Interview: Byrne Hobart
Byrne is, simply put, one of the best writers out there today. His insights:words ratio is off the charts, especially when the words themselves clock up easily north of 0.5m a year.
Finance often attracts those attracted to figuring out the whole world, or at least de-dimensionalise it to something tractable. And Byrne manages to make this unconscionably fun by re-introducing all possible missing dimensions; and using it as the wedge to explore philosophy, religion, beauty pageants, lobster pots, whaling, blimps and much much more!
Which is why I wanted to interview him, because there's so much to dig deeper that it's pure nerdy fun and about as strange loopy a thing as I could think of. The result is below. Byrne’s on twitter and if you don't already subscribe to his newsletter, tch tch, c’mon!
First question. Why hasn't there been a truly great literary work on the world of finance? There're the odd ones that are really popular like Bonfire or Liar’s Poker. But considering the centrality of the finance world, I find it odd that there haven't been truly great novels around that. Why do you think that is?
Yeah. It's a good question. So there are, there are writers who have written about financial markets and about working in finance. So I haven't read his novels, but there's a writer who I think may be the most frustrated novelist in the world, because he worked in finance for a while, he quit, he wrote two novels, they are about finance and, you know, working at hedge funds and investment banks and things. And they have a handful of reviews on Amazon. And it's a really wonderfully written book. And it's really enjoyable, well crafted, etc. So he knows he's a good writer, and he's done the writing thing. But then he went back to work and he wrote a different book, his most successful book, a nonfiction book on fracking. It's called The Green And The Black. And it has more Amazon reviews than the novels!
If you're at all curious about energy issues, and fracking, and how people who are involved in those companies and fund them how they think about it, then it’s definitely, definitely worth reading. And I do try to keep an eye out for a book that is, you know, the great American Finance novel, or the great financial sector novel, because I guess finance is pretty global at this point. There's a contender that I liked a lot. It's a book called Kapitoil. It's a book about working in that sector and dealing with some of the moral issues that come up. I thought it was well done.
There's some older ones, there's this book called JR by William Gaddis that is very complicated and very hard to read. And kind of surreal, like the book is written almost entirely in dialogue. And the speakers are never identified except through context. But that one, it's older, it's from the 70s. And it's kind of thinking through this very 1970s sort of way of doing business with conglomerates that just got into a whole lot of different random enterprises. But it is, it's certainly literary, certainly worth trying to slog your way through.
But yeah, I, I don't know for sure why, why there isn't a great finance novel, it may be that the people who have picked up enough information and context really understand it, they all signed non disclosure agreements, and they're all getting paid well. Like, you know, if you're trying to do a finance story, you can do a story that is a human interest novel about people who work in the lower levels of the industry. But really, what I think a lot of people would be looking for is what it's like at the top and how those people think. But the issue is, if you're close enough to those people, a) non disclosures as alluded to and b) you're probably also making a lot of money. So you have to really, really, really want to write that novel. You're sort of, you know, you have to befriend these people and then put them into a book, which is a pretty serious violation of trust. So yeah, I think I think there's just a number of forces arrayed against a really good financial novel ever coming out. That said it could happen, and I'm still looking for it.
Yeah, Kapitoil is one I remember. It was by Teddy something if I recall [Teddy Wayne]. What I found intriguing about that novel is that like almost all depictions of the world of finance and popular culture, it deals with the moral questions. Relatively clearly, right? I mean, that profit gained through oil futures prediction and things of that nature. And there's a bunch of discussions there about how to get out of this narrow minded mentality, so to speak, by reading, you know, Steinbeck.
I find that to be also representative of some of the broader conceptions of finance, right? I mean, if you think about movies, you think about, you know, something like The Big Short that came recently. Or go back to Gordon Gekkos time, almost all of it is dealing with that one aspect, which is greed, almost none of it with the other aspects, which is that, ultimately, it is an industry that sits at the crux of pretty much every other industry, which means you can actually have a pretty interesting worldview within it.
Yeah, you could say that maybe this is a stretch, but there is this notion that Berkshire is Warren Buffett’s sort of work of art. It's his life's work. And it is something that is, I think he's compared it to a mural, or he's at least compared the companies he buys to somebody else's masterpiece that they've spent a lifetime working on. And because finance touches on so many different domains, you could think of someone's portfolio or their track record as something that has some level of artistic expression.
Like there are definitely true aesthetically appealing trades. And a lot of people talk about that in the Structured Products world, that there are just some trades where it very cleverly handles several problems at once. e.g., it's usually one side that wants to hedge something, the other that doesn't care to hedge it, somebody has tax liabilities that they want to take care of and somebody else doesn't. And if you can, if you can wrap together all of the interests of all the different participants in a trade and take care of all of them in one little transaction or one big transaction, then that has some aesthetic appeal.
