What We’re Reading (Week Ending 20 February 2022) - 20 Feb 2022
Reading helps us learn about the world and it is a really important aspect of investing. The legendary Charlie Munger even goes so far as to say that “I don’t think you can get to be a really good investor over a broad range without doing a massive amount of reading.” We (the co-founders of Compounder Fund) read widely across a range of topics, including investing, business, technology, and the world in general. We want to regularly share the best articles we’ve come across recently. Here they are (for the week ending 20 February 2022):
1. Classic 05: A History of 5 Stock Market Crashes w/ Scott Nations – Sitg Brodersen, Preston Pysh, Scott Nations
Preston Pysh 01:18
So Scott, in this episode we’re going to be talking about the crashes of 1907, 1929, 1987, 2008, and the Flash Crash of 2010. I’m sure some, if not all of those years are very familiar for a lot of people in our audience. But let’s start all the way back to 1907. And for people not familiar with this crash, JP Morgan was a key character in this crash. So talk to us a little bit about him and talk to us about the characteristics of the 1907 Crash.
Scott Nations 01:46
JP Morgan was a fascinating man. He was a fascinating man. He was a man of privilege. He was born into privilege. His father was Judas Morgan, who was, well, essentially made the family even more wealthy by selling civil war bonds in London during the American Civil War. JP Morgan was educated as you would expect somebody of his wealth. He had a peripatetic education. He was obviously educated in the United States, but also in Switzerland and in Germany. As a young man in Germany, he developed an appreciation for art; actually a love for art, which informed his private life. But JP Morgan was really raised to be a banker in the early part of the 20th century. He was by far the wealthiest man on Wall Street. Well, maybe not the wealthiest, but certainly the most powerful. He was absolutely the most powerful man on Wall Street. He was called, “the Zeus of Wall Street.” And he really was involved in every aspect of finance in the United States in the first part of the 20th century.
Stig Brodersen 02:54
Now, if we go back to October 1907, the market crashed almost 50 percent from previous ESP. The panic might have been even worse if it hadn’t been for JP Morgan, and he pledged a huge sum of his money and convinced others to do the same. Perhaps you could tell us some of the factors about what caused the crash, but also the story about Morgan’s intervention in the market. It’s a very fascinating story.
Scott Nations 03:22
Before 1907, the United States was really beginning to understand that it was going to be the American Century. It was powerful. It was probably at that point, the most powerful country on the globe. And so, frankly, the United States got carried away with itself. And we’ll talk about some of the specifics of some of these crashes and how these shared some similarities in a bit. But you asked specifically about Morgan’s intervention in the market. This was before the US Federal Reserve existed. In fact, the Panic of 1907 was the cause the Federal Reserve was created. But if you were worried about the market; you were worried about the Panic of 1907, the person you went to see was JP Morgan because, again, he was so powerful. And one example occurred in the midst of the panic. On Thursday, the 24th of October of 1907, when the Press of the New York Stock Exchange went to JP Morgan. At the time, his office was directly across the street from the New York Stock Exchange. Very simply, “Mr. Morgan, we will have to close the exchange early. There’s simply too much selling.” And JP Morgan understood what that meant.
Scott Nations 04:37
His question was: How in the world do you ever reopen a stock market that you’ve been forced to close because there’s too much selling? And so, JP Morgan asked, “When do you normally close?” “Well, sir, we normally close at three o’clock, but we can’t get there. It’s too much sign.” He said, “Then, you will not close one minute early.” And his confidence was…it was not naked. He rounded up bankers in the Wall Street area. Got them all into his office. It was a time when if JP Morgan called, you came running. And he told the assembled bankers, “You have 15 minutes to raise $25 million to save the stock market.” $25 million back then was a colossal amount of money, but JP Morgan essentially said, “You have 15 minutes to raise this money or the stock market is going to close. And who knows when it will ever reopen?” And that’s just one specific story of his involvement. That’s not the first time he did something like that. Probably the most immediate, but within 15 minutes, they had raised actually more than $25 million. The officials were able to go on the floor and say, “We have $25 million to lend to investors, who are in trouble.” People were so desperate to get this money that the clerk, who was responsible for recording borrowers and amounts had his suit coat ripped off of him in the turmoil. So JP Morgan was really the man; the single man, who managed to save the stock market in 1907.
