Programming note: The Diff will be off Monday for Memorial Day, back Tuesday.
- Ed Conway on the surprisingly interesting history of car paint. The crux of this piece is that early in the history of the automotive industry, most of the time that elapsed between when supplies entered a factory and when a finished car left was devoted to waiting for the paint to dry. Which meant that early car manufacturing was even more capital-intensive than the modern version since so much in-progress inventory had to just sit around (and actually finding somewhere to keep those cars can't have been fun, either). Sometimes, the biggest impact of a new technology isn't on cost of goods sold, operating expenses, or revenue—it's actually on cost of capital.
- Trung Phan dives into the Blumhouse model of creating cheap horror movies, some of which turn into hits. One highlight: "The most economical way to shoot a movie location-wise it to have it take place all inside a house." Constraints breed creativity!
- Abraham Thomas, who has written several excellent posts on data businesses, looks at how the industry will change with the advent of AI. One of the important points from this piece is that as AI tools start interacting with each other, it will become more important to keep track of what information comes directly from a person and what's the output of an algorithm (not that the median quality of human outputs is necessarily higher, but people do have a reputational investment in the claims they make, and LLMs don't). Also fun: the claim that the drop in data costs had long-term deflationary effects—if ads get better, you don't have to bake in such a large ad budget or retail markup to make a product worth selling—so "ZIRP was a zero-cost-of-data phenomenon!" Whether that's true long-term depends on whether information is more of a universal substitute or a universal complement.
- Nilay Patel interviews Microsoft CTO Kevin Scott on AI, and how Microsoft is handling it. (Disclosure: I'm long MSFT.) The piece has some useful comments on how large companies navigate politics and coordinate their research, and on where Microsoft sees itself in the emerging AI stack—which is partly using AI to enhance its own products and partly using those products as proofs-of-concept for other developers to use AI tools themselves. One interesting irony here is that the larger the overall economic opportunity for a new technology, the more it makes sense to let other companies capture market share, as long as there's some part of the stack that Microsoft still owns.
- Tyler Cowen on ChatGPT as a triumph of marketing as well as technology. Perhaps the most important datapoint to consider when predicting the pace of AI's rollout is that ChatGPT is by some measures the most rapidly-adopted product of all time, even though at the time that it launched it was just a better interface for GPT-3.5, which had been available in some form for over a year. Figuring out the right context to use a technology, even a revolutionary one, is a very big deal: when Bell Labs first developed the transistor, the top applications they had in mind were military radios and missile guidance systems; the NYT article announcing the discovery also suggested smaller hearing aids. There turned out to be other use cases, too.
- In this week's Capital Gains, we look at how pension's assumptions lead them to make counterintuitive decisions about risk, which means that $10 trillion in assets are being managed in a surprising way. You can sign up for Capital Gains and get a weekly explainer on finance, economics, and strategy here.
- Am I Being Too Subtle?: Straight Talk From a Business Rebel: Sam Zell died last week after a long and successful career of being the only willing buyer in a variety of different asset classes. Zell’s book is a pretty good tour of the deals he did, from early property management deals to a record-setting leveraged buyout. The blow-by-blow description of the latter deal, for Equity Office Properties, is a good look at how deal terms evolve as both parties get closer to a resolution (that deal gets described briefly from the other side in What it Takes). One source of frustration: we get a lot of details on when Zell bought assets nobody else wanted—but quite a bit less on how he knew they’d bounce back rather than going to zero.
- Money Machine: There are decent books about the careers of successful investors, and good case studies on deals gone wrong, but there’s an unfortunate shortage of in-depth books about single investments that did very well. Weijian Shan is singlehandedly solving this with his previous book Money Games and with his latest, Money Machine. The latter covers the successful investment in, and turnaround of, Shenzhen Development Bank. The book is better as a look at private equity in China than at private equity in banking (for one thing, SDB was quite insolvent at the time of the deal, but in a Chinese context this didn’t matter because the government more explicitly backstopped the bank). A lot of the value in this book is that dives into the sorts of negotiations that you can get only a hint of from following public news coverage while they’re happening.
- Drop in any links or comments of interest to Diff readers.
- What are some things that are widely assumed to be zero-rates phenomena, but that really aren’t? And what are some still-unrecognized zero-rates phenomena. (A good intuition pump for this is that there are some practices and ideas that last past their obsolescence date because some temporarily external factor keeps them alive—think of the US auto industry in the 70s, where the oil crisis was the immediate catalyst for their problems but wasn’t the underlying issue.)
A recent subscribers-only piece on tail-risk hedges ($) and the difficulty of using them got some great feedback; Taylor Pearson sent along this piece on how tail-risk hedging can improve long-term returns by reducing volatility drag. One of the paradoxes here is that in one sense, investors are irrationally afraid of risk and will accept lower returns in exchange for smaller drawdowns. But in another sense, they’re irrationally averse to some volatility-reducing strategies that actually improve long-term returns. Options trading wouldn’t be much of a business if the biases were uniform in all contexts, and one category of mistakes people make (including me) is evaluating standalone strategies instead of looking at how they perform in the context of an entire portfolio.
A Word From Our Sponsors
This newsletter is brought to you by Tegus, your modern research launchpad.
Find your next bold investing perspective.
Tegus simplifies your research workflow by uniting qualitative insights, quantitative data and public financials on one platform. With the largest transcript library of public and private market data, affordable expert calls – at a 70% cost savings compared to traditional expert networks – and actionable benchmarking, charting and comps tooling, you can move quickly on valuable insights to make smart decisions and seal the deal.
Become the expert in the room with Tegus and make your workflows work for you.
Companies in the Diff network are actively looking for talent. A sampling of current open roles:
- A successful crypto prop-trading firm is looking for new quantitative developers with experience building high-performance, scalable systems in C++. (Remote)
- A proprietary trading firm is seeking systematic-oriented traders with ML experience—ideally someone who has displayed excellence in DS and ML, like a Kaggle Master. (Montreal)
- A company building zero-knowledge proof-based tools to enable novel financial arrangements is looking for a senior engineer with a research bent. Ideal experience includes demonstrations of extraordinary coding and/or math ability. (NYC or San Diego preferred, remote also a possibility.)
- A well funded seed stage startup founded by former SpaceX engineers is building software tools for hardware engineering. They're looking for a UX/frontend engineer interested in designing and developing software collaboratively with satellite, rocket, and other complex machine engineers. (Los Angeles)
- A company building ML-powered tools to accelerate developer productivity is looking for a mathematician. (Washington DC area)
Even if you don't see an exact match for your skills and interests right now, we're happy to talk early so we can let you know if a good opportunity comes up.
If you’re at a company that's looking for talent, we should talk! Diff Jobs works with companies across fintech, hard tech, consumer software, enterprise software, and other areas—any company where finding unusually effective people is a top priority.