Not bear feeding time just yet

Not knowing but hoping

Expectations (hope) and uncertainty are principal components of investing. Bull markets survive when expectations are strong enough to overcome concerns about uncertain (unwarranted) outcomes, keeping investors invested and allowing marginal buyers to push up prices. Expectations themselves need to be built on solid foundations – economic growth; rising corporate profits; promises of beneficial new goods, services and technologies; as well as stable interest rates and inflation. All of this was vividly illustrated in 2025. Between May and October, the S&P 500, alongside many other global equity markets, and the technology sector all went up, marking a succession of new all-time highs. The demand for semiconductors; Nvidia’s emergence as the most valuable company in the world; announcements of co-investment financing deals between technology players; and the emergence of an ever-growing number of use-cases for AI contributed to the zeitgeist. The backdrop was an improving macroeconomic outlook, particularly once there was more policy clarity following Liberation Day’s tariff announcements.

October break

But something changed in the third quarter of 2025. AI uncertainty became a larger burden for investors – would the significant capital expenditure ultimately deliver wealth benefits? This uncertainty rests on concerns that too much is being invested relative to future revenues, and that AI will prove so disruptive that certain businesses, especially in the services sector, will see their own revenues put under severe pressure. The technology bull market’s current run peaked on 29 October. Whether market performance since then is just a pause, or the start of a more profound re-rating - and what it means for US exceptionalism - are questions that are core to the outlook. While investors look for clarity around AI, bond and equity markets where AI concentration is less intense are still performing. Even within the US, the equally weighted S&P 500 is up almost 8% since then while the market cap-weighted index is flat.

Dispersion

The S&P 500 achieved its latest all-time high at the end of January and, as of 25 February, it was just 0.5% below that. But the IT sector, particularly software, has underperformed – with the latter being 25% below its October all-time high. Of the so-called ‘Magnificent Seven’ group of mega-cap technology shares, only Google’s parent Alphabet has enjoyed an all-time high in 2026. Some stocks have already displayed bear market price adjustments. The share price performance dispersion reflects the need for investors to now focus on winners and losers. The concentration of technology funds in the US market has meant the US has underperformed overall (the S&P 500 is up 1.5% year-to-date, the Korean market is up 49.9%).

Another consideration for investing in technology stocks is the shift towards borrowing. There have been large corporate bond issues by tech companies since October as capital spending is no longer being fully funded out of (still very strong) free cash flow. Leverage is low, balance sheets are strong, and ratings are solid. But bond investors are now in line for a slice of revenues in terms of coupon payments, which of course are senior to retained earnings, buybacks and dividends. That may also be a consideration for equity investors in technology – at the margin is a slightly higher equity risk premium (a lower price-to-earnings ratio) required?

Macroeconomic uncertainty as well

Numerous factors are troubling the economic outlook - policy uncertainty, geopolitics, the US labour market’s unsettled state, and sticky inflation suggest a recession, an inflationary boom, and stagflation are all distinctly possible. And AI creates the potential for a more dystopian additional scenario. Growth might well be bolstered by AI, but all the value will be seen by ‘machines’ – or the machine owners - while unemployment rises and consumer spending potentially collapses. This was the future scenario postulated by a couple of research reports doing the rounds recently.

Even the cyclical outlook is unclear, particularly in the US. The labour market is weak – there was no job growth in 2025 and most of the gains in the January employment report were in the healthcare sector. Real disposable income has been flat for months. Core Personal Consumption Expenditures inflation is still at 3.0% and markets still expect the Federal Reserve to cut rates at least twice this year. But this seems to be more in the hope that new Fed Chair nominee Kevin Warsh will push this through, without the data at the time justifying it. The US economy grew by 2.2% in 2025, with no employment growth and no wage growth in real terms. The machines might already be winning.

Still a bull market

On the whole, expectations (hope) still outweigh uncertainty. That equation has shifted in the software sector but some of that move may even be overdone. Elsewhere, as I say, market performance is strong. A lot of the macro and technology-related uncertainty is concentrated in the US – hence the underperformance of US equities and the dollar. We are likely seeing a steady shift in global asset allocations away from the US, given sentiment among investors over the last year. For Europe, the story might not be that exciting, but it is coherent – there is a recognition that Europe needs to own its destiny. There is less uncertainty over interest rates, fiscal policy is largely under control, and there is potential to unleash a lot of savings if the European Union’s Capital Markets Union initiative can progress.

Bonds

In bond markets, many total return indices are at all-time highs. Leveraged loans and other private credit strategies have suffered from the software shake-out but it has not had a broad-based effect. Credit spreads have widened a little, but owners of investment-grade or high-yield bond funds are up year to date, in total return terms. Expectations and hopes are more limited in this fixed income environment but if investors get a return close to the yield, with a little kicker from actively managed strategies, they should be happy. 

Don’t underestimate the future

The environment around AI is changing quickly. It is hard to have high convictions other than the technology is set to bring profound changes to the economic environment. For now, AI spending remains high and the realisation of what AI can do is creating anxiety as well as excitement. For the near term, the clear winners are the semiconductor and hardware manufacturers - just look at Nvidia’s latest set of impressive numbers. But even here there are some doubts as to whether the recent pace of revenue growth can be sustained. 

The challenge is that it is impossible to know when the AI tipping point is reached. In my opinion we are not there yet, as the scale of beneficial AI adoption and use throughout the economy is still in its relative infancy. But the hyperscalers need to demonstrate their ability to develop and sell cloud computing capabilities as well as AI services to sustain robust revenues and not let concerns about their new penchant for debt undermine the investment case. For the rest of the economy, it’s now either “adapt to AI” or “cease to thrive”. Ultimately, how that goes will determine stock price performance. For those investors who don’t want to bet on the outcome, then physical over digital is a strategy likely to push investors to seek even more diversification in global portfolios.

Performance data/data sources: LSEG Workspace DataStream, ICE Data Services, Bloomberg, BNP Paribas AM, as of 26 February 2026, unless otherwise stated). Past performance should not be seen as a guide to future returns.