And, you know, thinking also about the previous question on why people who are in a given career, why they write. One reason is that some people just have an itch to write, sometimes they have something they want to get off their chest, you know, there are just lots and lots of war memoirs, for example, because you go through a really messed up situation. And especially if it's a war that was big enough that you know, a large portion of the people in the appropriate age group had a choice. And that means that just through random chance, you're going to get some really good writers who also spent a while getting shot at.
So those tend to produce a lot of books, because you don't actually have that many reasons to do writing, while you're as part of your job if you're fighting a war. Whereas in finance, a big component of the job is writing and writing well. So a lot of them, you know, a lot there are these opportunities, not for not for especially creative writing, except with angry letters that are attached to 13D filings. But there's room to write down your thoughts, articulate why you're doing what you're doing. And, and so maybe, maybe that's just like this, this valve, like the safety valve where you vent a little bit of that, for sure, right? Because you are actually producing memos and explanations and reports and pitches all the time.
In that sort of vein, what parts of that world, and I'm gonna throw economics into the finance world as well, do you think are underrepresented in popular culture?
I think, so going back to some of the narratives, it is just easier to tell a story with good people and bad people or complicated people who are mostly bad, and you know, good people who can tell they're good. And I think there's a lot of people in finance who are just more complicated. And also, for some people, it's just a job. And you, because of the fact that the financial services industry interacts with so much of the rest of the economy, what you do does have these downstream effects that are actually pretty visible to you. But a whole lot of people who have a job, that is just a job, also have downstream effects.
So you know, if someone is a farmer, and they grow a lot of corn, they are contributing to obesity problems. If someone is working at John Deere, and they're at a factory that makes replacement parts or at, you know, designing the next generation of tractors, they're also contributing to this obesity problem. But it's in this much more indirect way where you sell a product, and then that product gets used, however it gets used, and you don't really have a lot of exposure to what he uses, it gets put to him.
With finance, you do have, like, if you make an investment, you are still tied to that company. There's a lot of data and disclosure around it. So it's easy to tell. It just makes a lot of those complex interrelationships more salient. So I don't know, I don't know how much of it is, is that finance is actually involved in more and more really troubling things. Or if it's just the case that living in a complicated economy often puts you in a position where you're participating in something you'd rather not participate in, but it's very hard to extricate yourself and finance makes it more visible.
And one of the interesting things I note is that at least recently, there has been a revealed preference, if not stated preference towards increasing the level of financialization of the world, whether it's sort of Robinhood changing the market microstructure or whether it's crypto. It shows that there is a demand, or at least a latent demand that they're tapping into. I wonder what you think about the revealed preference part of it being pretty strongly in favour of people actually wanting to play in this space versus the stated preference, in many cases, that we should not over financialize parts of the economy or the ecosystem.
So I will push back a little bit against the latent preference idea, because I think there are some ways that the supply and demand picture have changed. So the latent preference thing would be pure demand, it was untapped demand. And then Robinhood figured out how to tap into it.
But I think you have two important trends going on. One is that as the world ages, you actually have larger balance sheets. So if everybody dies before retiring, then nobody needs to really save for retirement. So if you have money being saved, it's because some people want to accumulate excess, they want to accumulate more capital and pass it down to their kids, or donate to charity, or just enjoy it and have a higher standard of living, whatever. But if you have a lot of people who are going to work until they're in their 60s, and then live until they're in their 80s, then they need to accumulate capital, because they have to live off of it for 20 years, which makes the plan even harder.
So if you have this, if you have the demographic transition, where there are a lot more countries where people can live to an older age, and also you have fewer people being born, you just you have an increase in average age, which means that increase in the overall demand for savings relative to the current amount of economic activity. So that's one piece; that's sort of why balance sheets everywhere are expanding. That's why Japan's has expanded a lot more than any other places. It's also why it's more sustainable, and it looks like Japan has had really high debt to GDP for a long time. But a lot of that is actually demand driven. Savers in Japan want to save money, they want to save it in a very cautious way if they actually keep a lot of their savings in cash relative to the rest of the world. Cash and cash equivalents. So that's one piece.
And then the other thing, the thing that is probably more of a driver for Robinhood, is that the transaction cost of executing a trade has been going down. And you know the margins on that they're always tricky to figure out, because they're always different ways to make money on a trade. But a couple things have happened, one of which is market making got automated. And the term you use for that, if you're complaining about it, is high frequency trading, but it's really that there used to be people whose job was they would keep an eye on dozens or hundreds of stocks. And if a stock deviated from the range they expected to be in, they would buy or sell. That was a decent way to make a living, but also really inefficient, because if the whole thing that you're doing is just looking for slight deviations from a number, that's something computers are really good at that humans are relatively worse at.
