🇸🇪 Svenska · 🇬🇧 English · 🇪🇸 Español
AI-generated briefing. Facts and links checked against current sources.
Nasdaq falls 2.9% as the AI sell-off closes the month
The Nasdaq fell 2.9% and the S&P 500 lost 1.8% on Friday, while Intel plunged nearly 29% after its results. October ended with mounting doubts about the financing and returns of AI data centres.
Why it matters: As investors prioritise cash flow over growth, the decline can spread from megacap technology to semiconductors, credit markets and Swedish industrial shares. Source: Reuters.
Chinese manufacturing slips back below the growth line
China’s official manufacturing PMI fell from 50.1 to 49.3 in October, below the 50.0 forecast and marking the third contraction in four months. Services and construction also slipped below 50 as domestic orders and export demand weakened.
Why it matters: Weaker Chinese demand is a direct risk for Swedish commodities, industrial and export companies and increases pressure for additional stimulus. Source: Reuters.
Loomis sets records for both revenue and margin
Loomis increased quarterly revenue by 6.8% to SEK 8.16 billion and EBITA to SEK 1.23 billion. Its EBITA margin rose from 13.2% to a record 15.1%, supported by secure logistics and automated cash handling.
Why it matters: The report shows that Loomis can grow while improving profitability, strengthening its capacity for acquisitions and shareholder returns. Source: Loomis.
AstraZeneca targets $1.5 billion in savings for new medicines
AstraZeneca launched a three-year programme designed to save about $1.5 billion annually and reinvest most of it in its drug pipeline. Third-quarter revenue rose 7% to $16.21 billion and core earnings per share beat forecasts.
Why it matters: The reshaping could lift long-term growth but brings up to $2 billion in one-off costs and potential cuts to older projects. Source: Reuters.
AI companies back a shared opt-out signal for web data
OpenAI, Google, Microsoft and Anthropic are supporting a common robots.txt signal that lets websites opt out of content use for AI training. OpenAI plans support within months and Google in 2027, while existing provider-specific controls will remain in place.
Why it matters: A shared standard could give publishers more control, reduce administrative work and reshape how AI companies obtain training data. Source: OpenAI and Google.
