This interview argues that we're in a tech revolution as big as the Industrial Revolution, and investors should focus on companies at the edge of disruption with lasting advantages. Fund manager Gavin Baker is most bullish on video games—they've evolved from entertainment into social platforms, but the market still undervalues their money-making potential. Key holdings: Apple (used iPhones hold value better, so users pay less than the sticker price suggests); Activision (Call of Duty has been top-selling for 20 years, with loyal users); Fortnite (called the first real example of the metaverse). Baker notes games earn $0.08-0.50 per hour vs. cable TV's $0.85-0.90, expecting that gap to close.
Gavin Baker (Founder of Atreides Management) explores core themes in technology and consumer growth investing on this episode of Invest Like the Best. Key insights include: Apple as a growth investment, with its future key lever being services revenue (e.g., App Store, Apple Music) rather than hardw
Gavin Baker (Founder and Portfolio Manager of Atreides Management) focuses on growth investing in technology and consumer sectors. The core theme of this issue: The current era is undergoing a technological disruption comparable to the Industrial Revolution, and investors should seek high-ROIC companies at the forefront of disruption with sustainable competitive advantages. Baker’s most compelling thesis: Video games are evolving from a "form of entertainment" into "social venues" and "metaverse platforms," with their user engagement and monetization potential significantly underestimated by the market — the global video game population has grown more than 10x over the past 10-15 years, yet less than one-third of the population participates, while game monetization rates ($0.08–$0.50 per hour) remain far below those of cable TV ($0.85–$0.90), leaving substantial room for convergence.
Baker argues that the current pace and scale of technological change rival the Industrial Revolution, and we may only be entering its peak phase.
> "I try to invest at the edge of disruption, just before it."
Baker believes that Apple, as a growth stock, derives its future value from three key variables rather than hardware sales alone.
1. Pricing power of used iPhones: Used iPhones retain value far better than Android phones. When users trade in their devices, an iPhone can fetch approximately $400, while an Android phone only yields about $150. This means that although Apple lists the iPhone at $1,000, the actual out-of-pocket cost for the user is only $600, whereas an Android user must pay $850. Apple has achieved a competitive strategy of "charging consumers a higher price, yet delivering a lower effective price."
2. The iPhone as an "edge server" for wearables: Drawing on the client-server architecture of the 1990s, Baker argues that in the future, everyone will own a "constellation of wearables"—phones, AirPods, AR glasses. The iPhone will serve as the edge server for these devices, and the "attachment rate" of wearables is a key variable for Apple's future growth.
3. Search competition potential: In 2018, Apple poached John Giandria, the head of search, from Google. Baker believes this signals a "tectonic shift" in Apple's privacy policy. Apple and Facebook are the only two companies with differentiated datasets that could potentially challenge Google's dominance in search. If Apple successfully enters the search market, it would unlock vast revenue opportunities.
Reader's Note: Baker holds a position in Apple. The above analysis reflects a long-side perspective, and its optimistic bias should be noted.
This is the most core and distinctive judgment of this issue. Baker argues that video games have evolved from "games" into "social venues," and their user engagement and monetization potential are systematically undervalued by the market.
| Metric | Value |
|---|---|
| Average mobile game monetization rate | $0.08–$0.10/hour |
| Average console/PC game monetization rate | $0.50/hour |
| Average cable TV channel monetization rate | $0.85–$0.90/hour |
| High-engagement content (e.g., sports) monetization rate | $2–$3/hour |
Baker believes that video game monetization rates should at least converge toward those of cable TV. Given that game users exhibit far higher engagement than average TV viewers (similar to passionate sports fans), the room for convergence is enormous. Moreover, the above figures exclude esports — people under 25 now spend more time watching esports than traditional sports.
1. "Games are a cyclical industry": In the past, game stocks fluctuated around console cycles, but today games have become social networks (e.g., Call of Duty has been the best-selling game for 20 consecutive years, interrupted only by "rock star" titles like GTA and Red Dead Redemption). User stickiness is extremely high, and cyclicality has been greatly reduced.
2. "The iPhone is a substitute for game consoles": Baker refutes this from first principles — playing Call of Duty on a 4-inch screen with fingers (an imprecise input method) versus a 50-inch screen with surround sound and friends online is not a substitution but a complement. In 2012, when the market sold off game stocks out of fear of iPhone substitution, Baker bought against the tide, holding 8%–10% of most publicly listed US game companies.
3. "Games are a hit-driven business": In reality, top game IPs (Call of Duty, Assassin's Creed, Battlefield) have extremely stable sales performance because they are essentially social networks — player communities find it difficult to migrate collectively.
Baker argues that most people will spend the majority of their waking hours in virtual worlds within their lifetimes. This is a linear extrapolation of current trends (mobile phone usage alone is already staggering). The platform layer of the metaverse will be dominated by existing top-tier video game IPs — want to visit ancient Rome? Enter Assassin's Creed; want to experience a soldier fantasy? Enter Call of Duty; want casual social interaction? Enter Fortnite. Fortnite has already been called "the first true instance of the metaverse" by Matthew Ball.
> "It is inevitable that a majority of people will spend a majority of their waking hours in virtual worlds within our lifetimes." (Baker believes it is inevitable that most people will spend the majority of their waking hours in virtual worlds within their lifetimes.)
Baker argues that the structural challenges facing value investing are deeper than the market recognizes, with the core issue being the weakening of the "mean reversion" mechanism.
1. Regulatory Capture Intensified: In the modern U.S. economy, the degree of regulatory capture has reached its highest level in the past 70–80 years. Companies hire former congressional chiefs of staff and former regulators to build teams of regulatory experts, and many important laws are actually drafted by corporations and their lobbyists. This naturally entrenches the advantages of incumbents.
