Find the audio here.
When I was 22, I thought I hit the jackpot with my job. After bonus season, management took the entire firm to Las Vegas. It’s still a weekend that looms large in my memory, not only because I got sun poisoning but also due to the sheer over-the-topness of it all. There was one moment when I was sitting in the basement of Drai’s at a table with the older traders and one of their friends — the friend was Vin Diesel. I mean, this was Vin pre-Fast and Furious, but still, I had never been to Las Vegas before, and my first impression was pretty overwhelming.
It was also peak.
What I didn’t understand was that all of this was a one-time dividend.
I’m going to talk about what I didn’t understand as a naive 22-year-old that I think is a universal lesson especially relevant today. I’m going to start with a simple fact. Prop trading firms pay 10x for new grads compared to when I graduated 23 years ago. Do the math. The natural log of 10 is 2.3. Divided by 23 years, that’s 10 percent compounded growth for a generation. That’s at least three times more than the inflation rate over the same period. The reason is a clue to the lesson.
Let’s consider what a more astute observer than my starry-eyed self would have seen 23 years ago. The option trading and market-making worlds were fragmented industries, lots of mom and pops, a crowded trading floor bursting with mosh pit energy. Just as connectivity was ramping up in the late 90s. The run-up to the 2001 dot-com bust was a moment of severe over-earning — a ‘tween moment where there was a boom in trading volumes and speculation to gorge the mom and pops one last time while ringing the dinner bell so loudly that it got the attention of all the suits.
The banks would commit resources to “facilitate” their customers rather than just hand them off to the floors for the equivalent of a waiter’s tip, when they could get paid for the whole meal. Specialist firm, Spear Leeds, and Kellogg top-ticked the business when it sold itself to Goldman for $6.5B, who felt the need to panic buy into the market-making game and evidently extrapolated just as poorly as this naive 22-year-old.
Our astute observer would have understood back then that a common bookie could make a great living skimming 10 percent on NFL games. Yet in this slice of time, traders were able to sell call spreads worth 50 cents for 75 cents on single-list option venues. This was not going to last. Either the customers would all go broke, or everyone would notice the free money.
So intelligent firms, knowing the margins were excessive, optimized for market share. They could undercut the mom and pops, offering prices that presented them with a worse risk-reward, although still quite profitable. But the undercapitalized members of the fragmented ecosystem would eventually give up. The surviving firms would increase market share, which meant better looks, which meant better info, which meant more profits, even if the margins were lower. Plow the profits back into technology capex, and you have a flywheel that is not only familiar today but probably over-internalized by Silicon Valley, by the VCs who spent the better part of a decade speed running this same cycle by subsidizing portfolio companies’ let’s “sell a dollar for 95 cents” customer acquisition strategy.
So, why is pay up 10x in a generation? Well, technology was leverage.
The headcount of the consolidated industry has likely not changed much since 2000, but the aggregate profits are much higher. Leverage means to do more with less. This is the textbook example. Instead of a fragmented industry, a couple of firms have over three-quarters of the option market-making market share today. In this process, the skill set to work at these firms has also converged with tech. This boosts prop pay further. Since the firms must hire from the same talent pool as the largest companies in the world, the victors of the software and therefore leverage-eating-the-world Olympics.
What does this mean for today’s 22-year-old or for that matter any ambitious person at any age, operating in a world where the waterline of automation keeps rising, submerging the obsolete and leaving only the tallest, still growing technology obelisks. Well, I think it urges awareness. You see, I was completely unaware. I didn’t realize that the business’s outsized profitability back then was an anomaly, and I just happened to walk into it. I was a warm body they needed to hire amidst the cash-grabbing scrum. I was merely a bubble hire. I mean, I’m a 95th percentile schoolboy. I’m not a bag of rocks, but I’m a dime a dozen when a firm only needs to hire 50 to 100 people a year.
Let me be more specific about what I mean about being aware.
Be aware of the background process of how technology works. It commodifies and flattens all tasks in a system that are adjacent to a differentiated bit. That differentiated bit is where all the value accrues; it’s the artist who can eventually cut out the middleman to go directly to the fans.
