair expenditure and £9 million of business rates. These costs are not interchangeable. Wages and rates affect continuing operations, while the timing of repairs may vary between reporting periods.

The latest trading period was more encouraging. Like-for-like sales, which compare pubs open in both periods, increased by 8.6% in the nine weeks to 27 September. Chairman Tim Martin said exceptional weather had helped, so readers should not assume that the recent rate will persist. Management continued to expect current-year profit before tax and separately disclosed items to be in line with market expectations. That remains a forecast rather than profit already earned.

The practical question is whether sales growth can continue to absorb wage, repair and business-rate pressures without weakening margins or demand. A nine-week sales comparison should not be treated as directly equivalent to profit from a completed financial year.

A dated FTSE observation

The FTSE 100 was last reported by the captured provider at 10,461.95 index points at 15:35:29 UTC on Friday 2 October. The provider described this as a last-reported quote and did not specify the market state, so it is not presented as an exchange-certified closing level. The UK cash market was closed at Saturday’s research cut-off.

US Market Update

Hiring slows without dictating the next rate decision

The US Bureau of Labor Statistics reported that non-farm payroll employment changed little in September, increasing by 29,000. The unemployment rate also changed little at 4.2% and remained within the 4.1% to 4.3% range recorded since March.

Revisions reinforced the evidence of slower hiring. July was revised from a gain of 21,000 to a loss of 10,000, while August was revised from 162,000 to 133,000. The two months together contained 60,000 fewer jobs than previously estimated. Average hourly earnings increased by 0.1% during September and by 3.0% over 12 months.

The rounded unemployment rate conceals a smaller change. St Louis Federal Reserve analysis put the increase at about 0.03 percentage points, from 4.141% to 4.175%. It also reported small improvements in labour-force participation and the employment-to-population ratio.

Slower payroll growth and downward revisions reduce the evidence of an overheating labour market. They do not remove inflation risk or commit policymakers to a particular decision. Market-implied probabilities can change quickly, but they are not promises from the Federal Reserve.

Nike’s Pace programme combines potential savings with substantial charges

Nike reported first-quarter revenue of US$11.2 billion, down 4% on a reported basis and 5% after removing currency movements. Nike Direct revenue fell 8%, Nike Brand Digital declined 13% and Converse revenue fell 28%. Greater China revenue declined 22% as reported, while North American revenue rose 2%.

Gross margin increased by 0.6 percentage points to 42.8%, primarily because of lower warehousing and logistics costs. Net income fell by 2% to US$712 million, while inventories were 3% lower at US$7.8 billion. Better cost control therefore accompanied continuing weakness in important products, channels and regions.

Nike expects fiscal-2027 revenue to decline by a high-single-digit percentage. Its new Pace programme includes supply-chain changes, a new campus in India, a realignment into three geographies and further organisational streamlining. The company estimates that the programme could deliver approximately US$2.5 billion of cumulative savings through fiscal 2031.

The accounting distinction matters. Nike also expects approximately US$1 billion of pre-tax charges through fiscal 2031. Its Form 8-K filing says the savings estimate is before those charges and future reinvestment, and warns that actual savings, charges and cash expenditure may differ materially. The US$2.5 billion is therefore an estimate, not cash already received or profit already earned.

Reuters, in a report carried by Euronext, described intensifying competition in China and questions about Nike’s product pipeline. The next useful evidence will be regional sales, digital performance, inventory and margins. Cost reductions can support profit, but the turnaround also requires products that sustain customer demand.

US shares ended Friday higher

Friday’s last-reported provider observations put the S&P 500 at 7,722.72 index points at 20:59:13 UTC, the Dow Jones Industrial Average at 51,176.96 at 21:00:17 UTC and the Nasdaq Composite at 27,190.863 at 21:15:59 UTC. The provider did not specify market state, so these are dated last-reported levels rather than exchange-certified closes.

Global Markets

Global shares recovered, but regional results remained uneven

Reuters reported during Friday’s session that global shares rose, government bonds found support and oil prices fell. The pan-European STOXX 600 was up 0.8% at the report’s observation, although it remained on course for a weekly decline.

