The Wooden Nickel is a collection of roughly a handful of recent topics that have caught our attention. Here you’ll find current, open-ended thoughts. We wish to use this piece as a way to think out loud in public rather than make formal proclamations or projections.
1. MoviePass vs Netflix
MoviePass was born in the early 2010s with the objective of becoming the “Netflix” of the movie-going experience. For a monthly subscription price, consumers could visit almost any movie theater in the country once a day to watch a film. With a lack of interest, management cut the initial subscription price of $30-$50 per month (depending on the city) to a nationwide price of $9.95. Subscriptions surged.
The business remains one of the worst business models in recent history, a pure exercise in stupidity, laziness, and chasing any recent fad, and in this case, the idea that everything must be sold as a subscription. An entry-level microeconomics student could tell you why: MoviePass had zero control over its costs, made no ancillary revenue (eliminating the well-known razor-blade business model) through selling concessions or other goods and services, and didn’t offer a new value proposition. With the average price of a ticket just north of $9, its hope to survive as a going concern was essentially that subscribers would sign up for the business and never use it, as the company barely broke even with one visit and was deep in the red with two or more visits to a theater in a month.
Netflix, by contrast, carved out a new market by recognizing a subset of entertainment consumers whose needs were not optimally met. First, the company offered something new, a no-ads experience of watching television. Second, there was a cost-effectiveness to the subset of consumers frustrated by a cable bundle; the latter’s rising price due to the high fees charged by sports and news programming, which a growing portion of the population did not find worth paying for. These two factors represented a different economic value to consumers compared to the MoviePass experiment. Finally, unlike MoviePass, Netflix had a stable cost structure by paying upfront for content. As a result, each incremental subscriber was almost pure profit, and each incremental view of the same program did not increase its marginal cost.
The contrast is about as clear as it gets for why doing something differently does not make one a “disruptor.”
2. Doing Nothing is an Option
While we’re on the topic of Netflix, the biggest story in corporate news is the media disruptor’s agreement to acquire Warner Bros., home of HBO. Zooming out ~30 years or so, it’s quite the path for HBO. The cable network redefined consumers’ expectations of the cable industry specifically and entertainment at large through its selling point of premiumization, an ad-free experience, and innovative digital content. Once a disruptor itself, it’s now being gobbled up by another.
The journey for these has been an interesting one (going back to Warner’s acquisition by Discovery in 2022) and is not near being finished, given a couple of different news stories from this weekend and yesterday. And the numerous scenarios have investment analysts and bankers playing game theory out in Excel for what Warner Bros should be trading at, given:
- Part of the proposed acquisition price comes in the form of Netflix shares.
- The deal, if it goes through as structured, will take time, and thus the future economic value needs to be discounted by some amount.
- Further, Netflix’s offer is only for part of the assets. The residual (aka stub) carries an implicit value in any current trading price. With shares at $27 and the deal value pegged at just over this level, shares are either saying the remaining value of the business is almost zero (unlikely) or expect other bids, which leads to…
- Paramount (home to Mission Impossible, Yellowstone, and Top Gun) has promised to go hostile with its original bid, while hinting it may go higher than the current offer of $30.
- Netflix may respond with a higher bid as well.
- The regulatory risk is real as members of Congress have spoken out against the deal, and the related party transaction nature of this administration continues.
- If offers don’t change and Paramount’s hostile bid fails, the regulatory risk offers serious downside to equity value, but…
- Warner Bros. gets to walk away with $5B in cash to pay down a material debt pile.
The permutations are numerous and far exceed the list above. In reading research and analyses about the situation this past weekend, one scenario caught my eye. Specifically, one bulge bracket sell-side firm was promoting the idea that the regulatory risk was so substantial that Paramount could end up lowering their offer. If the Netflix bid gets squashed for competitive reasons and without another buyer offering adequate value, this analyst argued Paramount could lower their offer to $27 or $28 per share due to having gained bargaining power.
It’s a scenario that, while improbable, can’t be eliminated, but it does have a dangerous underlying assumption that I think is essential for investors to always have in their mind: you are not obligated to do anything. There is no obligation or entitlement for Warner Bros. to sell. If they don’t get what is deemed fair value for their assets, they can continue to operate as is; they do not have to sell.
Similarly, investors do not have to buy (but often do…hence the FOMO trade we see in markets) just because the market is going up or because someone recommends it. You are not obligated to have an opinion on AI, let alone invest in it, if you don’t have a view. You do not have to be in nuclear energy or green energy, or in defense companies, or in private credit, etc. It would take no less than an hour to find a superficial justification for any and every security on the planet. If that’s all it took to invest, you’d end up with a portfolio of tens of thousands of securities and have duplicated the global market portfolio with much higher risk and at much higher cost.
You are free and entitled to invest how you wish, but in no way, shape, or form are you obligated to oblige the cacophony. It’s best to remember what the wise man does in the beginning, the fool does in the end. You do not have to swing at every pitch.
3. Gemini and other AI Thoughts/Questions
- The most important element in Gemini’s success is not how good the model and applications embedded with it are, nor the fact that it was trained on customer chips (prior Gemini models were as well..something people keep forgetting and overlooking). Rather the most important (near term) points are that: a) the model’s improvements were predicated on pre-training scale (i.e. adding more computing resources and data…a methodology that was supposed to be dead 12 months ago) and b) that Google chose to price the use of it’s models to third-parties at a premium to OpenAI’s GPT family of models, a new development. The former is interesting as it brings up other relevant questions. Does Google have some sort of secret sauce that other labs have failed to discover? Or does this accelerate the race for frontier leadership since adding compute is a tried and true method that other labs have experience with? If the race is back on, then will the diverging price performance of the Google ecosystem and the OpenAI ecosystem cross again?
As for the latter, does this mean Google believes its model is at such a premium that others will pay up for it? Does it believe its lead is sustainable?
- As models get bigger and better, small models get better as well. What happens when models small enough to fit on a consumer device (a laptop or a smartphone) no longer need access to data center computing power? Is this how Apple, with its innovative Unified Memory, wins in the AI era?
- If leads from having a frontier AI model only last 6 months (or much less), then where does value accrue? Is it all consumer surplus? Does it flow up a layer to applications which can route requests for intelligent tokens to the best or cheapest model?
- What does xAI do in the coming years? Very little information is available in detail, but by most accounts, it reimagined how to lay out a data center in a more efficient manner, giving it a cost advantage in generating tokens. But without the scale of a user base anywhere close to OpenAI or Google, how long can it afford to be on the bleeding edge of LLM development?
- As usage of LLMs grows (both in the number of users and in frequency/intensity of tasks), what happens to data walls? Will consumers naturally crave AI systems and agents that work across applications? Or will they accept AI as a natural feature across current use cases? Said better, will they want Gemini/ChatGPT/others to be able to read their Outlook emails, cross-reference articles in the WSJ, and explore pass-protected Substacks to have a single point of intelligence? Or will they accept the current boundaries the web holds and utilize a web of disjointed systems?
- Is Blue Owl’s (albeit failed) merger of its public and private vehicles the canary in the coal mine of this asset boom? If they are the incremental financier of the AI capex story, then who steps in should the private credit asset class fail? What if no one steps in?
4. Recommended Reads and Listens
