Here come the bots

Disrupt and maybe destroy

AI is hyper-charged creative destruction. It is a technology upending economic relationships and businesses. It is disruptive and markets are seeing the implications of that in real time. As of close of business on 12 February, the S&P 500 Software and Services Industry sector was down 19.7% year-to-date and 27% since the broader Information Technology sector peaked on 29 October. In the bond market, over the same period, the US High Yield Technology sub-index has seen spreads widen by 135 basis points. Individual equities and bonds have taken bigger hits. However, in their quarterly earnings announcements, the hyperscalers (cloud computing giants) increased capital spending plans once again, while semiconductor manufacturers have reported huge increases in revenue. Despite wobbles elsewhere, the S&P 500’s semiconductor index is up 4.5% this year.

Culling the coders

The AI-enabling theme is very strong. Huge commitments to develop data centres, cloud computing capabilities and large language models are generating massive revenues for companies involved in creating the infrastructure. Goldman Sachs publishes an equity index called AI Data Centers and Electrical Equipment and that index is up 26.7% this year, and more than 100% compared to a year ago. Nonetheless, companies at risk of disruption are in an existential panic. If more powerful AI agents can perform marketing, accounting, legal and human resource management tasks quicker and cheaper than existing Software as a Service (SaaS) incumbents, then users win from lower costs and productivity gains, while suppliers see their revenues undercut. This was certainly the theme of the last couple of weeks. AI adoption will be good for margins for some, and disastrous for revenues (and share prices) for others.

However, as always, the story is more nuanced. Enterprise AI adoption is still relatively limited across the broader economy, and its growth has not been as fast as some had hoped. It takes time to replace enterprise software when it is embedded in the corporate structure (imagine companies getting rid of Microsoft Office!). There are opportunities for big software providers to respond to AI competition by incorporating more AI value-added services into their own software applications. Customer service and business models are often based on legacy frameworks, and as such, AI agents cannot simply barge in and totally replace existing workflow management. It’s going to be interesting. But growth expectations and software companies’ valuations have been rightly challenged. This story has more to run and could end up undercutting broad US equity returns. The story in 2026 has been one of US index underperformance, with Asian markets delivering the strongest price gains.

More tech bonds

AI is disruptive in other ways. The hyperscalers have ramped up their corporate bond issuance to finance capital spending plans. The money being raised is sizeable. Two deals in the US corporate bond market raised $25 billion and $20 billion in the last week or so. One hyperscaler tapped the Swiss franc and sterling bond markets as well. The technology sector is rapidly becoming the source of the biggest corporate bond issuers. They are highly rated; they need to borrow and there is demand.

Corporate bond investors are going from generally funding highly regulated businesses like banks and utilities, to funding growth companies in a technology which is still new, and which might not deliver the revenues needed to make the current levels of spending viable.  In the short term this is unlikely to be a problem as borrowers have strong balance sheets and recent deals provide new opportunities for income-focused bond investors. For sterling investors, starved of issuance in recent years, the 100-year maturity bond from Alphabet, representing an additional 1% of long duration credit assets in the market, was a welcome source of coupon payments (6.125%).

In the US market, the spread on the ICE BofA Technology and Electronics bond index above US Treasuries has risen to match the overall market spread after being tighter for the last decade. So far, in the euro and sterling corporate bond markets, technology spreads are still tighter than the overall market, although these sectors make up a very small share of the entire market. As these sectors grow, they are likely to attract more retail investors who may want to mix equity and fixed income exposure to the most dynamic economic sector around. The Alphabet sterling issue is a big deal for that market.

Disrupting jobs

Another area of disruption is employment. If AI can do things quicker than humans, and carry out tasks too complex for the human brain, then why employ a person and not a machine? The extent to which this is happening today is hard to assess. There are anecdotes about graduates finding it difficult to get a job and there are fears the latest AI models could potentially render some roles such as paralegals and financial analysts redundant. In the asset management world, reporting on portfolio performance – an extremely important service offered to clients – might mostly be done by machines. Ahead of the release of the January US employment data I looked at the year-on-year change in employment across US economic sectors. The largest declines were in areas like Computers and Electrical Products, Computer Systems Design, Data Processing, and Business Support Services. AI might be responsible for at least some of these job losses - and readers will have their own views, some of which might be based on real-life experience. But legal and accounting services – supposedly prime for disruption – were amongst the biggest job gainers over the last year.

The cyclical position does not point to near-term Fed rate cuts 

Overall US jobs growth has been weak. Both official and household surveys of employment show flat growth compared to a year ago. Despite this, the January employment report was better than expected. The number of non-farm payroll jobs increased by 130,000 compared to a revised 48,000 in December and a market consensus expectation of 65,000. The unemployment rate also fell to 4.3% from 4.4%. The Fed’s three rate cuts last year were based on policy having proven too restrictive and a rising unemployment rate (it was 4.0% in January 2025). However, in January the unemployment rate fell to just below the estimate of the natural rate of unemployment (i.e. the unemployment rate estimated to be consistent with a stable inflation rate). There is a scenario in which the Fed could remain on hold for some time which may see market yields resume the upward trend they were in between October and December last year.

That contrasts with the UK. The economy ended 2025 with growth slowing. Since then, there has been renewed political uncertainty. The Bank of England strongly hinted that it might cut rates in March. Growth is soft, the government appears weak, and inflation is yet to return to target. That combination might make life difficult for investors in the gilt market, despite its attractiveness on a relative value and real yield basis. The optimistic scenario is that the government survives its current difficulties, inflation drops to 3% in January, and the BoE cuts the bank rate to 3.5% in March. It might even stop raining!

Performance data/data sources: LSEG Workspace DataStream, ICE Data Services, Bloomberg, BNP Paribas AM, as of 12 February 2026, unless otherwise stated). Past performance should not be seen as a guide to future returns.