Preston Pysh 06:11
So Scott, you briefly mentioned that the Panic in 1907 was the reason for the creation of the Federal Reserve. Talk to us a little bit more about that idea.
Scott Nations 06:19
It became obvious to everybody after things had settled down after the Panic of 1907, the United States government needed a way to inject liquidity into the system and didn’t have it. And that there needed to be a lender of last resort, if you will, for the financial market. And that didn’t exist before 1907. And so, the Federal Reserve was created in 1913 because everybody realized, if nothing else, JP Morgan is not going to live forever. And we can’t rely on one man, one person to essentially bailout the stock market in times of stress.
Stig Brodersen 06:57
Let’s turn to the next crash, the crash of 1929. And to really understand what happened, we also have to understand how crazy the market behaved in 1927 and 1928. And I think you do a fantastic job of that in your book explaining everything leading up to the crash. In 1928, the Dow closed in 300, which probably to the listeners out there seems outrageously cheap, but that was definitely not the case. And this was at the end of the second biggest two-year run ever. It was actually more than 90%. So Scott, what drove the all-time highs leading up to the crash in 1929?
Scott Nations 07:38
In the late 1920s, actually much of the entire decade, but particularly in the last half of the 1920s, there were simply a euphoria at work in the United States. It was not just financial. It was…it had to do with the United States’ place in the world. And from a military point of view, also from an industrial point of view. So as you pointed out, in 1927 and 1928, the stock market gained more than 90%. We had come out of World War I. We felt great about our place in the world, but there were also some other things that were…for example, in a situation like that; with an economy roaring like that; a stock market booming like that, you would expect the Federal Reserve, which was new at the time, you would expect them to raise rates. One Federal Reserve officer at one time described it as “taking away the punch bowl, when the party really got going.” And the Federal Reserve did not do that. In fact, they kept rates low. They kept rates too low, largely because they wanted to help England return to the gold standard after World War I. That was a tragic mistake–keeping rates that low. There was also a roster of new technologies that were unleashed following World War I. Radio would probably be the biggest, but also the automobile industry really got going; really came into its own. And then, America just felt good about itself. And so, all of those things spawn this euphoria that eventually made its way into the stock market. And the stock market got carried away with itself.
Stig Brodersen 09:13
It’s very interesting. And you mentioned that before here with the Federal Reserve. Now, could you talk more about which actions did it take? You already, again, briefly touched upon that. But if you should outline and put some years on like before, during and then, especially after the crash. It was very interesting, the type of monetary policy that the Federal Reserve decided to carry out.
Scott Nations 09:34
The Federal Reserve really started making errors in policy in 1924, when they were essentially begged by the British government to help them get back on a gold standard by lowering interest rates here in the United States. And they did that. And they continued that sort of policy, and eventually the Federal Reserve simply lost control of the monetary situation in the United States. There was so much money being made by industry and individuals that they were happy to loan that money, the stock market speculators, and that’s sometimes called, “the call money market.” Call money is money that’s available to investors to speculate with, and for a long time that money had been provided by banks. And now, outside investors were providing it, and they did it in droves because interest rates were so low, otherwise. And the bubble was undeniable. In the late 1920s, American brokerage firms paid a $100,000 to put a brokerage office on a single transatlantic liner, the Barren Garea. A $100,000 just to open up; the opportunity to open a brokerage office. Another brokerage firm opened a tent at the US Amateur Golf Open in Pebble Beach. The stock market was such a phenomenon, and the rally was such a phenomenon that people didn’t want to get away from it.
2. Latest success from Google’s AI group: Controlling a fusion reactor – John Timmer
As the world waits for construction of the largest fusion reactor yet, called ITER, smaller reactors with similar designs are still running. These reactors, called tokamaks, help us test both hardware and software. The hardware testing helps us refine things like the materials used for container walls or the shape and location of control magnets.