Asia had already produced a mixed session. MSCI’s broad index of Asia-Pacific shares outside Japan fell 0.1% on Friday and 1.3% over the week. Japan’s Nikkei fell 0.9% on the day but gained almost 3% over the week, while Hong Kong’s Hang Seng fell 2.7% after a holiday. Mainland Chinese markets were closed for a public holiday.

The broader lesson is that the phrase “global markets rose” conceals meaningful regional differences. A broad index can provide context, but it should not replace checking the particular country, currency and asset held.

Bonds, Currencies and Commodities

Treasury yields eased during Friday trading

Reuters reported the benchmark US 10-year Treasury yield down by six basis points, or 0.06 percentage points, at 5.1717% during Friday trading, after it had reached 5.3445% on Thursday. Bond yields move inversely to prices.

The captured provider later reported a 5.277% yield at 18:59:55 UTC, with market state unspecified. These observations came from different times and should not be treated as contradictory closing figures. Inflation, government borrowing, energy prices and future central-bank decisions remain relevant to longer-term yields.

The dollar and bilateral exchange rates measure different things

The captured provider last reported sterling at US$1.324 per pound and the euro at US$1.1257, both at 21:29:05 UTC on Friday, with market state unspecified. Reuters separately observed the dollar index, which measures the US currency against a six-currency basket, down 0.1% at 101.86 during Friday trading. An index level and a bilateral exchange rate are different measurements.

Brent reverses part of Thursday’s rise

Reuters reported Brent crude futures down 2.7% at US$99.45 per barrel during Friday trading. This was an intraday futures observation, not a spot price or final settlement. It followed Thursday’s US$102.31 settlement for the December contract.

The comparison shows why the delivery contract and observation time belong beside an oil figure. A fall during one session does not establish the future cost paid by consumers or companies.

Gold rebounds without reversing its weekly decline

Reuters reported spot gold up 1.1% at US$4,223.49 per troy ounce at 13:05 UTC on Friday while trading was open. It remained about 1.5% lower over the week at that time. US gold futures were separately quoted at US$4,254.10.

The captured provider later reported COMEX gold futures at US$4,162.30 per troy ounce at 20:59:58 UTC, with market state unspecified. Spot gold and futures are different instruments, observed at different times. Friday’s rebound therefore did not erase the weekly decline or establish one definitive price for the entire session.

Crypto

Chainalysis publishes a new account of the Bitget theft

Chainalysis said it attributed the US$387 million stolen from Bitget on 24 September to North Korea-linked actors. This is the analytics firm’s assessment, not a newly announced government or court finding.

The firm said the assets left Bitget in 23 transfers during the first three hours, initially reaching Ethereum, XRP, Zcash and Tron. Investigators then traced movements through cross-chain protocols, instant swaps and other services. Cross-chain protocols allow value to move between blockchain networks, complicating investigation even though the individual transactions remain recorded.

Chainalysis said its investigators used internally developed artificial-intelligence tools to automate parts of the work while retaining control over the logic and review. It estimated that more than 20 hours of manual bridge reconciliation was reduced to under ten minutes. That is Chainalysis’s description of its own workflow and has not been independently tested here.

The practical lesson extends beyond attribution. Withdrawal resumption does not reverse a completed theft, and movement across several chains can make recovery harder. Users should check which legal entity holds their assets, what withdrawal controls exist and how a platform says losses would be handled. For personal transfers, Cristoniq’s first-transfer runbook explains how to match the asset, network and destination.

Arbitrum pauses new Stylus activations, not the entire network

The Arbitrum Security Council temporarily paused new Stylus contract activations on Arbitrum One and Arbitrum Nova. According to the official notice, new and expired Stylus contracts cannot be activated while the restriction remains in effect.

The scope is narrower than a network shutdown. Active Stylus contracts can continue until expiry and may be renewed beforehand. Solidity and Ethereum Virtual Machine deployment and execution are unaffected. Arbitrum said the reviewed findings posed a denial-of-service risk to network availability and that no user funds had been at risk. Its status page reported no incident on 2 October.