So we automated a lot of that, but because it's automated, that meant that there are these more skewed returns to being good at it. So it used to be that there were a lot of people who were market makers who made a decent living. And now you have a smaller number of high frequency trading firms that make a whole lot of money. But it's really because they've eliminated a big inefficiency in the market.
And then, and then there is also efficiency in terms of acquiring lots of users. Although that's a very unstable equilibrium. Robinhood grew at a time when there were a lot of organic marketing strategies you could use that would just get you really cheap users. So you can find analyses of how Robin Hood marketed itself where it was things like once you sign up, they tell you, you can move up on the queue in the queue, if you tweet, and if you promote it, etc. They basically got people to acquire additional users for them. That stuff, you know, can still work. But as more timelines have gone from purely chronological to algorithmic, it's given companies like Twitter and Facebook, the ability to turn those dials so that somebody else's monetization strategy that's running through Twitter is going to take the form of, paying Twitter money, CPC or CPA rather than giving people some reason to tweet.
When you look at these kinds of changes that are happening, forgetting the marketing side for a second, do you feel like we have too much financial innovation or too little financial innovation?
I think it's just a hard, hard metric to capture. Because there's a lot of stuff that it's measurable when it's part of the financial sector, it happens whether or not there's a financial sector. And so as the cost of doing it through institutions and doing it at scale goes down, you actually see a larger financial sector, even though what’s actually happening is there's less investment in that category.
So if you think of a category of taking care of people who can't work, the biggest category, people who can't work or don't want to or people who are retired. Basically, the three ways you can take care of them are one, family, two, charity to government programmes, three, the financial system that allows them to save money and then live off their capital afterwards. So the US has mostly opted towards three, a lot of countries in Europe are more at two, but the default for most of human history was option number one.
And what option number one means is that everybody is basically providing unemployment and disability insurance to all of their relatives. And because you have a small, small risk pool, that either means that you're over saving for that you have too much slack in the system, or you don't have nearly enough and you know, one family member with an illness or disability ends up just kind of making your life making your standard of living significantly lower and making that a threat or family significantly lower for as long as you have to take care of them.
So if we switch some of that stuff, not so much the disability stuff, but the saving for old age stuff, if we switch that from doing it through family ties to doing it through the market, then you actually get a lot more diversification and diversification is, you know, to return to the free lunch, it is just if you don't have an edge in picking what vehicle you save, then it's better to pick a bunch of them, than pick a smaller number and be exposed to the Socratic risk.
So I think given that kind of noise, it's just really hard to look at any kind of aggregates and say, you know, financial services used to be X percent of the economy, and now they're Y% of the economy. Because a lot of that is that it's more like financial services used to be 0.5Y% of the economy, and were under-measured, and they've gone down as a percentage of the economy. But the percentage of them that gets measured has gone up. So it looks like it's growing.
On an individual basis, I think I agree with you. On an aggregate basis, I think one of the interesting things about the sector is that ultimately, capital allocation and efficient movement of capital helps unlock innovation across sectors. So in some ways, you can look at something like a CDO, and can say, as many newspapers did, that's just complete fraud. But that's not right. I mean it's a better way of assessing and allocating risk with its own downsides. That's kind of what I mean by saying financial innovation, whether it's new products, or new ways of analysing parts of the economy that remained illegible or intangible, and making them legible and tangible, so we can actually build a future around it.
Yeah, if something goes down, it goes back to the Jim Barksdale quote, about bundling and unbundling. You know, there's this value you can add by finding people who are in a position where their business is set up to make individual whole loans, but they really want exposure to the asset class, there's a broader asset class of returns of those loans. And so if you give them the opportunity to sell those loans into CDOs, and buy a piece of the CDO, instead, they do get that upside from diversification.
On the other hand, it's a kind of limited upside. And there are these costs of actually keeping track of the loans and things. So it's not not a pure win. And it is, again, there's this big measuring problem of what is the trade off between getting diversification versus weakening the incentives to analyse things or maybe you do set up a system to analyse things because somebody has to buy the equity slice of that CDO.
On the other hand, if you have a cohort of people who own the equity slice, they actually have a weird set of incentives where their real incentive is to produce CDOs, whose internal correlation between the assets in the CDO are higher than it looks, because the equity slice is designed. So it usually is a very high chance of going to zero, but if the returns of the assets within that CDO are more correlated, then its odds of going to zero are actually lower because the odds that nothing defaults are actually higher. So it does create some weird perverse incentives which are a little bit hard to manage.