Option author and trader Euan Sinclair talk about how in the ’80s, you could make money by doing the tedious task of compiling historical prices. The edge wasn’t in the Black-Scholes model, but in the labor. But as automation cheapened and democratized the labor; the source of the edge moved higher up the stack. You could say it has moved all the way up to the bottleneck. Today, the direct or quasi-direct relationship to the customers via the exchange. Market segmentation is the biggest source of edge, and that has some part to do with politics and the provisioning of access. In other words, dynamics that are not strictly dependent on technology.
Again, technology is leveraged, doing more with less. It’s the definition of productivity growth. This is a good thing. It’s the foundation of real, not just nominal, economic growth per capita. But its benefits are not even.
[This reminds me, by the way, of Burn Hobart’s piece about tradeable goods. He points to this chart that conservatives and libertarians alike like to parade around, showing that inflation in healthcare, housing, and education — the three sectors characterized by government regulation — have outpaced everything else by a large margin. But what’s overlooked is that these are also the sectors where the benefits of scalability have lower limits. Low student-to-teacher ratios are selling points that would literally be undermined by automation. If you read about the economic concept known as Baumol’s cost disease, then your baseline expectation is that service sector industries should show lower productivity gains on a relative basis to the tradable goods sector. The chart might still have merit, but it’s almost certainly weaponizing a false prior that misunderstands the substantial impact of relative productivity on long-term inflation rates.]
All right, back to where we were — the prescription for an individual here is to differentiate, it’s to be unique. My friend, an author, trader, and entrepreneur, Augustin Lebron, wrote the following about LLMs like ChatGPT, I’m quoting here:
The LLM is a put option on cognitive tasks. Almost a year into the world’s experience with ChatGPT, it’s pretty clear that its biggest value is helping people with things they are not good at. It’s not going to write the next great American novel, but it will write that annoying blog post you need to churn out if you’re not great at data analysis. A code interpreter will do a fine job. It won’t make you an edge-level researcher, but it can help you learn faster. Basically, it makes you decent at what you’re not good at. That’s the definition of a put option. It gives you a cap on the amount you’re going to lose relative to a competitive world in situations where you’re not that good. Using an LLM is like cheaply buying puts on the holes in your cognitive skill set.
So, how should you rationally adjust your behavior given that you own these cheap puts? Embrace variance. Your value to the world is less dependent on what you’re worst at, but more dependent on what you’re good at.
That’s a very profound statement that he’s making, and I totally agree with him. I responded:
If you subscribe to lean into your strength and just spend the minimum amount to get your weaknesses serviceable, the LLMs have just raised the strike for the same given cost.
This is permission to go deep.
I do want to anticipate an obvious rebuttal, best embodied by David Epstein’s book “Range.” The subtitle of the book captures the gist of it: “Why Generalists Triumph in a Specialized World.” Epstein offers many compelling arguments. But he’s not arguing for what bullshit job apologists want to interpret it as — a celebration of being a mile wide but an inch thick. The people who do this, see a renaissance man in the mirror, but is really more of a superficial good-at-trivia nuisance, not a true polymath.
A useful way to bridge the going deep versus rangy person is to consider some advice from Marc Andreessen’s old blog. (His old blog was very good, as opposed to the marketing hack jobs I feel like he’s given to these days). Anyway, he talks about getting to the 90th percentile in two fields as an alternative and often easier path than becoming the 99th percentile in one field.
[In joint probability space, this is just as rare. Since 10 percent times 10 percent still leaves you in the top one percent. But this assumes the two fields are totally independent, which is unlikely if the fields are both valued by the market. However, I do think that this sentiment holds here.]
So, being great, but not elite in several domains, is an alternative to striving to be in the top one percent of a single field. I think this reconciles the range idea with my sense that you need to go deep. Your creativity, what makes you unique, is going to be your edge in the world. And in general, I think the best long-term strategy for both prosperity and satisfaction is to lean into your creativity. What makes you different is your strength.
Well anyway, that’s at least until our insides are refined into robot fuel.
Comments