5 min read

Kohlberg, Kravis, Roblox

Kohlberg, Kravis, Roblox
Photo by Florian Schmetz / Unsplash

The Roblox Roll-Up

We were talking the other day about the bizarrely active market for mergers and acquisitions on Roblox, which raises the obvious question: Why? Why would a professional studio pay a teenager millions of dollars for a game that might be forgotten in six months? The answer is a classic bit of financial engineering, a move straight out of the 1980s private equity playbook. It is, in its way, a perfect roll-up.

Here is how a roll-up works in the real world. You find a fragmented industry full of small, independent businesses—say, dentist offices or car washes. These little businesses trade at low valuation multiples, maybe three or four times their annual earnings, because they are small and risky. You, the private equity firm, create a “platform company” and borrow a lot of debt to buy up dozens of these little businesses. You professionalize their operations, find some efficiencies, and then, a few years later, you sell your now-big, diversified, less-risky company for ten times earnings. It’s a form of financial alchemy: turning a 3x asset into a 10x asset just by bundling it.

Well, this is also what the studios are doing on Roblox. An individual game, created by one 19-year-old, is a high-risk asset. Its popularity is subject to the whims of millions of children; the creator might get bored and move on to something else. And so, it trades at a low multiple. Reports suggest a hit game can sell for as little as one or two months of its revenue.

But a studio like Voldex Entertainment or Do Big Studios is a platform company. It owns a portfolio of games, diversifying its risk. It has a professional team of programmers to sustain and improve the games after the original creator has cashed out. And crucially, it can secure and deploy high-value intellectual property that an independent creator can't.

Roblox itself is now formalizing this advantage, having just launched an official licensing platform to bring in partners like Netflix, Lionsgate, and Sega. Here is Reuters on the development:

[Roblox] has signed licensing partnerships with companies such as Netflix, which would allow developers to use characters and names from the streaming giant's popular shows "Stranger Things" and "Squid Game" in their games.

The License Manager would enable rights holders to register and create licenses, offering them quickly to creators in days or hours rather than months, Roblox said.

The ability to instantly and officially embed characters from Squid Game into your portfolio of games is how you professionalize the asset. It makes the studio a more stable, predictable business, and so, theoretically, it should command a higher valuation multiple. The arbitrage is simple: buy a dozen assets valued at a 1x multiple, bundle them together, and create an enterprise valued at a 10x multiple.

What is so delightful about all of this is the context. The financial logic is KKR. The vibe is a teenager’s bedroom. The deal-making happens on Discord. The sellers are sometimes literal minors. The assets are virtual worlds where you grow a garden or play a knock-off version of Grand Theft Auto. But the math is the same. It turns out the logic of a leveraged buyout works just as well for a digital paintball simulator as it does for a chain of laundromats.


The Compete Doctrine

One theory of the US-China tech war is that the US should do whatever it can to deny China access to the most advanced technologies, thereby containing its rise. Another theory is that this is a mistake, because it only forces China to build its own formidable, non-American technology stack, which the US cannot control or influence. In that second theory, it is better to let China buy some advanced US technology, to keep it dependent on the American ecosystem. For a long time, the first theory seemed to be winning. But a funny thing happened this week.

Nvidia said it received assurances from the Trump administration that it can once again sell its H20 artificial-intelligence chip in China, just days after CEO Jensen Huang met with President Trump. The move is a significant turnabout; the Commerce Department had just restricted sales of the chip in April, costing Nvidia billions of dollars in lost revenue and inventory write-offs. The news sent Nvidia’s shares up more than 4%.

What is going on here? One answer is that the second theory is gaining ground. Denying Nvidia a presence in China created a vacuum that was quickly being filled by a surprisingly capable Huawei. Allowing a calibrated flow of good-but-not-the-best American chips into China keeps Chinese developers hooked on Nvidia’s all-important CUDA software platform. It is a shift from a policy of pure containment to one of managed competition.

Another answer, of course, is that these chips are just a useful bargaining tool in a much bigger negotiation. A recent WSJ story on the decision makes the timing here feel particularly relevant:

The U.S. decision to allow more Nvidia chips to flow to China again was viewed in Beijing as a gesture of good faith in trade talks, said people close to official thinking. Access to chips and advanced technology has been a main priority for Chinese negotiators.

The two countries reached a trade truce in June that includes a Chinese commitment to step up approvals of rare-earth minerals needed by Western manufacturers. This week, China granted conditional approval to a $35 billion acquisition by U.S. chip-design software company Synopsys that had been held up by Beijing’s review since last year. 

Huang, Nvidia’s founder and CEO, long preferred to stay out of politics but has emerged as a central player in U.S.-China tensions in recent months, hopping from Beijing to Washington and back in the hopes of maximizing Nvidia’s access to China and global markets.

This is not a permanent peace treaty in the chip war. The politics in Washington remain fraught, with influential senators scrutinizing Huang’s trips to Beijing and a major deal to sell chips to the UAE still stalled over concerns they might be diverted to China. Nvidia has spent years designing chips to meet US rules, only for the “goal posts to be moved repeatedly.” What the H20 reversal suggests is that the game is now more fluid. The goal posts can be moved back, too, if the price is right.


The Scoreboard

  • Semiconductor: Nvidia to Boost H20 Chip Sales to China After US Export Restrictions Ease (Reuters)
  • Semiconductor: TSMC Posts Record Quarterly Profit on AI Demand, but Wary About Tariffs (Reuters)
  • Cloud Infra: OpenAI Says It Will Use Google’s Cloud for ChatGPT (CNBC)
  • AI: Huawei Brushes Aside Accusations of Copying Alibaba AI Model (Nikkei Asia)

Enjoying these insights? If this was forwarded to you, subscribe to ARPU and never miss out on the forces driving tech:

Subscribe