Moving from one sort of markets to another, to prediction markets? Why do you think they're not mainstream or used widely enough yet?
Hmm. Well, there is this adverse selection issue, which I've written about. The more fine grained the prediction market is, the more likely it is that someone who's making a bet actually has unique information. And since it's a market, you need people on either side of the bet. So you need a market maker who says, I am going to bet against people who come to me because they want to make a bet, because I assume the average person is overconfident, which is true on average.
But the more fine grained those bets are, the more likely it is that the person you're betting against has a really strong and accurate view. So I think the example I gave was like, you can imagine a prediction market on “Will there be a coup in the next month”, “a coup anywhere in the world next month”, versus a prediction market on “Will there be a coup in this particular month in Venezuela”. And if you're a market maker, and you start to see that, everyone, every counterparty you have is betting there will be a coup in April of next year in Venezuela, then you assume that they're better because they have good information.
And so the prediction market has better information, but the market maker has bigger incentive to dissipate. And the natural response for market makers is okay, if we know that there can be informed traders, we have to rapidly respond to that by widening the spread which we quote. But if they do that, then that means that the prediction market actually doesn't have doesn't give you very much alpha, if you are using that information. So it becomes not worth doing, it becomes easier for the person who's trading on that information to just use Treasury futures, or oil futures or some other broad asset class that's going to react to that news in a predictable way.
But presumably, those are going to be only a smaller subset of the available questions. So for example, questions about things like the spread of Covid, or questions about new products that are actually going to be created, would be created. There's a bunch of these types of questions where you can legitimately have disagreements about what's going to happen in the future, because it's not a determinant event that the person can actually have a disproportionate impact on. But we don't even see them used much for that, right? I mean, they're getting slightly more popular, I feel in the sort of small blogosphere-esque circles, but they've definitely not broken out. And the few times that they were tried out in corporate settings, they didn't really get used all that often. I remember, Google tried to use it, for instance, at some point.
Yeah, Google was using it. And I don't really know what happened with that. I think there was someone who was doing automated market making there, So they had a prediction market. And I have seen some of the Covid markets, those do get pretty interesting.
But then it turns out that you also have this issue with exactly what you're betting on. Like, you can imagine someone who hears about delta, the Delta variant really early. And they make a bet that in 2022, the Delta variant will be the predominant variant. And you know, for a while it looks like they're right. But once Omicron comes along, then they are actually wrong, even though the core thesis they had, which is that on variants, there will be mutations, and they're going to spread. That core thesis was right, but then the exact implementation of it was wrong.
And yet, going back to the financial markets, you can view financial markets as probably this evolved way to elegantly break up what it is that people actually want to bet on and what they don't. So think about commodities markets. There are oil and gas trade magazines that will give you quotes on several 100 different varieties of crude oil, because there's just a lot of different kinds, and you know, a lot of different things you could extract for them. But really, there are a handful of global benchmarks like WTI, and Brent, and a lot of other things are just priced in reference to those. So you will, you'll have something that's priced as WTI minus X dollars a barrel or plus X dollars a barrel or something.
It's just hard to have a market where everyone wants to trade if there are too many things to trade, and nobody can coordinate on what the main asset is. So it's taken me a long time to actually figure out what those assets are, what the parameters are, and actually make the trades doable so that there are counterparties on both sides. And prediction markets have a hard time solving that coordination problem because we've used oil for a long time.
So there have been markets, futures markets and oil for a long time. We haven't been betting on Covid for all that long. Hopefully, we will still not be betting on it in a couple years. So they just get less of that institutional knowledge on how to set up these bets so that everyone can work on what to bet on.
On a similar note, one of the interesting things that I noted about Covid especially was that there were a lot of folks who called it, shall we say, early. Even if by early I mean they called it, say, more than a few weeks ahead of the March market crash last year. The proportion of those people who called it early but actually took that thought to the logical conclusion and actually did something with it in the markets were much, much lower. a) I guess, do you feel my characterization is correct, anecdotal as it is, and b) why do you think that is?
So I think there are a couple things going on. One is what you had to bet on with Covid. If you're betting let's say, It's January, you've seen these disturbing headlines at Wuhan. You've seen some of the videos, you think it's a big deal. But you notice nobody else seems to care. You have to not so much bet on the disease, but that the people will start to care. And, you know, in January, there were news stories about it. And I remember it was trending on Twitter, in late January, briefly, or the hashtag Coronavirus is trending. Yeah, I think it was because of the Wuhan lockdown. So people knew about it, but the market didn't react.