But arguably, the software is the most important. To enable fusion, the control software of a tokamak has to monitor the state of the plasma it contains and respond to any changes by making real-time adjustments to the system’s magnets. Failure to do so can result in anything from a drop in energy (which leads to the failure of any fusion) to seeing the plasma spill out of containment (and scorch the walls of the container).
Getting that control software right requires a detailed understanding of both the control magnets and the plasma the magnets manipulate, or, it would be more accurate to say, getting that control software right has required. Because today, Google’s DeepMind AI team is announcing that its software has been successfully trained to control a tokamak…
3. EE380 Talk [on cryptocurrencies and their externalities] – David Rosenthal
“Blockchain” is unfortunately a term used to describe two completely different technologies, which have in common only that they both use a Merkle Tree data structure. Permissioned blockchains have a central authority controlling which network nodes can add blocks to the chain, and are thus not decentralized, whereas permissionless blockchains such as Bitcoin’s do not; this difference is fundamental:
- Permissioned blockchains can use well-established and relatively efficient techniques such as Byzantine Fault Tolerance, and thus don’t have significant carbon footprints. These techniques ensure that each node in the network has performed the same computation on the same data to arrive at the same state for the next block in the chain. This is a consensus mechanism.
- In principle each node in a permissionless blockchain’s network can perform a different computation on different data to arrive at a different state for the next block in the chain. Which of these blocks ends up in the chain is determined by a randomized, biased election mechanism. For example, in Proof-of-Work blockchains such as Bitcoin’s a node wins election by being the first to solve a puzzle. The length of time it takes to solve the puzzle is random, but the probability of being first is biased, it is proportional to the compute power the node uses. Initially, because of network latencies, nodes may disagree as to the next block in the chain, but eventually it will become clear which block gained the most acceptance among the nodes. This is why a Bitcoin transaction should not be regarded as final until it is six blocks from the head.
Discussing “blockchains” and their externalities without specifying permissionless or permissioned is meaningless, they are completely different technologies. One is 30 years old, the other is 13 years old.
Because there is no central authority controlling who can participate, decentralized consensus systems must defend against Sybil attacks, in which the attacker creates a majority of seemingly independent participants which are secretly under his control. The defense is to ensure that the reward for a successful Sybil attack is less than the cost of mounting it. Thus participation in a permissionless blockchain must be expensive, so miners must be reimbursed for their costly efforts. There is no central authority capable of collecting funds from users and distributing them to the miners in proportion to these efforts. Thus miners’ reimbursement must be generated organically by the blockchain itself; a permissionless blockchain needs a cryptocurrency to be secure.
Because miners’ opex and capex costs cannot be paid in the blockchain’s cryptocurrency, exchanges are required to enable the rewards for mining to be converted into fiat currency to pay these costs. Someone needs to be on the other side of these sell orders. The only reason to be on the buy side of these orders is the belief that “number go up”. Thus the exchanges need to attract speculators in order to perform their function.
Thus a permissionless blockchain requires a cryptocurrency to function, and this cryptocurrency requires speculation to function.
Why are economies of scale a fundamental problem for decentralized systems? Participation must be expensive, and so will be subject to economies of scale. They will drive the system to centralize. So the expenditure in attempting to ensure that the system is decentralized is a futile waste…
…The costs that Proof-of-Stake imposes to make participation expensive are the risk of loss and the foregone liquidity of the “stake”, an escrowed amount of the cryptocurrency itself. This has two philosophical problems:
- It isn’t just that the Gini coefficients of cryptocurrencies are extremely high[4], but that Proof-of-Stake makes this a self-reinforcing problem. Because the rewards for mining new blocks, and the fees for including transactions in blocks, flow to the HODL-ers in proportion to their HODL-ings, whatever Gini coefficient the systems starts out with will always increase. Proof-of-Stake isn’t effective at decentralization.
- Cryptocurrency whales are believers in “number go up”. The eventual progress of their coin “to the moon!” means that the temporary costs of staking are irrelevant.
There are also a host of severe technical problems. The accomplished Ethereum team have been making a praiseworthy effort to overcome them for more than 7 years and are still more than a year away from being able to migrate off Proof-of-Work.
