Sunday Snapshots (10th November, 2019)

Jim Simons and Renaissance Technologies, Ethical publication, Airbnb, and Krispy Kreme

Hey everyone,

Greetings from Chicago!

This has been a relaxing week as I transition from the hectic start of the quarter towards the holidays. Everyone is going into a low-gear and things are moving just a bit slower.

I’ve also been writing two long-form essays that I want to release before Thanksgiving:

  • How queueing theory explains Starbucks’ marketing campaigns

  • Understanding Flexport through the lens of growth loops

If you’re interested in either/both of these topics and would like to give me feedback, reply to this email and I’ll share a draft with you in the next few days.

In this issue of Snapshots, I want to discuss:

  • How Jim Simons created the best money-making machine in the history of investing

  • What nuclear weapons can teach us out about the ethics of publishing AI research

  • The merging of the Airbnb and hotel experience

  • Hustle and Krispy Kreme

  • And more!

Book of the week

This week, I read Gregory Zuckerman’s The Man Who Solved the Market on Jim Simons and Renaissance Technologies. Simons started out as a mathematician but after having a theorem named after him decided that there was an opportunity to apply his talents to the stock market. What followed was less about Simons mathematical prowess and more about his genius in creating a team that works. The Medallion Fund at Renaissance has had annualized gains of 71.8% over the last 20 years. That record is unbeaten at the scale of Medallion.

Here are my takeaways from the book:

  1. Emperors need empires: One of the counter-intuitive things I learned about investing from this book is that different strategies work at different scales. For example, strategies that work for a $10,000 personal portfolio will almost certainly not work for a $10,000,000 portfolio. When Renaissance started, Simon traded commodities, currencies, and bonds. This meant that he had a limited fund size – there is not as much money in these securities as equities. The firm was very good at trading these and their initial attempts at building a trading model for the equities market were unsuccessful. It seemed that the best way forward was to remain in their circle of competence and limit themselves to commodities, currencies, and bonds. But according to a staffer, “emperors need empires” and Jim Simons is an emperor. He saw equities as the silver bullet that could make him a billionaire. So he conquered it and created the best money-making machine in the history of investing.

  2. No insights needed: Here’s how Jim Simons is different from Warren Buffett.

    Buffett looks at a company, breaks it apart into its most granular levels, evaluates those granular levels, and decides if it is undervalued or not. If it is undervalued, he invests. If not, he doesn’t.

    Simons doesn’t care why the company is doing what it’s doing. He looks at trends in the market. For example, how do technology stocks move in the first 5 minutes after the Fed announces an increase in interest rates. How do they do after 10 minutes? 30 minutes? By analyzing historical patterns and finding micro-trends in different sectors and equities, Renaissance’s algorithms are able to predict what they will do in the short-term future. These micro-trends don’t need to be correct all the time. They just need to be correct in aggregate.

    Warren Buffett needs to be right most of the time. Jim Simons just needs to be right often enough.

  3. Team building: Recruiting was tough for Renaissance. Since they didn’t need to figure out why the market was moving, just how it was moving, Simons didn’t recruit finance or economics majors with a track record on Wall Street. He went into elite Math PhD departments and skimmed the top of the class. But most mathematicians are not driven by money. Academics also had to get over the fact that most of their achievements would not be public. In academia, the goal is to publish and when you publish, your name is at the top of the paper. At Renaissance, they would have to work as a part of a team.

    This was a tough pill to swallow for some potential recruits, but Simons motivated them by offering them a shot at modeling human behavior through the stock market. This was an ugly, messy bundle of noise. It was the mathematicians’ job to find the signals. The fact that they became multi-millionaires in the process was an added bonus.

The book is honestly not that well-written. Where you need brevity, such as the location of Renaissance offices, it is expansive. Where you want it to dig deep, such as Simmons motivations behind starting the Medallion fund and his psychological outlook at this system he was trying to crack, it under-delivers.

Simons’ story makes up for it though. From a mathematician to a code-breaker to solving the stock markets, he’s had an amazing life. The book is worth reading for that alone.

Long read of the week

Release Strategies and the Social Impacts of Language Models

When Robert Oppenheimer saw the first test of a nuclear bomb, he is said to have quoted the Bhagavad Gita:

I am become death, the destroyer of worlds.

At the risk of sounding like a luddite, progress in atomic physics has left humanity hostage to the decision-making of a handful of head of states around the world. While the use of advancements in nuclear energy could have been substantial, we have failed to capture those gains – nuclear power plants supply only 10% of the world energy. So we are left with a bad trade where we capture little of the upside and most of the downside.

We want to make sure that doesn’t happen with AI. We need to follow the nuclear model of not letting things fall into the wrong hands, but we also need to guide and improve how innovations in the field are applied.

Open AI is incorporating this approach into their research. Since their large scale unsupervised language-learning model can be applied to fake text generation across social media platforms, they chose a staged approach to publishing their results and their model. This is to evaluate how others are using their models and also to get people accustomed to text written by AI.

Innovation is important. But when it comes to potentially dangerous innovations, distribution is more important.

Business move of the week

Airbnb to Verify All Listings

Over the last decade, there has been an explosion of a new breed of companies. These companies are neither pure tech companies nor are they non-tech companies. They are somewhere in between. They are enabled by technology but interact with a part of the real world. Uber was the granddaddy of these hybrid companies.

Another darling child of this generation of companies is Airbnb. They have decided to verify all listings on their platform by December 15, 2020. The announcement comes after five people were killed in an Airbnb rental in Northern California. It’s the right thing to do and since Airbnb has practically zero marginal cost for an extra rental, they definitely have the cash to do this – CEO Brian Chesky cryptically says, “We have raised $3 billion and we have more than that in the bank.”

The announcement is interesting from a broader sense though. One of the original value proposition of Airbnb was lower prices. Today, prices on the platform are very close to what you can get at a hotel. Another proposition was a sense of whimsy of being in a home. Now, there are “super-hosts” who buy a house to host guests full-time – kind of like a hotel. The new verification process will certainly lead to a more uniform experience across different rentals. A uniform and safe experience across different locations is why people stay at hotels.

The endgame for Airbnb might be to become a mega-hotel company with a better-than-others technology layer on top of their supply of rooms. It would have also increased overall supply for stays. As the era of VC-subsidized consumerism draws down, I think we’ll see a merging of the new and old players to a point where they are not that different from each other.

Gesture of the week

College student Jayson Gonzalez would drive 540 miles round trip to buy Krispy Kreme and re-sell them to his peers in Minnesota. Despite the entrepreneurial hustle, he got banned by Krispy Kreme for not following food safety regulations. They eventually turned around though – they are now making him a distributor and will help him graduate college debt-free. Kudos to Krispy Kreme.

Meal of the week

Velvet Taco in River North is the best late-night food in Chicago. You can’t really go wrong with anything on the menu. Best consumed after 1AM.

That wraps up this week’s Sunday Snapshots. If you want to discuss any of the ideas mentioned above or have any books/papers/links you think would be interesting to share on a future edition of Sunday Snapshots, please reach out to me by replying to this email or sending me a direct message on Twitter at @sidharthajha.

Until next Sunday,