The Winners That Get Away | Chris Davis and Robert Hagstrom on Holding Great Businesses

Episode Summary

Executive Summary: Value investors Chris Davis and Robert Hagstrom discuss why investors often sell great businesses too early, how position sizing must balance compounding with client and volatility risks, and why high returns on capital and durable competitive advantages matter more than headline growth. They debate index concentration, AI infrastructure spending, and whether hyperscalers can turn today’s capital outlays into lasting pricing power and cash flows.

Main Topics: Holding great businesses and sizing positions: Davis argues that the most common mistake after finding a great business is selling it too soon. Both investors discuss position limits, noting that concentration can capture exceptional compounding but volatility, client behavior, regulation, and fiduciary responsibilities can make oversized holdings difficult to sustain. Index concentration and incentives: Davis warns that index funds have become increasingly concentrated and momentum-sensitive. He argues that continued flows into passive strategies may amplify price movements, while advisers’ focus on relative performance can conflict with investors’ absolute returns and risk. When and why to trim a winner: Hagstrom describes trimming when a position exceeds limits or its margin of safety declines. Davis’s pre-mortem approach asks what would need to happen for holding or selling to prove mistaken, while recognizing that business pivots and unexpected growth can make forecasts wrong. Returns on capital and valuation: The investors argue that earnings growth and valuation multiples alone do not capture a business’s value. Returns on invested and incremental capital determine how much reinvestment growth requires, whether cash can be distributed, and whether compounding can continue without repeated investor decisions. AI investment, capital cycles, and pricing power: Hagstrom sees the current hyperscaler build-out as potentially cyclical, with cash generation possibly improving after investment peaks. Davis emphasizes uncertainty over whether a few providers will retain pricing power or whether competition and low switching costs will transfer much of AI’s value to users. Competitive advantage, balance sheets, and financing: Davis highlights the difficulty of predicting how long high returns will persist and urges scrutiny of balance sheets and the links among financial statements. Both investors note that opaque or circular financing in AI deserves attention, while distinguishing complexity from evidence of wrongdoing.

Key Arguments: Once an investor has identified a genuinely strong business, selling too early can be more damaging than ordinary price volatility; major winners may compound far beyond initial expectations. Position size should reflect both opportunity and survivability. A concentrated holding can generate exceptional returns, but a severe drawdown may trigger client withdrawals, poor investor behavior, or regulatory and fiduciary complications. Indexing is a sound strategy in principle, but its growing concentration and momentum effects may make it riskier than investors assume; relative-performance incentives can also cause advisers to prioritize benchmark results over absolute outcomes. A rising share price alone is not a reason to sell or hold. Investors should reassess the margin of safety, business quality, competitive position, and what would have to be true for the market price to make sense. Two companies with similar earnings growth can deliver different investor returns because one may require substantial reinvestment while the other can distribute cash. High returns on incremental capital can justify a premium when they are durable. The AI infrastructure boom may not require hyperscalers to remain permanently capital-intensive. However, the investment case depends on whether providers can sustain pricing power and capture returns, rather than enabling users or customers to capture most of the value. Competitive-advantage duration is the central uncertainty: current leadership in chips, data centers, or AI models does not establish who will earn durable returns years later. Balance sheets and cash flows should be examined alongside income statements, especially where financing arrangements are complex; investors should determine who ultimately pays for AI services and infrastructure.

Data Points: Typical position limit discussed: Raised from 5% to roughly 10% - Davis describes increasing his limit; Hagstrom also says his firm adopted a 10% threshold, balancing concentration with client and portfolio risks. Naspers investment in Tencent: $32 million for 46%; later worth about $235 billion - Davis offers this as an example of the extraordinary value that can result from holding a successful investment rather than trimming it under a fixed position cap. Hagstrom portfolio technology exposure: About 45% previously; now about 26% pure technology, or 35% including Amazon and Google - Hagstrom describes reducing exposure as positions grew and margins of safety declined; he characterizes the broader measure as approximately market weight. Estimated hyperscaler investment-cycle peak: Second half of 2027 - Hagstrom cites Michael Mauboussin’s analysis as suggesting capital reinvestment could peak around then, with cash-flow yields and returns on capital potentially recovering afterward. Illustrative returns on capital: 50% versus 12% - Hagstrom uses these hypothetical returns to ask why businesses with otherwise similar growth should necessarily receive the same valuation multiple. Costco valuation and subsequent growth: Bought around $28; sold around $80; earnings multiple rose from roughly 15–16 to about 28 - Davis recounts selling Costco after its valuation expanded, then explains that store growth, Kirkland Signature, and international success exceeded expectations. Illustrative relative-versus-absolute return scenario: −1% annually versus a market at −2%, compared with 13% versus a market at 15%, over five years - Davis argues that advisers may prefer the first scenario because it beats the benchmark, although investors may prefer the second because it produces higher absolute returns.

Pivotal Quotes: "The biggest mistake you make is selling it." — Chris Davis: His answer to what most often goes wrong after an investor has found and bought a great business. "The big money is not in the buying and the selling, it's in the waiting." — Chris Davis, quoting Charlie Munger: The discussion of Naspers’ long-held Tencent investment and the difficulty of allowing winners to compound. "The key word, and this is the word we're the most focused on, is the word they." — Chris Davis: His framing of the AI investment question: whether infrastructure builders themselves will capture durable returns or whether value will accrue to users.

Implications: Investors should pair conviction with position-size discipline, evaluate reinvestment needs and competitive-advantage duration, and avoid treating recent index or AI trends as permanent. For AI, returns hinge not just on adoption, but on who ultimately captures pricing power and cash flows.