4. Geoffrey Moore – Building Gorilla Businesses – Patrick O’Shaughnessy and Geoffrey Moore
[00:03:07] Patrick: Geoffrey, your books were some of my earliest education in the world of the competitive landscape of technology. I’d actually start at the end in terms of how I think about your work, which is with the concept of a gorilla as a business. Everyone’s going to know, Crossing the Chasm. We’re going to talk a lot about all the insight from that book and that thinking. But I think the gorilla as a concept is for me a great unifying theme of your work, aspirationally we all are going to want to be gorillas or invest in gorillas, or start gorillas at some point. Maybe just begin there. What do you mean by a gorilla company? Define that for us to begin.
[00:03:43] Geoffrey: The simplest definition is a market share leader in a powerful category. In order to sort of take that model apart, we created something called the hierarchy of powers. The idea behind the hierarchy of powers was, go back to investing. If you want to invest in a successful company, you want to invest in one that has more competitive advantage than the alternative investments. How would you actually analyze competitive advantage? That led us to something called the hierarchy of powers. This is the core investment model by on the Gorilla Game and a book called Living on the Fault Line and Going Forward. The hierarchy of power says the most powerful power is what we call category power. It has to do with the technology adoption life cycle, and where is the category that this company specializes in monetizing, where is it in its adoption life cycle? For most businesses most of the time it’s on what we call main stream. In other words, the category’s been established for a decade or more, there’s budget for it. It’s settled out. There’s a pecking order of vendors in the category and the category probably grows close to GDP growth rates. Value investors spend most of their life with categories in that world.
Tech investors, and my whole world is tech. We invest at the beginning of these life cycles. Sometimes before this, there’s not even a category, it doesn’t even exist yet, it’s called category creation. But the key moment in that category development life cycle or what they call the technology adoption life cycle, is when all of a sudden the world goes all in on the new paradigm. The way we went all in on cloud computing, the way we went all in on mobile apps, the way we’ve gone all in on streaming video. When it goes all in what happens is, all of a sudden the world which in a prior year did not have budget for this category, now everybody has budget for this category. And so it creates this huge secular uplift and spend, we called it the tornado. We had a book called Inside the Tornado, huge secular. That’s category power. If you are in that category, that rising tide floats all boats, that is the number one predictor of your future success for the next several years. That’s why you see these incredible valuations in companies that are losing money because the investing community said, yeah, but they’re in the hot category. Having said that, the next thing we said is, well, that category is going to sort out with a pecking order. The power law of returns from that pecking order is, the gorilla is going to get the lion share or the gorilla’s share if you will.
Number two will probably get half of what the gorilla gets. And number three will get half of what the chimp gets. And so that led to gorilla to chimp monkey sort of returns. And so the idea behind the Gorilla Game was, you would see a category going into tornado. You would buy a portfolio of companies that could win. As you saw who was winning, you would gradually exit the ones that are monkeys and chimps and put more and more money into your gorilla. And then you would hold the gorilla because the gorilla’s power position, what happens is the ecosystem forms around the gorilla, which instantiates the gorilla permanently in that category. Now you can screw it up, but in general, it’s not just that the gorillas powerful during the tornado, even on main street, the world is now organized permanently around the gorillas de facto standards and whatever. There was just a clear sense of the sooner you could identify the gorilla and then concentrate in the gorilla, the better it would go…
…[00:09:43] Patrick: I want to come back to category creation and how you think about the idea of a category itself, but this is a great excuse to talk about your notion of architecture. If we were to think about a two by two matrix or something with open and closed architecture on one access and proprietary and non proprietary on the other, this is for me, a critical unlocking idea. Salesforce is a great example. Most power full version of a gorilla. I love the litmus test that you never get fired for blank. The blanks are the gorillas. But talk us through this concept of architecture. Why is this so important within a category? What does it mean? And what does that two by two matrix mean?
