[Week 28, 2025] Blurred Lines
Welcome back to Price and Prejudice with a few musings from Week 28 of 2025.
Arbitrage Police
During my college years, I interned at an Indian hedge fund, and one aspect that surprised me most was the sheer volume of derivatives trading. What seemed like a niche market has since exploded into something remarkable: according to recent reports, India now accounts for over 60% of global equity derivatives trading volume, with retail investors driving much of this activity.
This article describes how Jane Street has been banned from Indian markets after regulators froze $566 million over alleged index manipulation. According to regulators, Jane Street would aggressively buy large amounts of stocks and futures early in the trading day, then place substantial bets that the index would decline. They would subsequently sell off their earlier positions, dragging the index lower and making their options trades profitable. In a textbook index arbitrage, you maintain balanced exposures – buying roughly as much stock as you sell in options to capture risk-free profits as prices converge. But Jane Street sold far more in options than they bought in stock, creating an asymmetric position that deviated from standard arbitrage practice. When you stray from the textbook definition of arbitrage, I suppose it becomes increasingly blurry what constitutes legitimate trading strategy versus (alleged) market manipulation.
Code Without Credentials
Whether you like it or not, one needs to think at least a bit about how to live with AI. (I like this take from Neal Stephenson who compares AI to animals) Of course, one important question is who wins and loses from the introduction of AI. This article talks about one particular kind of people that may benefit: "high-agency" individuals – those with curiosity and a defiant streak who challenge the status quo and believe the world around them is changeable. The piece argues that AI tools are democratizing not just information access but actual creation, allowing these driven individuals to build apps, systems, and businesses without traditional technical expertise. Examples include barbershops building booking systems, restaurants creating inventory management tools, and an 18-year-old building a calorie-counting app that's been downloaded six million times. As a personal case in point, here are some songs that I created using AI.
This notion of "high agency" feels particularly relevant to finance, where success has always depended on identifying opportunities and acting decisively when others hesitate. With AI, what once required teams of quantitative researchers and engineers can now be prototyped by individuals using AI-powered coding tools. If AI truly enables millions of high-agency individuals to build trading systems and financial tools, we might see both increased competition (driving down returns) and increased innovation (creating new sources of alpha). The question is then whether this democratization leads to more efficient price discovery or simply more noise and whether traditional financial institutions will adapt by hiring these AI-enabled individuals or find themselves disrupted by them entirely.
Internal Capital Markets
External financing can be expensive, time-consuming, and subject to market timing – factors that led many large corporations to develop internal capital markets as an alternative. These systems reflect the ability to allocate resources across business units based on internal assessments rather than external market conditions. GE famously operated this way for decades, moving cash from mature business (e.g. appliances) to high-growth divisions (e.g. finance). The theoretical benefit is obvious: management can direct capital to its highest-return opportunities without the friction and information asymmetries of external markets. In practice, however, conglomerates often trade at a discount to the sum of their parts, partly because investors worry about empire-building and misallocation of resources across unrelated businesses.
This article describes a modern reincarnation of internal capital markets through Musk's business empire, where SpaceX is investing $2 billion in xAI despite the companies being technically separate entities. The pattern is familiar – Musk previously borrowed $20 million from SpaceX to fund Tesla and took a $1 billion loan during his Twitter acquisition. While this cross-pollination allows rapid resource deployment without external financing constraints, it also creates valuation complexities. Increasingly, the valuation of each Musk company will reflect not just its standalone fundamentals but also its role as both a source and destination for capital within the broader ecosystem. Investors buying SpaceX or Tesla shares are effectively getting exposure to Musk's entire portfolio of ventures, whether they want it or not.