What this covers

Patrick interviews Gavin Baker, a growth and technology investor, in a conversation that runs from chip manufacturing to video games to the mechanics of online ad auctions. Baker's stated method is to notice what is already happening rather than predict the future, and he works through concrete cases. Intel lost its manufacturing lead in 2018 because TSMC reached its 7-nanometer process node first. Apple's used-iPhone resale value lowers the effective price of a premium phone. Game franchises like Call of Duty behave less like hits than like social networks that players find hard to leave.

The advertising material is the most specific. Baker argues that ad dollars lag consumer attention, as they did with radio and TV, and uses 2012 figures to project roughly $11 billion in US ad revenue for Facebook. He explains how second-price Vickrey auctions keep costs low for uncontested bidders until competitors arrive, and why that led him to short Netflix. Apple's hiring of a former Google search chief and the iPhone's role as an edge server for wearables also come up, along with why Chinese internet companies innovate faster.

The broader claims are more contested. Baker says AI progress comes from more data and cloud compute, not new algorithms, and that relative scale is the barrier to entry that matters most. He also argues that regulatory capture has made top-decile returns more persistent, that breaking up large tech firms may be the most logical regulation, and that most people will eventually spend most waking hours in virtual worlds. He draws on the Carlota Perez framework to expect decades more disruption. He closes on why value investing has struggled, citing career risk for human managers, and why tech offers an easier route to deep knowledge than other sectors.

Sharpest takeaway

Gavin Baker argues that growth and consumer-tech investing works best by underwriting durable business outcomes over 5-7 years, using simple, historically grounded frameworks (adoption lags, scale, data quantity, competitive intensity) rather than near-term estimate revisions, and that video games and Big Tech scale dynamics are underappreciated in public markets.

  • Long-horizon underwriting of business outcomes beats short-term estimate revisions, since valuation dominates only in the short run
  • Simple frameworks (attention share vs. ad share, data quantity drives AI quality, relative scale) provide conviction through volatility
  • Alpha is richest in tech and consumer because of uncertainty, dislocations, and cumulative knowledge advantages

The argument · threads4 threads · 87 claims
0.85

Competitive intensity in tech markets is the key variable investors must track closely.

6 pointscentrality 5/5
  • Chinese internet companies will likely never be regulated in the same way as U.S. internet companies, which makes them structurally different competitors.

    “I don't think they'll ever be regulated in the same way as US internet companies.”

  • Netflix was Baker's biggest career mistake because he never owned it in size, and Atreides was short Netflix despite the long-term bull case of 500 million subscribers at $10 per month and $60 billion in revenue, because competitive intensity and content uncertainty make the path hard.

    “Netflix, throughout my career, is by far my biggest mistake. I never owned it in size.”

  • Since mid-2016 forward-thinking DTC brands have reported that even a few physical stores significantly improve online advertising efficiency, which is the most important e-commerce metric alongside customer retention.

    “if we have a few stores, it actually dramatically impacts our online advertising efficiency”

  • Google and Facebook ad auctions behave like second-price Vickrey auctions, so uncompetitive segments enjoy low costs, but when new competitors enter, customer acquisition costs blow out in ways few models anticipate, making competitive intensity a key variable for tech investors.

    “These are both Vickrey auctions, they're second price auctions. So if you operate in a relatively uncompetitive segment, you have incredible advertising efficiency”

  • Baker avoids stocks with a very high degree of difficulty, and Netflix has become a meaning-of-life stock for him, where the long-term outcome is believable but execution difficulty is too high for a core position.

    “But now it's what I'd call 'a meaning of life stock' for now, where I do still believe in that long-term outcome, but the degree of difficulty is so hard. And I avoid really high degree of difficulty stocks.”

  • For technology investors, changes in competitive intensity are among the most important things to track, more so than in other sectors, because of the structure of ad auctions.

    “These are both Vickrey auctions, they're second price auctions. So if you operate in a relatively uncompetitive segment, you have incredible advertising efficiency”

0.61

Investors gain edge from private, cohort, and present-tense data rather than public forecasts.

4 pointscentrality 4/5
  • Regulation Fair Disclosure produced a sharp change in investing, after which companies became more circumspect about what they said to select investors, and alternative data's impact is comparable to or greater than that change.

    “In some ways, even bigger than Regulation Fair Disclosure (Reg FD).”

  • If customer cohorts are stable or improving, investors can gain reasonable conviction on how customers will behave two to three years out and model revenue more confidently than is typically possible with public equities, and cohort data can be constructed for public companies.

    “if you see the cohorts, you can get a reasonable degree of conviction looking out two-to-three years”

  • Private market activity, such as venture-stage company growth, influences public equities without being immediately observable, and an app that alerts on private companies growing over 200% can identify interesting venture opportunities.

    “The other dark matter is what's happening in private markets. And they have a big influence on public equity companies”

  • Investors should focus on noticing what is already happening in the present rather than predicting the future, a principle drawn from Benchmark's Matt Cohler and applied to conclude that the iPhone was a complement to, not a substitute for, console gaming.

    “My job is not to predict the future, it's simply to notice the present first.”

0.43

Interactive gaming keeps improving while non-interactive media has plateaued in quality.

3 pointscentrality 3/5
  • No movie made today will be better than Empire Strikes Back, Godfather II, or Deer Hunter, and no music better than Bach, Beethoven, or Mozart, implying that non-interactive media has reached a quality ceiling while interactive games keep improving.

    “you can't make a movie that is better than Empire Strikes Back, Godfather II, Deer Hunter.”

  • In 2012, the console cycle was the longest in history, so the gap between consoles and iPhone and PCs was at its smallest and the PS4 and Xbox One would deliver the biggest processing uplift ever, which led Baker to build an 8 to 10% stake in most public US video game companies.

    “So the new console would actually have the biggest uplift ever.”

  • Talented gamers could be a recruitment pipeline for businesses, since gaming ability correlates with leadership and pressure performance, though Baker had not thought of the idea.

    “it is a great recruitment pipeline, because it goes back to that movie long ago”

0.29

Passion and primary-document study build expertise beyond crowded conventional value wisdom.

3 pointscentrality 2/5
  • Value investing has been diminished by its popularity because everyone is steeped in Benjamin Graham and Buffett, so the chapters of The Intelligent Investor that matter most are widely known and the approach is crowded.

    “everybody's steeped in Benjamin Graham now. Everybody knows that chapter 8 and chapter 20 are the most important chapters”

  • Building top 1% knowledge in tech is realistic for someone who has focused on the sector for 20 years, and it involves reading primary documents such as the Google, Netflix, and Salesforce S-1 filings when they came out.

    “How many people read the Google S1 when it came out? The Netflix S1 when it came out? The Salesforce S1 when it came out?”

  • Passion for an area of the market is a strong predictor of becoming a good investor, because knowledge and hard work matter, and the top 1% of knowledge is hard to reach if the subject is not fun.

    “focus on what you are passionate about”