tag

At the turn of the year, the outlook for investors is unusually difficult to call. Signs of a possible investment bubble in AI-related stocks have prompted some falls in technology stocks, but often followed by a rally. Valuations are stretched, and investment in data centres by the hyper-scalers is ambitious, but while individual institutions building AI capability may falter, many investors are looking at a broader picture in which AI may improve or even transform business performance across the economy.BlackRock, the world’s largest asset manager, declared itself continuing to be ‘pro-risk’ in its Outlook for 2026. It remains bullish regarding US tech stocks, but selectively so, stating that it is a good time for active investing, ‘for those with insights on who will capture the revenues’. The ‘micro’ has become macro – the scale of the investments in building AI capability by a small number of large tech firms is such that it has a macroeconomic impact.The projected investment in AI is projected to be as high as $8tn by 2030, while the increase to revenues for the AI hyper-scalers is estimated to be around $1.6tn per year. This, however, is not the full story. One reason that high AI-related valuations do not obviously constitute an investment bubble is that the technology has applications across the economy. BlackRock compares the AI revolution to earlier industrial revolutions such as the steam engine and mains electricity, increasing productivity and spurring economic development.Taking on debt to front-load the investment is necessary to realise the returns over a longer period, but this does come with risk to the financial system, at a time of high public sector debt. ‘Bond yield spikes could pose a risk to this financing,’ the report states. A structurally higher cost of capital raises the cost of AI-related investment.A small number of stocks with potentially high but uncertain prospects mean that divesting or staying is a big active call: There are no easy passive diversification options.The report notes that the long-term trend growth rate in GDP for the US has been around 2% per year for the past century. If the AI revolution significantly increases this, it would be historic. Such an outcome is a ‘tall order’, but is feasible, BlackRock concludes. The likely uneven impact on AI, with some businesses realising the gains more than others, calls for active investing and a rethink of traditional approaches to portfolio construction.If the AI spending fails to lift the trend growth rate above 2%, then it could crowd out non-AI investment, and potentially cause inflation to rise.A major constraint in the US is energy capacity. AI data centres could be consuming up to a fifth of electricity generated in the US by 2030. On energy, China may be winning the competition as it is proving to be able to build scale affordably in different energy sources, including hydropower, solar and nuclear as well as coal. DeepSeek, the highly competitive Chinese large language model, is more efficient on energy use than US rivals.The Franklin Templeton Institute, in its Global Investment Outlook, describes the ‘Age of Intelligence’ as a major theme. The AI revolution is just beginning, the report advises, and the potential for economy-wide efficiency gains is ‘enormous’.Franklin Templeton identifies three themes, which are interlinked: Broadening, steepening and weakening.Broadening: US equities have outperformed others for many years, and while prospects remain strong, there are opportunities emerging more broadly both within the US, and internationally. The profits of small cap US businesses are set to rise, owing to reduced interest rates and lower debt servicing costs. Assets in emerging markets show potential for attractive returns. Steepening: The combination of falling interest rates while inflation is above target will result in steeper yield curves. The Federal Reserve is likely to continue cutting interest rates in 2026, while two other factors are set to cause the yield curve to rise: Increased demand for capital to fund investment, especially in AI and energy capacity; and continued increases in borrowing by governments in the largest economies. Weakening: The US dollar weakened by around 10% in 2025, and is set to weaken further. Inflation and interest rates are lower in Europe there will be a lower cost of currency hedging. Commodity prices will likely rise, with the exception of oil. The weakening dynamic further encourages the broadening effect by enhancing potential returns in non-US assets. In emerging markets, import costs and inflation will fall allowing central banks to ease policy, underpinning bond prices. Credit ratings in emerging markets have improved.Franklin Templeton also pointed to the rise in government intervention: In trade, through tariffs and other barriers, and subsidising or otherwise protecting certain businesses or sectors. While there are legitimate reasons for some policies that curb globalisation, such as security, in general markets are better than governments at allocating capital, the report states.It is notable that neither report makes a definitive call of a bubble in AI stocks, although they do identify risks in the scale of investment, as well as systemic financial risks around leverage and high government borrowings.While there is a rational basis for being bullish, perhaps ‘carefully bullish’ might be an apt phrase.The author is a Qatari banker, with many years of experience in the banking sector in senior positions. 