Developers should check whether a planned deployment depends on Stylus activation. Ordinary users should not interpret the notice as evidence that all Arbitrum transactions or contracts stopped working.

NEAR Intents says exploited funds were returned

NEAR Intents general manager Alex Shevchenko said the US$3.8 million taken in an exploit had been returned in full before a 4 October deadline. Crypto Times reported that a successful BNB Chain transaction carried a message saying the funds had been returned, while noting that the message did not independently establish who controlled the sending address.

The reported return changes the immediate loss position, but it is not a substitute for a technical post-mortem. A complete independent transaction reconciliation was not captured, and Shevchenko did not disclose how the alleged exploiter had been identified. Users should distinguish recovery of funds from confirmation that every affected service and security weakness has been remedied.

Illinois tax delay remains conditional

CoinDesk reported that Illinois officials and industry groups agreed to seek a six-month delay to the state’s 0.2% digital-asset tax, moving the intended start from 1 January to 1 July 2027. The agreement still requires approval from a judge.

This is neither a repeal nor a completed court order. Affected businesses may receive more preparation time if the court approves the request, while the legal dispute over the tax continues.

The fixed crypto sample moved lower

At approximately 05:07 UTC on 3 October, while crypto markets were trading continuously, every asset in the fixed five-coin US-dollar sample had a negative rolling 24-hour change. Chainlink recorded the largest percentage decline within the group.

These are rolling observations rather than formal daily closes. The selected assets differ substantially in size and design, and this crypto-only sample does not represent the whole digital-asset market or any traditional market. The figures do not establish why prices moved.

AI, Technology and Investor Tools

Gemini 4 Argon begins with restricted cybersecurity access

Google announced Gemini 4 Argon on 30 September, describing it as a model for long-running software, professional-research and cybersecurity tasks. It is initially rolling out only to selected trusted cyber defenders through Google’s Fairwind programme. It was not generally available to consumers, developers or enterprises at the research cut-off.

Google said wider access would follow after further safety testing and feedback. The company also published performance and productivity claims based on its own tests and internal examples. Those claims have not been independently tested here. Investors should distinguish a product announcement from evidence of customer adoption, revenue or lasting operating savings.

A practical workflow for AI-assisted market research

Keep the model between the question and the source, not between the source and the decision

1. Ask the tool to identify the exact filing, release or dataset needed.
2. Open the original document and record its publication date, event date and reporting period.
3. For each figure, record the instrument, currency or unit and observation time.
4. Separate reported facts, company forecasts and the model’s own inference.
5. Keep research separate from order execution while assessing a tool’s reliability.
6. Before acting, recheck the asset, legal entity, quantity, order type and price limit in the relevant platform.

A larger context window may help a model process lengthy filings, but it does not guarantee that the system will notice the decisive qualification. For cybersecurity and financial research, use read-only access where possible, restrict permissions and preserve a record of the source material used.

What to Watch Next

• Monday 5 October: watch how UK, European and US markets reopen after Friday’s employment report. Compare any rate repricing with the actual 29,000 payroll increase and the revised earlier months.
• Nike: watch Greater China revenue, digital sales, inventory and margins, alongside any changes to the estimated Pace charges and savings.
• Wetherspoon: compare the next trading update with the recent 8.6% like-for-like sales increase and management’s warning that exceptional weather helped.
• Arbitrum: check the official notice for the reopening of new Stylus activations and any fuller explanation of the denial-of-service risk.
• NEAR Intents: look for a technical post-mortem, service-restoration details and a complete account of the returned funds.
• Illinois: watch for the Sangamon County court’s decision on the proposed tax delay. Until approval is documented, 1 July 2027 remains conditional.
• Friday 6 November: the Bureau of Labor Statistics has scheduled the October US Employment Situation report for 08:30 Eastern Time. Revisions to September may be as important as the new headline figure.

This briefing provides educational information, not personalised investment, trading, tax or legal advice.