Equity markets continued to shrug off the tariff-induced volatility that roiled stocks earlier this year. The S&P 500 returned 14.8% year-to-date through September 30. This performance was driven by AI-related names and more speculative, lower-quality stocks. Low quality outperformed high quality by nearly ten percentage points among both large and small cap stocks. The 109 stocks in the Goldman Sachs US TMT AI Basket (which is a mix of low- and high-quality stocks) returned 42% YTD and their combined market capitalization totaled approximately $25 trillion dollars – over 40% of the market capitalization of the S&P 500. Foreign equity indexes outperformed but with significant return dispersion among sectors and countries. European banks and defense stocks drove the MSCI EAFE. China, along with South Korean and Taiwanese semiconductor stocks contributed to the outperformance of emerging markets. India, the best performing emerging market over the last decade, significantly underperformed.
This process of creative destruction is the essential fact about capitalism. – Joseph Shumpeter
The rise of Artificial Intelligence dramatically increased investment risks over the past two years. The boom so far has been limited to three primary groups of companies – the chip designers and makers such as NVIDIA and Taiwan Semiconductor, their hyperscaler customers such as Microsoft or Meta and a group of companies selling ancillary products and services to this investment boom. The latter group includes everything from suppliers of HVAC services and traditional electricity generation for new data centers to more speculative stocks with yet unrealized plans to develop small modular nuclear reactors or fuel cells. It would be a mistake to dismiss this simply as a ‘low quality’ phenomenon. NVIDIA and the hyperscalers meet all the definitions of quality stocks.
However, current valuations and the magnitude of the investment pose risks to these stocks and the broad market. Capital investments by the top four hyperscalers – Amazon, Alphabet, Meta and Microsoft – averaged around $100B in 2020 and 2021 before the launch of ChatGPT. Spending this year is expected to exceed $300 billion and surpass $400 billion next year. Total revenue for these firms, after backing out Amazon’s retail businesses, runs around $1 trillion. Tech industry research firm Garter estimates total AI-related investment worldwide at $644 billion1. Most of these funds have been used to fill data centers with servers running NVIDIA’s latest chips.
Chips are not long-lived assets. NVIDIA’s latest GPUs have a 3–5-year lifespan under heavy data center workloads. Revenue from generative AI-based services must scale rapidly from current levels for providers to earn sufficient returns for their shareholders. This creates a risk that the current AI boom will repeat the pattern of past investment booms around transformative technologies, from railroads in the 19th century to the buildout of internet infrastructure in the early 2000s where consumers reaped the benefits of a glut of capacity while producers struggled to earn returns on their investments.
The concept of investing in high-quality companies is not new. The early pioneers of value investing, notably Benjamin Graham, emphasized the importance of financial health and earnings stability as crucial criteria for stock selection. Traditional active managers applied these principles in a discretionary, qualitative manner. With the development of academic finance in the 1960s and 70s focus shifted to the idea than an efficient market valued stocks fairly in proportion to how risky they were relative to the overall market. Under this view, high quality stocks – companies with high profitability, strong balance sheets and sound management – should deliver lower returns than higher risk, more speculative names. A large body of research over the past 30 years has demonstrated the opposite is true – higher quality companies in fact delivered higher returns, contradicting the prediction of established financial theory.
Why should objectively lower risk stocks outperform? Four explanations have emerged that likely all contribute to the phenomenon:
Preference for “Lottery Tickets”: Many investors exhibit a preference for assets with the perceived potential for “lottery-ticket” like payouts, particularly if they are accompanied by an exciting story of a new technology or business model. Similarly, distressed companies on the verge of bankruptcy offer astronomical returns if their financial health recovers (it typically does not). This systematic demand can bid up the prices of junk stocks, leading to their long-run underperformance.
Overconfidence and Representativeness: Investors may be overconfident in their ability to analyze and turn around distressed, low-quality companies or predict the success of long-shot speculative growth stocks. Conversely, they may see the stable, predictable nature of high-quality firms as “boring” and offering limited upside, thereby undervaluing their long-term compounding potential.
Present Bias and Myopia: Investors tend to extrapolate recent performance and chase the performance of speculative stocks, buying after a substantial increase in value.