[00:10:19] Geoffrey: The difference between open architecture and closed architecture was, Apple has a very closed architecture. You don’t participate in Apples architecture. Whereas Android has a very open architecture. Okay. The idea is, do you want other people to complete your solution? The cable box was contained, but the Roku is an open architecture. In general, I think originally it was all closed. The IBM architecture was just IBM. DEC was just DEC. The Sun just Sun. No, actually it was Sun, they began to do open architecture. They would buy their storage from a different vendor or you’d get your operating system from the Berkeley operating system. That was the beginning of open architecture. I think what we learned during the last 20 years is in general open architecture beats closed architecture, because closed architectures always have a single point of failure. Meaning if any part of the closed architecture doesn’t work, you can’t ship. In an open architecture if you have a failure of one component, you can get it from another vendor and get back in the game.
Now open architecture is harder to manage for quality, and so that was always the challenge, but that was closed versus open. Proprietary versus non-proprietary, has to do with who gets to control the next release of this thing. Open source is not proprietary. Open source, there’s no locking. But proprietary there is locking. And so the most powerful idea was proprietary open architectures, where you had proprietary control of an ecosystem that involved other companies, but they had to eventually play to your standard. That’s what gave the gorilla the most power. Because now what the gorilla can do is, you have to stay with me, I’m a market mover. I don’t just move my own products. I move everybody’s products. By the way, by staying with me, you leave my competitor behind. Every time I differentiate from my competitor, then you conform to my standards, you just made yourself incompatible with their standards…
…[00:16:40] Patrick: What’s that been like watching more recently, if we think about the most pure play enabling technologies today? It might be the API companies, the Stripe, the Twilio, the Okta’s of the world. Well, how did those companies solve this problem of you need the actual use case, the actual application? Amazon AWS is enabling technology, but it was its own best first customer on the retail side. So that application problem was solved by them. How do you approach these pure play, hire this API for this one function in your application type company?
[00:17:09] Geoffrey: It’s typically around the use case. Like Okta, I think started with single sign on. People were just saying, this is such a pain in the neck, that I have to sign into this, have sign. Okta said, okay, we’ll do single sign on. And then once you did single sign on, you thought, well, wait a minute, we’re sitting in a very interesting piece of real state here. There’s a bunch of highways coming together. Maybe we should have some service stations and a restaurant, we should build some hotels. That’s what Okta did. But enabling infrastructure always starts with a problematic application use case that you can’t solve with existing infrastructure.
And so initially, first of all it looks like, well, your market is so small, there’s only this one use case and there’s only this one application and you’re building all this technology to make that better. Are you sure you want to do that? And if that was the only return, the answer would be, well, no, it doesn’t make any sense at all. But if you’re say, no, that’s my point of entry. And then expand, we called it the bowling alley phase of the technology adoption life cycle, where you’d say I, okay, I’ve got my first use case in my first industry. Can I get a second use case in that industry? Or can I find that use case in the second industry? Either way you were going to expand outward. And then at some point, if you can get enough expansion, the world goes well, hang on, this is the new infrastructure. That’s when the tornado starts.
5. Twitter thread on the importance of alignment within a company – Jean-Michel Lemieux
Another common question I’m answering working with scaling tech companies is…
Q. What’s the worst leadership advice you’ve heard?
A. By far the worst is “Hire great people and get out of their way”.
Let me explain… 🧵 (1/32)
2/ After a year leading engineering at Atlassian @scottfarkas told me in my perf review that I was doing ok but wondered why I didn’t talk and involve him more regularly.
3/ My answer was “I thought that was my job — to take away all this crap from you and let you do your CEO thing. I thought you wanted me to be autonomous. I need autonomy.” He said sure, but you should cheat “and use my brain to help you”
4/ At that point I changed some habits, involved him more in different ways, got over the autonomy complex, and we got a lot more done together. I learned a lot and we made better decisions.
5/ Since then I’ve hired many leaders and had to repeat the same conversation that @scottfarkas had with me over and over and over. Most people default to expect the wrong version of autonomy.
6/ These experiences sent me down a multi-year reflection. Why did I feel like success was maximizing autonomy and showing that I could take care of things without bugging my boss?
7/ I prioritized autonomy over alignment. It’s a million times easier to measure and feel high autonomy than it is to measure high alignment.