And I started looking, keeping an eye on whether or not people were wearing masks on the subway in New York. And I didn't really see any change in mask wearing. So it looked like people just wouldn't care. And that was disturbing, like a bunch people are gonna die and no one's really gonna care. So you had to be willing to make this bet that at some point, it'll get bad enough that people will notice and care. So, you know, you start with this pretty simple question that you're asking of will a lot of people get this disease?
And then you have to start asking what will the response be, and you also had to make the bet that the response is not going to be too effective. So you're actually betting on this narrow window, where you realised this is a big deal. And so we shut down travel from China, like every country, and we all start testing people for Covid all the time. You know, that's an inconvenience. And, you know, maybe, maybe clips a couple of basis points off of GDP growth, but not a catastrophe.
So, you know, shorting the market ahead of that maybe maybe you make 2%, because the S&P goes down a bunch when it goes down a little bit for a few days when this is enacted, but then it comes back up. So you had to bet that there wouldn't be a really strong reaction, but there would eventually be a reaction.
And that was just a tougher call to make. There's some people I knew who bought puts on airlines and cruise lines. And that worked out really well. Assuming they exited those positions at the right time.
And then, as it turns out, there was this other problem that you had to think about if you were shorting in anticipation of Covid, which is that the economy of the US and the economies of the developed world, were generally performing below their capacity, and they just they had capacity to absorb a lot of government deficit spending. Without having runaway inflation, we eventually did get higher inflation, but not, you know, not hyperinflation. So there wasn't a limit to that spending as many expected. But where Covid just prompted a strong enough economic response, initially on the monetary side, than what they did on the fiscal side, that it offset a lot of the direct impact.
And also then it turned out that there were a lot of people who were out of the labour force but could get back in, there's a lot of just latent potential economic activity, that given a sufficiently stimulative fiscal policy, that activity would actually happen.
So that turns out to be from the short sellers perspective, the biggest risk, which is what if the Covid response actually pushes us to engage in more more GDP maximising economic policy, and then that policy sticks around and you know, you could end up with this worst case scenario from the Covid short seller thesis, which is basically, we lose a huge chunk of GDP in Q2 2020. In exchange for which we get GDP growth that is half a point higher indefinitely.
And when you look at how effective companies are at capturing incremental economic growth, it ends up meaning that you're discounting 1% or 2% higher corporate earnings growth over the life of every company, and especially for the companies that have these long duration cash flows that ends up compounding out to a huge number. And you as a short seller would’ve been betting against that and ultimately lost. So it's hard.
Maybe that this is an argument for prediction markets before using markets in assets as a de facto prediction market. Worst case scenario, someone is smart enough to figure out what is the long term macro impact of Covid. Then when they hear about Covid spreading, they buy and stocks don't crash. And then, because stocks don't crash, nobody realises it's a big deal and more people die.
Very interesting. Would it were that was the alternate reality that we had to look at! Sticking with the theme of prediction markets or predictions in general, what probability would you place on FAANG, or MAAMA at this point, Or MAGMA, being dominant the way they are in, say, two decades.
Being dominant like MAGMA? So the base rate is you want to bet against that. And you can go back and look at who were the largest stocks in the 70s and 80s and 90s, the early 2000s. And the list changes a lot like early 80s used to be of oil. IBM, one of the biggest companies of course, of the 2000 people. But yeah, typically there is a mean reversion at the top.
And the instinctive counterargument I have is these companies are really well-run. But of course, if you ask the average person in the year 2000, what is the best major company? You know, they'd say, Well, obviously General Electric and obviously Jack Welch. So these things are just really hard to judge at the time.
One of the things that I find interesting is that if we had this conversation about tech ten years ago or maybe even five years ago, Google would have been near the top of the list. And now they're near the bottom of the list for very smart young techies, one that seemed to want to go work. They might still do it because they get paid well and you get a foosball table, but that doesn't seem at the top of what they’d want, which means arguably the base rate is accelerating.
Yeah. You know, it's always hard to judge these things, especially as a company gets bigger, it becomes more conventional and it becomes less of a question of like, would you want to work there or not. For a lot of people, it's a default they would totally accept given the opportunity, and a lot of them get that opportunity. And except they just don't have to think that much about it. Google still has a whole lot of employees and a lot of them are engineers, so a lot of people are making that choice.
Yeah, you're right. It's not that they are necessarily that excited about it, but it's almost like you have to think of the base rate of excitement about jobs in general, and that's low for the majority of people. A job is just a job. The things you want to do in life are not the things that are about going to the office. You want to do things at the office. They give you enough money to do the other things you want to do and give you enough time to do those things. And right now, the salary package in general and the Google one in particular does seem like you can have a very nice standard of living without having to work crazy hours. Amazon might seem the exception there, but you know a lot of people or people I talk to at Alphabet, I guess we call it, they seem to have a nice work-life balance.
