Talk Your Book: AI Winners & Losers

Episode Summary

Executive Summary: Dan Chung of Alger argues AI is a transformative but very different boom from the 1990s internet era because today’s leaders are profitable, mature giants with real revenues and cash flows. The discussion focused on identifying AI winners and losers, managing valuation and volatility, and why Alger’s concentrated ATFV ETF leans heavily into highest-conviction names like NVIDIA, Nebius, and Google while remaining cautious on Meta’s longer-term risks.

Main Topics: AI boom vs. the 1990s internet bubble (Priority: 5/5): Chung contrasts today’s AI investment cycle with the late-1990s internet boom, emphasizing that current leaders are large, profitable, established companies rather than speculative startups. This changes both the quality of capital allocation and the probability of long-term winners surviving. Concentrated portfolio construction and conviction (Priority: 5/5): The conversation explains why Alger 35 ETF is highly concentrated and how the team justifies large position sizes through bottom-up fundamental research, scenario analysis, and a focus on companies with durable moats and major AI exposure. AI winners, losers, and market volatility (Priority: 5/5): Chung says the market is still sorting out which incumbents and newer entrants will win from AI. He sees rapid stock swings driven by sentiment, but believes the market eventually separates strong companies from weak ones over months and years. Data centers, infrastructure, and neo-clouds (Priority: 4/5): A major theme is the buildout of AI infrastructure: semiconductors, data centers, liquid cooling, power, and related suppliers. Chung highlights Nebius as an example of a less obvious but highly compelling AI infrastructure winner discovered through fundamental research. Meta vs. Google as AI platform cases (Priority: 5/5): The discussion contrasts Meta and Google. Google is viewed as stronger long term because it combines search, cloud, YouTube, and Waymo, while Meta is seen as more exposed to ad saturation and existential uncertainty despite strong current profitability. Risk management through scenarios and position sizing (Priority: 4/5): Chung describes Alger’s approach to risk as balancing bear, base, and bull cases, then sizing positions according to risk-reward. He stresses that price can move much faster than fundamentals, especially in volatile software and AI-related stocks.

Key Arguments: Today’s AI leaders are fundamentally stronger than the dot-com era leaders: they have revenues, profits, and mature management teams. AI spending may be justified overall, but not every company participating in the buildout will win; many will fail or be disrupted. Market volatility around AI names reflects sentiment and narrative shifts, not always underlying fundamentals. Concentrated portfolios can be appropriate in disruptive periods if backed by deep research and careful scenario analysis. Nebius was identified early because fundamental research uncovered assets and capabilities the market had not yet recognized. Meta may be a short-term AI beneficiary but has more long-term existential risk than Google because it relies mainly on advertising. Google’s AI strategy is more durable because it spans search, cloud, YouTube, and Waymo, not just ads. Position sizing and valuation matter more than predicting exact timelines; being right on the theme is not enough if expectations get ahead of reality.

Data Points: ATFV top 10 holdings weight: 64% - Alger 35 ETF concentration level as of April 30. NVIDIA weighting in ATFV: 14.3% - Compared with 7.8% weight in the S&P 500 as of April 30. NVIDIA active overweight vs. S&P 500: 6.5 percentage points - Difference between ATFV and S&P 500 weights as of April 30. Nebius weight in ATFV: 5.0% - A non-index holding highlighted as a high-conviction AI infrastructure name. Western Digital weighting in ATFV: 4.8% - Compared with 0.24% weight in the S&P 500 as of April 30. Meta valuation discussion: 14-15x earnings - Approximate forward P/E cited as evidence the stock looks cheap despite strategic concerns. Meta user scale: 3-4 billion users - Referenced across Meta’s family of apps, underscoring its scale in advertising. Google Cloud growth: 100% last quarter - Used to argue Google’s AI and cloud businesses are compounding strongly. Alger history: 62 years - Used to support the firm’s experience in identifying disruptive growth opportunities. Dan Chung tenure at Alger: Since 1994 - He joined through Alger’s Analyst Training Program and has been there for decades. Market-cap example for Nebius: $50B+ - Described as having grown from small cap status to a large market-cap company.

Pivotal Quotes: "I think this generation of leaders clearly are proven leaders here." — Dan Chung: Chung explains why today’s AI leaders are different from the speculative companies that led the 1990s internet boom. "The market is clearly saying that Meta is not going to be an AI winner." — Ben Carlson: Used in a discussion about whether the market is correctly discounting Meta’s long-term AI prospects. "We do struggle a little bit with the controversy around it." — Dan Chung: Chung discusses Meta’s long-term risk profile and why Alger is cautious even though the stock looks inexpensive.

Implications: Investors should expect AI leadership to keep shifting and avoid treating every participant as a winner. The best opportunities may be in infrastructure and proven platforms, but concentrated bets require disciplined valuation, scenario analysis, and patience through volatility.