By conventional indicators, the valuation of technology stocks has risen this year to take them into bubble territory. Valuations that reached multiples of forward earnings scarcely seen before indicated that they were priced for perfection. So a fall in valuations since mid-October was not a surprise.This dip may reflect caution and profit-taking. It may presage a bigger fall, or it may be a pause in a long bull market accompanying the AI revolution. The indicators are not all pointing in the same direction.So far, the stock market slide is just a correction. Markets fell in the first week of November. They nudged upwards in the week commencing Monday 10th, but then fell at the end of the week. The market as a whole remains at elevated levels compared with April, when there was a drop associated with President Donald Trump’s announcement on tariffs. To take just one example: Nvidia fell around 10% in the first week of November, but it was still around 60% higher than it was just six months earlier. The S&P 500 dipped to 6,700 on 14 November, which compares with a high of 6,920 but a low of 4,835 over the previous 12 months, and remains nearly 70% higher than November 2022. The dominance of large technology companies in aggregate market valuations has become pronounced. By the end of October, while the S&P had risen through most of the year, during that period some 397 of the stocks actually fell in value. Eight of the 10 biggest stocks in the S&P are tech firms. They account for 36% of the entire US market value, and 60% of the gains since April.Palantir Technologies, a business applications software specialist, reached a peak valuation of 230 times future earnings. In early November, it was revealed that prominent hedge fund manager Michael Burry took a $912mn position against Palantir, whose stock has fallen from over $200 per share to around $170, though remains more than 150% higher over the year. Mr Burry later closed his hedge fund Scion Asset Management.There has been an uneven pattern to the sell-off, following earnings reports in late October. Meta, the owner of Facebook, fell 12% over concern of its high investment levels in AI, given disappointing returns from its investment in virtual reality, though has recovered slightly. Alphabet, the owner of Google, rose 3%, though has since dipped by around 5%, and Microsoft fell by just 3%, then fell further before a partial recovery.Unlike the dotcom start-ups of 25 years ago, the tech firms have strong revenues and a sound business model. Their services extend far beyond AI, covering business application software and cloud computing. In the case of Amazon, it is a general retailer as well as a tech firm. A strong argument is that much of the investment in AI is from large, profitable companies with a strong cash position.The hyper-scalers, Amazon, Meta, Alphabet and Microsoft, all have strong underlying global businesses. The scale of the investment in data centres being planned has caused some investors to be concerned, however. Some tech firms have been issuing bonds; for example, in late October Meta announced a $25bn bond issuance to finance AI investment, following Oracle’s $18bn bond sale in September. In early November, yields on big tech firms’ bonds started to rise. Oracle’s stock suffered bigger falls in the middle of November, with investors concerned over debt, heavy reliance on OpenAI, negative free cash flow. It emerged that the outgoing CEO Safra Catz sold $2.5bn of Oracle shares this year.Also, 10 loss-making AI specialist start-up companies have between them been valued at nearly $1tn, while there have been patterns of circular financing, especially concerning OpenAI.Another dimension is that there is softening economic data from the wider economy. With the US government lockdown entering its second month, there has been no official jobs data since 5 September. Analysts and economics have been relying on private sources. Data from the private company Challenger, Gray and Christmas showed the highest level of October job lay-offs since 2003, while the payroll company ADP reported that US companies shed 32,000 jobs in September, the biggest fall in two and a half years.Earnings from mainstream businesses have disappointed. The stock of the popular restaurant chain Chipotle fell 13% in late October following disappointing quarterly results.The price of bitcoin has been unpredictable in recent weeks. It often rises in a counter-cyclical manner, increasing as stock market falls, but cryptocurrencies generally were off their highs at the time of the wider market correction. Bitcoin fell from around $125,000 on 7 October to just below $100,000 by mid-November. Many bitcoin investors are leveraged, and some forced, automated sales are likely to have occurred, accelerating and drop in price. Gold has fallen from a high of $4,400 per ounce to around the $4,100.The AI investment industry is one of the few sectors to be registering growth, so if technology firm leaders fall short of their ambitions, the impact would ripple outside the industry.Against that, the bearish commentators point to the relatively narrow foundation of asset price investment, and debt and macro-economic fragility in the higher-tariff era, a combination that compares unfavourably with the more benign macro-economic picture of 2000-2001.The emerging technology of AI comes during an extended period of cheap money and globalisation, including of retail investing, and rapid growth of private credit. The worldwide exposure of investors to US stocks are part of a highly inter-connected system. High levels of government debt limit the extent of any fiscal stimulus following a shock. A loss in market value of the same proportion as the dotcom crash would have a far bigger impact on the real economy. The economist Gita Gopinath has estimated that it would cause a loss of $20tn for US households, or 70% of GDP. Foreign investors might lose $15tn.There are two dimensions to risk assessment: Likelihood, and impact. The likelihood of an asset price collapse that is equivalent in proportional terms to that of the dotcoms would not appear to be high, although it is possibility. The impact would be seismic, and felt across the global economy.The author is a Qatari banker, with many years of experience in the banking sector in senior positions.