That's a good way to put it. Moving on to sort of something slightly broader. One of the things that I noticed, at least in the world, as you know, Keynes was a wonderful economist and an active market participant. And I think as his views evolved over time, a lot of it was impacted by the fact that at least once he lost a substantial proportion of his net worth in the markets. We don't seem to see that happen much anymore. Is the gap between theory and practice much wider today than it should be?
I don't know how many people start out as theorists and then go into finance, but there are definitely people who have a finance background and go into theory. You do have books like Weapons of Math Destruction was, I think, written by Cathy O’Neil who had worked as a quant at D.E. Shaw. One of the big promoters of modern monetary theory and previously worked at a hedge fund.
Soros has been trying to quit running a hedge fund, because he has wanted to be a public intellectual for the majority of his career at this point. And I think that the appeal, it's just too fun to speculate as to fund. So he ends up getting pulled back. Maybe he's finally fully retired at this point. We'll have to see.
I think part of the reason Keynes is an outlier is that there were a lot more low hanging fruit in terms of understanding how the economy works when he was writing and working. And so it was just more plausible for him to make those kinds of contributions. And also the market was more inefficient. I think the story is that every morning he would spend an hour in bed reading the Financial Times and looking at the stock quotes and looking at currency and commodity quotes and making his investment decisions and then would go off and do his academic stuff and lead his very busy social life. It's harder to do that now. You do have to specialise more in one thing or the other. So I think a lot of this is less about the decisions people make and more about what the efficient frontier is for doing both theory and practice.
On to something a bit more abstract, organisational design. I've had a general thesis that the amount of experimentation we see in the types of companies that actually exist around the world is surprisingly less. You can jump from Apple to Samsung, and for the most part you will kind of know what department means, what you need to do, etc. Can we see more types of experimentation actually happen here in a practical sense?
On the question of experimenting with company forms, it is interesting that it's hard to find that much diversity in how companies are set up, and it's also kind of funny that sometimes a company will try to do something some totally radical as an organisational structure and it turns out that what they evolve into is something a lot closer to a more standard corporate hierarchy. Maybe that's because we've all just been trained to expect that.
If you work in an organisation where you suddenly find out there are no managers, no job titles, you still have to pick somebody who is going to break the tie. When you're trying to figure out who's working on what etc. And maybe it's just that this is sort of how humans end up organising their behaviour is that you have groups of people who work together and someone ends up basically in charge, usually not in charge, like they're the only ones making any decisions and showing any agency. And everyone does what they say.
But if everyone has slightly different views on what needs to get built and in a 10 person team everyone builds 10 percent of a different thing, then you end up with nothing. You end up with just a worthless pile of spare parts. But if everyone builds 10 percent of a specific thing, then you end up with the thing. So you need to coordinate that somehow. And maybe a natural way to do that is just somebody ends up being in charge.
That's one piece I do think, and this is actually something I plan on writing about at some point in the near future. I do think that one of the things that DeFi can help with is that it can include new forms of management, whether that is something that is a more democratic system or something that has more non-negotiable rules rather than norms that don't necessarily get enforced.
I think the other thing that's interesting about it kind of parallels this is that DeFi gives you a limited sort of IP protection for interesting financial innovations. And this is a problem for investment bankers. Not an incredibly sympathetic class, but it is a problem that they run into, which is if somebody invents a novel new security they can't get the upside. So if it's cool and adds value everyone will start using it but they don't capture the upside from their efforts.
There have been counter examples where one company was able to figure out a new product category and actually capture a lot of the upside. The most salient one of those is Milken with high yield bonds, where he figured out that high yield bonds had better risk adjusted returns than other things. Notably, he figured this out. This is especially salient in the 1970s because one mathematical fact about a high yield bond versus an investment grade bond is that for a given maturity, the high yield bond has a shorter duration because more of the cash flows are interest payments rather than long term payment, rather than the principal being returned at the end.
So they actually are less sensitive to inflation, and they're also historically less sensitive to interest rates. They're less sensitive to interest rates because there are two offsetting factors that affect how bond price reacts to interest rates. One is, of course, part of the net present is the net present value, the bond changes if interest rates change. So that's just mathematically what the reason is.
The other one, though, is in a low or high interest rate environment, how does the company’s likelihood of making payments change? And in general, if rates are going up, it's because the economy is strong. So credit worthiness gets better. For a highly rated bond, the interest rate effect predominates. But if you look at the historical sensitivity of interest rates by bond rating, bonds get less and less sensitive to interest rates as credit ratings go down until by the time you get to the junk bonds, they actually go up when interest rates go up because interest rates only go up when the economy's doing better!