Constraints on Leverage: Investors, particularly professional active managers, want to outperform the benchmarks against which they are measured. As market downturns are relatively infrequent, they prefer to outperform during positive years. They typically cannot utilize leverage but can get similar results by overweighting more volatile and cyclical lower quality stocks. This leads to an overvaluation of lower quality stocks.
We tracked the benefits of quality in the performance of active investment managers beginning in the early 2000s, noticing that investment processes that focused on profitability, balance sheet strength and quality of management tended to have more consistent outperformance, particularly during market downturns. This became integrated into our overall investment strategy and a focus on quality became a requisite characteristic for our manager selection process. Academic studies of the quality factor also began emerging during this period, along with the first investment products that attempted to systematically invest in higher quality companies. These early attempts at creating “quality index funds” suffered some missteps during the 2008 Global Financial Crisis where
an over-emphasis on earnings consistency and return on equity and a lack of risk controls on sector weights allowed many of these strategies to overweight highly leveraged financials such as Citigroup or AIG, which led to underperformance during the downturn.
As research continued in the 2010s, a consensus emerged that quantified quality around three primary pillars – profitability, balance sheet strength and the integrity of financial reporting. Profitability encompasses several measures of how well a company generates financial returns for shareholders. Balance sheet strength is a measure of safety and correlates highly to outperformance during a market downturn. Quantifying the integrity of financial reporting utilizes several metrics, such as measuring how much of a company’s stated earnings come from actual cash flow vs accruals or other financial reporting tricks. This last part would screen out companies with questionable accounting methods such as Enron in the early 2000s. By the late 2010s, products were being launched that utilized multiple measures for each of the three pillars and imposed top-down risk controls on position and industry limits. We have been tracking these systematic approaches to quality and believe they warrant consideration for portfolio construction. Importantly, quality is not the only factor in equity investing –
valuation and growth matter as well. The goal is to integrate a sufficient tilt toward quality stocks that allows for smoother, more consistent compounding while continuing to participate in market environments where quality may be out of favor.
Rapid technological change, shifting market leadership, and evolving investor behavior make this a challenging investment environment. While the rise of artificial intelligence has introduced new risks and opportunities, the enduring value of quality—profitability, balance sheet strength, and transparent financial reporting—remains clear. History shows that periods of exuberant capital investment often benefit consumers more than producers, underscoring the importance of disciplined investment strategy and a focus on fundamentals. As we navigate the uncertainties ahead, a thoughtfully diversified portfolio anchored in quality offers the best chance of achieving long-term investment goals.
Home » Market Commentaries » 3rd Quarter 2025 Market Commentary
3rd Quarter 2025 Market Commentary
Equity markets continued to shrug off the tariff-induced volatility that roiled stocks earlier this year. The S&P 500 returned 14.8% year-to-date through September 30. This performance was driven by AI-related names and more speculative, lower-quality stocks. Low quality outperformed high quality by nearly ten percentage points among both large and small cap stocks. The 109 stocks in the Goldman Sachs US TMT AI Basket (which is a mix of low- and high-quality stocks) returned 42% YTD and their combined market capitalization totaled approximately $25 trillion dollars – over 40% of the market capitalization of the S&P 500. Foreign equity indexes outperformed but with significant return dispersion among sectors and countries. European banks and defense stocks drove the MSCI EAFE. China, along with South Korean and Taiwanese semiconductor stocks contributed to the outperformance of emerging markets. India, the best performing emerging market over the last decade, significantly underperformed.
The rise of Artificial Intelligence dramatically increased investment risks over the past two years. The boom so far has been limited to three primary groups of companies – the chip designers and makers such as NVIDIA and Taiwan Semiconductor, their hyperscaler customers such as Microsoft or Meta and a group of companies selling ancillary products and services to this investment boom. The latter group includes everything from suppliers of HVAC services and traditional electricity generation for new data centers to more speculative stocks with yet unrealized plans to develop small modular nuclear reactors or fuel cells. It would be a mistake to dismiss this simply as a ‘low quality’ phenomenon. NVIDIA and the hyperscalers meet all the definitions of quality stocks.
