--- title: "Investors monitor AI-related stocks in US and South Korea amid leverage concerns; Debt funding for AI tops $250B in 2026." sdDatePublished: "2026-08-08T11:33:00Z" source: "https://www.schroders.com/en-ch/ch/wealth-management/insights/spotlight-on-technology-why-have-ai-related-stocks-been-so-volatile/" topics: - name: "financial service" identifier: "medtop:20001370" - name: "artificial intelligence" identifier: "medtop:20001298" - name: "semiconductor and electronic component" identifier: "medtop:20000230" - name: "stock activity" identifier: "medtop:20000186" - name: "market trend" identifier: "medtop:20000284" locations: - "United States" - "China" - "South Korea" --- Investors monitor AI-related stocks in US and South Korea amid leverage concerns; Debt funding for AI tops $250B in 2026. Spotlight on technology: why have AI-related stocks been so volatile? Spotlight on technology: why have AI-related stocks been so volatile? Ongoing concerns about infrastructure spending, competition from low-cost models and speculative trading have all played a role. Yet periods of volatility are an inevitable part of the technology cycle and do not undermine the long-term investment case for AI. Stocks connected to the AI supply chain have been volatile over the last month. After weeks of declines, some of the largest US technology firms – including Microsoft and Amazon – saw share prices jump by over 15% following their latest results. South Korea's stock market, which is dominated by two large semiconductor makers, fell as much as 40% from its peak, before a near-20% recovery on July 31st. 1 These dramatic swings have been attributed in part to leverage (borrowing money to gain or increase exposure to markets), with significant use of margin trading 2 and leveraged Exchange Traded Funds (ETFs). 3 Retail investors in South Korea used these ETFs to gain exposure to memory-chip makers SK Hynix and Samsung and were a significant driver of market volatility. We have also seen institutional investors using leverage. Situational Awareness, an AI-focused hedge fund that reportedly managed over $20 billion at its peak, is just one example. As the sector came under pressure, the fund was forced to unwind positions, potentially amplifying the sell-off. Yet recent volatility is not just a function of retail trading activity or one hedge fund’s poor risk management. Investors are also grappling with fundamental questions about the sector’s economics. The first issue is the sheer scale of investment required to support AI – and how it is being funded. The world's largest technology companies continue to commit vast sums to data centres, semiconductors and other infrastructure. As the sector moves away from its traditional “capital light” business model 4 , shareholders are increasingly concerned about the level of returns on these investments. With spending running ahead of cashflows, much of the investment will be funded by borrowing: debt issuance from the largest technology companies is set to exceed $250 billion in 2026, according to research from Goldman Sachs. 5 So technology companies must also reassure bond markets that their spending will not undermine balance sheets that were once regarded as among the strongest in the corporate world. The second concern is the rapid emergence of lower-cost AI models, especially from China. For many everyday tasks, these models may prove more than adequate – and businesses may turn to them instead of the most advanced models from “frontier” developers such as OpenAI and Anthropic. While lower cost models should support broader AI adoption, they may also intensify competition, putting pressure on Large Language Model pricing. This would have implications across the AI ecosystem. For some observers, the situation evokes memories of the late 1990s, when investment in telecoms infrastructure ran significantly ahead of demand. The picture today is very different. Unlike in the 1990s, the current AI investment boom is accompanied by substantial customer demand. Despite the risk of competition from low-cost rivals, Anthropic has seen its revenues surge from an annualised rate of $9 billion at the end of 2025 to approximately $50 billion at mid-year. More recently, OpenAI announced that revenue in July exceeded that of the prior three months combined, providing further evidence of acceleration. Established technology companies are also benefiting: Alphabet, Amazon and Microsoft all reported stronger-than-expected revenue growth in their cloud divisions, largely driven by demand for AI-enabled services. Backlogs remain substantial, suggesting revenue growth will remain high for the foreseeable future. There is also growing evidence that businesses outside the technology sector are incorporating AI into their operations, supporting the view that demand is broad and durable. This does not guarantee that every company in the AI supply chain will prosper. But it does suggest that recent market volatility is not primarily a story about insufficient demand. Rather, it reflects uncertainty over where value will ultimately accrue within the AI ecosystem. The AI ecosystem is complex, as our infographic at the end of this article shows. What remains unclear is who will benefit most from the industry's development. Will value accrue primarily to the companies building AI models? Or to the cloud providers supplying the computing power? Or to software companies embedding AI into products and services? The market is still trying to answer those questions. Importantly, the investable AI ecosystem extends far beyond model developers such as OpenAI and Anthropic. Behind every AI application sits an extensive network of cloud suppliers, semiconductor manufacturers, data centres, networking equipment providers and power infrastructure companies. In several areas, genuine bottlenecks have emerged as demand outpaces supply. Memory chipmakers such as Micron and SK Hynix are a good example. Demand for advanced memory used in AI systems has risen sharply, but expanding production facilities takes several years. As a result, prices for memory chips have surged, as have profits and share prices of manufacturers. However, the recent sell-off suggests investors have their doubts about the sustainability of such elevated profitability. We may well see other