And if you have a company that is nearly bankrupt, one of the things that can save them is if the economy comes out of recession or starts growing. So Milken found that for the time, these bonds were pretty illiquid. Like a lot of them, they were junk bonds because they'd been issued as investment grade and then defaulted. And because bond trading is not centrally organised, you had to know who owned it and you had to know who wanted it.
So it had some network effects. As Milken pioneered the original issue junk bond market, he was able to create a larger market. One is that the network effect was growing. Another was that a whole lot of people would have a very big favour owed him because he had been able to raise money for them at a time when they otherwise couldn't.
So back at the net assessment, part of what it was dependent on was just one person with basically photographic memory who was really, really into bonds and also worked 18 hour days for a big chunk of his career. It's hard to replicate that.
The ultimate upshot of that is if you have a new financial innovation, you should be really proud of yourself and you should expect all the big investment banks to copy it if it's good. So you won't get most of the upside.
Back to DeFi. If you can somehow create a token that participates in that innovation and you can create a DAO that issues whatever clever, structured product you've invented, it's possible, not certain but possible, that you actually capture more of the upside. And I think people will start finding interesting ways to do that. Interesting ways to create new financial instruments and then create a way to monetise some form of ownership of them or to instantiate the network effect of them in something that has a market price.
We're going back to a corporate organisation that, you know, hopefully, I guess the optimistic view would be there are a lot of other ways to organise human activity to do things that turn a profit or at least a break even, and that if we have dollars, we'll be able to experiment more cheaply with them.
And then the pessimistic but base rate compliant way to think about it is though they're just not that many ways to get a bunch of people to work together towards one end and that we have mostly figured these out, that a lot of things like equity comp have been around for a long time. U.S. Steel had an employee stock purchase programme that had 70000 participants in the 1920s. So we've been finding ways to give people salary plus equity for a really long time. And we've had org charts pretty much since the dawn of the railroad or the railroad industry, where a few had a lot of different people who are performing the same function in different places and you have the telegraph so you could actually coordinate them.
So a lot of this stuff, we sort of didn't know the optimum. We figured out something that was pretty close to the outcome, and we're still tweaking things around the edges and figuring out little details. But it does seem like a lot of companies, the companies that do weird stuff as they get bigger, they get more conventional, which sort of implies that there is a cost to weirdness. Maybe there's a cost to assimilating people into a weird way of doing things, but in general it seems like there's just a lot of convergence.
I think the other relevant headline there is, I guess, two related headlines. One one that argues to your point, one argues against it. The one to your point is that Wall Street Journal article about fintech bankers, Steve McLaughlin. He sets up these deals that are not just a straightforward sell, where you get paid a percentage of sale price. It's more like I will help you sell the company and I will get paid a much larger percentage, but only the sale price is above X. And also you have to agree to work with me on future deals, etc.. He was able to capture a lot more upside from working with those companies by setting up an unusual deal.
On the other hand, you have Cravath, I believe, who used to have a partner compensation structure that was purely lockstep, seniority based. And they just said they're going to switch it to partly performance based and partly still lockstep. So it's just very hard to preserve idiosyncratic things. And maybe that points to one of the reasons that you're competing with people who are more normal. And so you have to be really, really confident that you've got the right unusual method for organising things. Or it just won't won't work and you'll lose all your best people.
Cravath is particularly interesting because they also created one of the original sort of management consulting style reforms internally, right? So they're definitely interesting. But going back to organisational structures and growth. In the hedge fund world, there have been several great funds that started out as newsletters or quasi newsletters, right? I mean, Ray Dalio being the biggest one. We haven't really seen that happen in other sectors. I'm just curious, why do you think that is?
It’s happened sort of in venture capital where it's very dependent on deal flow. And there are definitely companies that have gotten big, in part because they really understood email marketing and did some form of newsletter. I think with finance, it goes back to the point earlier that a lot of the output is actually writing and by thinking really hard and writing down some numbers and then prose explaining those numbers. So it maps very well to the news of the world.
Like a newsletter, you can do a newsletter that is basically in the same format as a pitch that you would send to a portfolio manager of a hedge fund. Probably wouldn't be as interesting to general people, but it would certainly be if you're looking for a job at a hedge fund. An easy, straightforward way to do that is write the pitches for stocks that you would want to pitch at a hedge fund and publish them on Substack. And eventually, if you're good, people will find you. Some people will copy you, but some people will find you and will try to hire you.