However, current valuations and the magnitude of the investment pose risks to these stocks and the broad market. Capital investments by the top four hyperscalers – Amazon, Alphabet, Meta and Microsoft – averaged around $100B in 2020 and 2021 before the launch of ChatGPT. Spending this year is expected to exceed $300 billion and surpass $400 billion next year. Total revenue for these firms, after backing out Amazon’s retail businesses, runs around $1 trillion. Tech industry research firm Garter estimates total AI-related investment worldwide at $644 billion1. Most of these funds have been used to fill data centers with servers running NVIDIA’s latest chips.
Chips are not long-lived assets. NVIDIA’s latest GPUs have a 3–5-year lifespan under heavy data center workloads. Revenue from generative AI-based services must scale rapidly from current levels for providers to earn sufficient returns for their shareholders. This creates a risk that the current AI boom will repeat the pattern of past investment booms around transformative technologies, from railroads in the 19th century to the buildout of internet infrastructure in the early 2000s where consumers reaped the benefits of a glut of capacity while producers struggled to earn returns on their investments.
The concept of investing in high-quality companies is not new. The early pioneers of value investing, notably Benjamin Graham, emphasized the importance of financial health and earnings stability as crucial criteria for stock selection. Traditional active managers applied these principles in a discretionary, qualitative manner. With the development of academic finance in the 1960s and 70s focus shifted to the idea than an efficient market valued stocks fairly in proportion to how risky they were relative to the overall market. Under this view, high quality stocks – companies with high profitability, strong balance sheets and sound management – should deliver lower returns than higher risk, more speculative names. A large body of research over the past 30 years has demonstrated the opposite is true – higher quality companies in fact delivered higher returns, contradicting the prediction of established financial theory.
Why should objectively lower risk stocks outperform? Four explanations have emerged that likely all contribute to the phenomenon:
We tracked the benefits of quality in the performance of active investment managers beginning in the early 2000s, noticing that investment processes that focused on profitability, balance sheet strength and quality of management tended to have more consistent outperformance, particularly during market downturns. This became integrated into our overall investment strategy and a focus on quality became a requisite characteristic for our manager selection process. Academic studies of the quality factor also began emerging during this period, along with the first investment products that attempted to systematically invest in higher quality companies. These early attempts at creating “quality index funds” suffered some missteps during the 2008 Global Financial Crisis where
an over-emphasis on earnings consistency and return on equity and a lack of risk controls on sector weights allowed many of these strategies to overweight highly leveraged financials such as Citigroup or AIG, which led to underperformance during the downturn.
As research continued in the 2010s, a consensus emerged that quantified quality around three primary pillars – profitability, balance sheet strength and the integrity of financial reporting. Profitability encompasses several measures of how well a company generates financial returns for shareholders. Balance sheet strength is a measure of safety and correlates highly to outperformance during a market downturn. Quantifying the integrity of financial reporting utilizes several metrics, such as measuring how much of a company’s stated earnings come from actual cash flow vs accruals or other financial reporting tricks. This last part would screen out companies with questionable accounting methods such as Enron in the early 2000s. By the late 2010s, products were being launched that utilized multiple measures for each of the three pillars and imposed top-down risk controls on position and industry limits. We have been tracking these systematic approaches to quality and believe they warrant consideration for portfolio construction. Importantly, quality is not the only factor in equity investing –
valuation and growth matter as well. The goal is to integrate a sufficient tilt toward quality stocks that allows for smoother, more consistent compounding while continuing to participate in market environments where quality may be out of favor.
Rapid technological change, shifting market leadership, and evolving investor behavior make this a challenging investment environment. While the rise of artificial intelligence has introduced new risks and opportunities, the enduring value of quality—profitability, balance sheet strength, and transparent financial reporting—remains clear. History shows that periods of exuberant capital investment often benefit consumers more than producers, underscoring the importance of disciplined investment strategy and a focus on fundamentals. As we navigate the uncertainties ahead, a thoughtfully diversified portfolio anchored in quality offers the best chance of achieving long-term investment goals.
1 www.hostinger.com/tutorials/llm-statistics