parts of the AI ecosystem fall in and out of favour as the story evolves. The winners of the past two years may not be the winners of the next two. This is one reason why we believe an active approach is particularly important. While AI is often discussed as a single investment theme, it is in fact a collection of very different businesses, each exposed to different drivers. As the economics of the industry become clearer, we would expect opportunities and risks to shift across the ecosystem. Active managers can adjust exposures accordingly. The key uncertainty is not whether AI will have a significant economic impact, but where value will ultimately accrue within the ecosystem. With that in mind, we think it is helpful to consider three broad scenarios. AI adoption continues to broaden across industries. Lower-cost models capture significant volume, but leading models retain pricing power through superior performance, reliability and support and capture more of the value. This would be similar to the experience of patented and generic drugmakers in the pharma industry. In this scenario, cloud providers and chipmakers remain important beneficiaries given continued demand for computing capacity, although returns become more differentiated than they have been to date. Frontier model providers such as OpenAI and Anthropic can generate significant revenues if they maintain leadership. Investment remains high because demand continues to exceed expectations. Cloud providers, AI developers and much of the supporting infrastructure ecosystem benefit as adoption expands into new use cases, including AI agents, robotics, autonomous systems and applications that have yet to emerge. Demand for AI begins to slow as use cases fail to generate significant value to justify the cost. This could be driven by a slowdown in AI development due to a technical hurdle of some kind. The gap between premium and lower-cost models narrows significantly. Pricing power weakens, margins come under pressure and returns on investment disappoint. In this scenario, software companies may emerge as some of the biggest winners because they gain access to powerful AI capabilities at ever lower cost. Cloud providers and model developers would face greater pressure. As markets grapple with the economics of the AI industry, further periods of volatility are likely. Such growing pains are a normal part of the adoption cycle for transformative technologies. The industry's next phase may well produce different winners and losers, with leadership shifting as adoption broadens and competitive dynamics evolve. The current volatility is therefore best seen not as a sign that the AI story is running out of steam, but as evidence of a rapidly developing industry moving towards maturity. Source: Cazenove Capital, as of 31 July 2026. This document is for information purposes only and is intended for clients of Cazenove Capital. It is not personal advice, and it does not take account of individual circumstances. Any views expressed are those of the author(s) as at the date of publication and may change. Investing involves risk, including the risk of loss of capital. Past performance is not a reliable indicator of future results. The value of investments and the income from them can go down as well as up, and you may not get back the amount originally invested. Active management does not guarantee outperformance. Forecasts and scenarios are illustrative, not guaranteed, and not a reliable indicator of future performance. Any reference to individual stocks and sectors is for illustrative purposes only and is not a recommendation to buy or sell any financial instruments. 1 Source: LSEG Datastream, August 5th 2 In a margin trading account, investors borrow money from a broker to increase the size of their investment positions. Small price movements can therefore lead to larger gains or losses. 3 Exchange-traded funds designed to deliver a multiple of the daily return of an underlying stock or index. They typically use derivatives and borrowing and can be very volatile. 4 A capital-light business model is one that requires relatively little capital investment, allowing growth with limited spending on physical assets. 5 Source: Goldman Sachs, Expectations for Hyperscaler Debt Issuance: A Top Down Approach, 27 July 2026. Hyperscalers include Meta, Microsoft, Alphabet, Amazon and Oracle. This communication is marketing material. The views and opinions contained herein are those of the author(s) on this page, and may not necessarily represent views expressed or reflected in other Schroders communications, strategies or funds. This material is intended to be for information purposes only and is not intended as promotional material in any respect. The material is not intended as an offer or solicitation for the purchase or sale of any financial instrument. It is not intended to provide and should not be relied on for accounting, legal or tax advice, or investment recommendations. Reliance should not be placed on the views and information in this document when taking individual investment and or strategic decisions. Past performance is not a reliable indicator of future results. The value of an investment can go down as well as up and is not guaranteed. All investments involve risks including the risk of possible loss of principal. Information herein is believed to be reliable but Schroders does not warrant its completeness or accuracy. Some information quoted was obtained from external sources we consider to be reliable. No responsibility can be accepted for errors of fact obtained from third parties, and this data may change with market conditions. This does not exclude any duty or liability that Schroders has to its customers under any regulatory system. Regions sectors shown for illustrative purposes only and should not be viewed as a recommendation to buy sell. The opinions in this material include some forecasted views. We believe we are basing our expectations and beliefs on reasonable assumptions within the bounds of what we currently know. However, there is no guarantee than any forecasts or opinions will be realised. These views and opinions may change. The content is issued by Schroder Investment Management Limited, 1 London Wall Place, London EC2Y 5AU. Registered No. 1893220 England. Authorised and reg