I think the implementation of the finance idea once you have the writing done is very, very simple. Whereas it's not as simple even if you could write a really good newsletter on how you would redesign the Google user interface, which is a challenge. If you had some interesting ideas there, you know you could write an interesting piece on this. You could do your mock ups, you could explain your reasoning. But it's not like you go to Interactive Brokers and order. “I want to build a new Google” and they give you the Google. There's a pretty big cost to actually implementing things outside of finance.
You could still show it, though, right? The benefit in finance is that you can tell whether you're right or wrong pretty easily. Buy Google, then you just wait awhile and see if it goes up or down. So there are clear ways to say whether you know whether what you said actually added or subtracted value. But even otherwise, I feel like it is underused as a strategy across spaces. Perhaps a new one where it is coming back is on the media company side, but I haven't really seen it anywhere else.
I think the big determinant is, what is the implementation cost versus the cost of creating a memo! You might be able to think of categories where implementation is happening, no matter what the memos determine what gets implemented first. Think about Procter & Gamble where they had a pretty memo focussed culture where you write a terse, direct memo that exactly explains what you're trying to build. You know, if you wanted to get into the consumer packaged goods industry, it might be interesting to do that. Just come up with pitches that are basically for things that a large CPG company should offer. As a product, you can throw in your market research and mock up what the product would look like on the shelf.
And then, you know, if you find one that you've actually done a bunch of thinking about, it seems appealing. And maybe it's, you know, a food product. And after you send this newsletter note, you get a bunch of emails from people who are buyers at grocery stores who say it does this exist or are you just making it up? If it exists, we'd like to order boxes of it. You know, at that point, you can probably find a manufacturer who's able to produce the product to your specs.
There's a case study that Justin Mares wrote about a while back where he said that he wanted to know if there was demand for bone broth. So he bought the domain. And then once he got enough orders, he told everyone, you know, we don't actually have the product right now, but we'll give you half off if you can wait three months. And then he has three months to find a manufacturer.
So that kind of strategy where you build the sales funnel first, that works. But it actually does come back to the idea that implementation is relatively cheap compared to strategy where there are contract manufacturers. You could totally do with electronics. It'd be interesting.
But then you end up with another alternative that doesn't matter, which is with a good component with a tree pitch. Your target audience is always going to make the trade. But in this case, where you're writing about a product that would get sold, it's not like you're writing something that is an ad for the product to the consumer. You're actually writing something. There's an ad that is directed to somebody else who's going to help you sell that product to the consumer. So there is another step of intermediation that makes it less of a clean way to participate.
You can think of a finance newsletter as just a hedge fund with zero AUM. But a newsletter about interesting variants on Kraft Mac and cheese or interesting ways to compete with Kraft in the mac and cheese market, that is not really a food company in any meaningful sense. It’s a food company that's just missing every single thing that a food company needs. Instead of 90% of the effort, 90% of the input to whatever the output is which used to be, it's more like 5% of the end. But it may be a very valuable 5%, but it's just harder to accrete the rest of it around just that.
Last question. How have you found the world to become more explicable or less explicable since you started writing The Diff?
It's gotten a lot more explicable because one of the things that The Diff gave me a very strong incentive to do was learn a lot about several industries. I used to sort of do that by traversing the supply chain a little bit. So I knew a bit about internet media companies and ended up learning about traditional media companies and so on. But now part of what I do is just pick an industry and try to figure it out.
Like the energy industry does not have a huge effect on the internet companies, but it is an interesting industry which has a ton of capital invested in it and affects everyone's lives all the time. And it's going through a transition where at some point we will probably not be drilling for oil, pulling it up, burning it and just rolling that way. We will stop doing that because there are cheaper alternatives or they'll stop doing that because policy makes the existing version untenable. Or we'll stop doing that because of something apocalyptic that happens once you get sufficiently high CO2 concentration in the atmosphere.
One way or another it'll go away, but very uncertain as to whether it goes away in 20 years or in 200 years. And in the meantime, a lot of interesting things have to happen. And a lot of these companies have been very thoughtful, they also know that the industry is not going to be around forever and that if you're doing an investment that has a five year life, then it's an oil and gas world and that's that's probably sensible. You don't have to think too much about excessive risk. But if you're building an oil refinery and it has a 50 year useful life, you have to think pretty seriously about what the policy environment will look like in 2060, because it can turn your investment from pretty good to a waste of capital. If it turns out that the thing you built becomes illegal at some point before, it's before you've fully amortised the cost.
This is brilliant. I know we could have kind of gone on for a lot longer, but I notice we're already 15 minutes over, so I'm going to pause here for now. Byrne, thank you ever so much for this. Can’t wait to do it again!
No worries. I appreciate it. This was fun.