Turbulence in technology stocks could ratchet higher in the coming week as investors react to the quarterly report from Nvidia Corp, the world's largest company by market value that is at the heart of Wall Street's artificial intelligence trade. On Thursday, the benchmark S&P 500 equity index gave up gains from earlier in the week, as uncertainty about the economic outlook and path for US interest rates undercut optimism over the end of the longest-ever US government shutdown. Investors remained skittish about vulnerability to technology shares, which stumbled this month on concerns AI exuberance has driven up valuations to expensive levels. With its AI chips, semiconductor giant Nvidia has been a bellwether for the theme that has lifted shares of an array of tech names as well as other companies involved in the vast infrastructure expansion to support AI use. Nvidia is the "epicentre" of the build-out of AI, so its results after the bell on Wednesday will be important to the tech sector as well as areas such as industrials and utilities, said Matt Orton, chief market strategist at Raymond James Investment Management. "If you don't see the growth that I think the market is expecting around Nvidia or the positive commentary that we are likely to get from Nvidia going forward, I think you're going to see more of a dent to those sorts of trades," Orton said. Nvidia shares have soared about 1,000% since the launch of ChatGPT in November 2022. This includes a year-to-date gain of nearly 40% that made Nvidia the first company to surpass $5tn in market value last month. That market heft means the stock's moves can sway equity indexes. Nvidia carries an 8% weight in the S&P 500 and a roughly 10% weight in the widely followed Nasdaq 100. **media[381893]** Analysts on average expect the company to post a 53.8% year-over-year rise in fiscal third quarter earnings per share, on revenue of $54.8bn, according to LSEG. Analysts have also been getting more bullish about the company's future performance, with expectations for the company's fiscal 2027 revenue rising 15% since late May to about $285bn currently, according to LSEG data. "The assumptions that the market is making are positive, it's getting priced into the stock, and how the company guides will be very important," said Melissa Otto, head of research at S&P Global Visible Alpha. Investors will also focus on commentary from Nvidia related to demand or spending trends. Capital expenditures from hyperscalers such as Microsoft and Amazon earlier in the reporting season indicated no signs of slowing in the build-out of data centres and other AI infrastructure. "You're not supposed to have any weakness given all the capital spending commitments from various companies," said Jimmy Chang, chief investment officer of Rockefeller Global Family Office. "Demand should still be looking pretty solid in the current environment." Nvidia's report is one of the biggest remaining market catalysts in 2025. The S&P 500 is logging a roughly 15% year-to-date gain, but Wall Street is wary of concerns stocks are in an "AI bubble." Investors appear to be bringing more scrutiny to AI investment announcements, said James Ragan, co-CIO and director of investment management research at DA Davidson. **media[381894]** "We're moving into a stage where investors are going to demand a little bit more proof of concept in terms of what are the returns, what are the cash flows," Ragan said. Aside from Nvidia's results, quarterly earnings from retailers are due in the coming week including from Walmart and Home Depot. There could also be a batch of economic data releases that were delayed during the shutdown. While the S&P 500 tech sector has struggled so far this month, other sectors are logging solid gains in that time, including healthcare, materials and financials. "There's a realisation that for investors, maybe that AI is not the only game in town," Ragan said.