2. Scale Advantages Matter More Online Than Offline: Baker believes that if forced to choose just one competitive moat, it would be scale, especially relative scale. Offline, a 40% market share against two 30% competitors is not ideal; but online, data scale directly determines AI quality—multiple papers from Google and Microsoft Research confirm that for every order-of-magnitude increase in training data, AI quality (prediction accuracy) doubles. This creates a powerful positive feedback loop: most users → most data → best AI → stronger competitiveness.
Baker believes that the most reasonable regulatory approach is a breakup—these companies suffer from a "conglomerate discount," and a breakup could lead to more optimal capital allocation across businesses (e.g., an independent Google Search might increase its buyback rate). However, excessive regulation would severely harm U.S. interests, as Chinese internet companies would then dominate the global market.
Baker admits that Netflix is the biggest mistake of his career (he never held a significant position), but at Atreides, he shorted Netflix. The logic:
Reader's Note: Baker's current stance on Netflix is "wait and see" (meaning of livestock). He believes the long-term outcome is still possible but extremely difficult to achieve, and he tends to avoid high-difficulty stocks.
Baker argues that achieving top 1% knowledge in the technology sector is easier than in traditional industries such as banking and insurance, because the 2000 dot-com bubble and the 2008 financial crisis washed out a large number of investors, leaving very few who have consistently focused on technology for 20 years. Technology is a "cumulative knowledge game" — reading the S-1 prospectuses of Google, Netflix, and Salesforce, and continuously tracking them, allows knowledge to compound.
Baker quotes Benchmark's Matt Kohler: "My job is not to predict the future, just to notice the present first." Two cases:
1. 2012 Gaming Stocks: The market feared that the iPhone would replace console gaming. Baker judged from first principles that this was a complement rather than a substitute, and went heavily against the consensus.
2. 2017 Retail Stocks: Starting in 2016, DTC brands reported that "opening a few stores significantly improves online advertising efficiency." In 2017, Amazon acquired Whole Foods. Baker believed this was a complete validation of physical retail value by the world's most e-commerce-savvy company, while the market knocked retail stocks down 10%-20% on the day. He screened retail stocks based on: exclusive supply (not available on Amazon), a high proportion of cash payments (not competing with Amazon Prime), and a certain degree of digital maturity.
| Position | Guest Stance | Key Data |
|---|---|---|
| Apple | Bullish (holds position) | Used iPhone trade-in value ~$400 vs Android $150; hired Google search head John Giandria in 2018 |
| Intel | Risk warning | TSMC's 7nm node in 2018 led Intel's 10nm node, causing Intel to lose 15-20% of its embedded advantage, with a 20-30% swing in competitiveness |
| Bullish (previously heavy position) | In 2012, accounted for 7% of US user time, <4% of ad revenue; Baker model projected $35B revenue (based on time share convergence + search non-advertising + measurability premium + CAC substitution rent) | |
| Netflix | Neutral (previously short, now on the sidelines) | Content has a 400% markup from creator to consumer; customer acquisition costs will surge as competition intensifies; pricing power uncertain |
| Risk warning (regulation) | Search is "demand fulfillment" rather than advertising, analogous to a digital "shelf fee"; capital allocation could improve after a breakup | |
| Amazon | Risk warning (regulation) | Whole Foods acquisition validated physical retail value |
| Tencent | Neutral (no position, but researched) | Pioneer of China's internet super-app; Baker visits China at least once a year for research |
| Activision/EA/Ubisoft | Bullish (previously held 8-10% position) | Call of Duty has been the best-selling annual title for 20 consecutive years; gaming has become a social network, not hit-driven |
| Fortnite (Epic Games) | Bullish (unlisted) | Considered "the first true instance of the metaverse"; is the third user entry point in the gaming industry (after Nintendo and iPhone) |
1. “We may be entering the peak of technological disruption, with another 30-40 years of rapid disruption ahead” (Baker) — Using the Industrial Revolution as an analogy, the microchip (1959) corresponds to the steam engine (1761), with the peak of change occurring 70-100 years later.
2. “Video game monetization rates should at least converge with those of cable TV” (Baker) — Current games monetize at $0.08–$0.50 per hour, cable TV at $0.85–$0.90, and high-engagement content at $2–$3. Game user engagement far exceeds that of average TV viewers.
3. “Apple can sell at a higher price than competitors, but consumers actually pay less” (Baker) — The high resale value of used iPhones means Apple’s sticker price of $1,000 results in an actual consumer outlay of $600, while an Android priced at $850 costs the consumer $850.
4. “Every order-of-magnitude increase in data doubles AI quality” (Baker) — A conclusion from Google and Microsoft Research papers. This creates a positive feedback loop of “most users → most data → best AI,” which is the deepest moat for large tech companies.
5. “My job is not to predict the future, but simply to notice the present first” (Baker, quoting Matt Kohler) — Two cases: gaming stocks in 2012 and retail stocks in 2017. The market mispriced them out of fear, and Baker identified the truth of the “present” from first principles.
6. “Most people will spend the majority of their waking hours in virtual worlds within their lifetimes” (Baker) — This is a linear extrapolation of current trends, and existing top gaming IPs will become the platform layer of the metaverse.
7. “Reaching the top 1% of knowledge in tech is easier than in banking/insurance” (Baker) — Because the 2000 and 2008 downturns washed out many investors, and few have remained focused for 20 years. Tech is a “cumulative knowledge game.”
8. “Regulatory capture is at its highest level in the past 70-80 years, with many important laws drafted by corporations and their lobbyists” (Baker) — This is one structural reason why the mean-reversion mechanism of value investing has been weakened.