8/ What I’ve learned since, way too slowly, is that companies are performing a monumental balancing act trying to decide what 98% of their problem space to focus, what to ignore, and how to ship. That’s your strategy. And it’s complex, ambiguous, and changing.
9/ The hardest part of building a company is alignment on strategy and clearly communicating it. Think about it this way, a 1 degree deviation in course of a rocket heading to the sun means it will miss the sun by 1.2 million miles. A lack of alignment compounds quickly…
…14/ There’s mass confusion on what to align on? Most people just want to know the very high level goals. And this causes most companies that I work with to align superficially. Their strategy is a superficial incomplete map, they don’t remove scope, communicate clearly, etc…
…17/ The biggest alignment problem is the gap between how much people think they have to align versus what they should align on. There are many strategic decisions in the “how”. eg, what technologies to use, new system vs integration, build in core or in an app.
18/ Alignment forces you to talk with your boss and peers to: – define a strategy, narrow focus – communicate it clearly together – and ensure you’re hired enough people who “get it” and can fill in the implementation details with their teams.
19/ So…when you hire someone or you have a new leader, your number one job is alignment. And you do this continuously. It’s definitely never going to be “hire them and get out of their way”.
6. Owning the funnel – Lillian Li
Since I started writing Chinese Characteristics, I’ve been puzzled by a few observations: why is there a fanatic fixation on internet traffic? Why do firms distinguish between private and public traffic? Why did every consumer app become a super-app? And why are B2B offerings are going the same way? Why is every player worth their salt is moving into payments now? Finally, why do Alibaba’s acquisitions tend to languish while Tencent’s investments tend to thrive?
My current framing for Chinese tech’s underlying logic is that every player is always working on owning the awareness-to-fulfilment funnel (or customer journey). This is a descriptive product strategy that builds on a foundational ethos of owning the user. It outlines the offerings that a tech platform needs to provide to achieve that goal. It looks like Western players are converging in the same direction, Shopify and Google’s move into payments and Facebook’s store fronts are all part of the trend.
This behaviour pattern is stark in Chinese tech is for two reasons. The first is that a defined geographical market constrains Chinese tech. It’s no secret that Chinese companies tend to struggle with internationalisation. Unlike their western counterparts, who can build sizable companies being the best-of-breed for different geographies, Chinese tech companies have to focus on owning the user (and funnel) to grow. As I mentioned in my Bilibili piece:
Relative to western consumer tech companies, who tend to focus on “serving a function” as their core mission, Chinese companies tend to focus on “owning the user” as their core mission (though the initial wedge into the consumer is always through a function – Meituan through food delivery, Ofo through bike-sharing, etc.). Owning the end-user and their attention is what led to the rise of the super apps and Bilibili is no different….Put another way; they want to own the Chinese Gen Z population’s attention through providing a comprehensive entertainment service rather than be the platform that caters for all Chinese UGC video needs.
In an ecosystem where hundreds of competitors spring up overnight, functions and features get commoditised as soon as they are made. Owning the user by providing the whole monetisation funnel is the closest thing to a moat.
The second reason is Tencent. Tencent is the default operating system for the Chinese population, and it has a particular trait. It doesn’t rely on advertising to monetise. What started as a shrewd product decision to prioritise the user experience has had a lasting impact on the nature of internet traffic in China. Inventory on WeChat is scarce, and it commands a premium from advertisers. In chat, links are earned, Tencent’s anointed portfolio can share links, while the associates of arch-nemesis Alibaba and Bytedance get blocked for spam. Baidu has been stagnating in their ad revenue strategy from lacklustre search (owning the funnel means a walled garden approach). When traffic hegemons are capricious, everyone suffers.
Regardless of the origins, once some tech players have started the game of owning their proprietary funnel (or user), everyone has to move in the same direction just to keep up.
7. Six Questions For Derek Thompson – Morgan Housel and Derek Thompson
What aren’t people talking about enough?
For the world, carbon removal technology. It’s the most obviously important nascent technology on the planet, given the fact that even if we shift immediately to 100% renewable energy, that still leaves all the carbon dioxide we’ve already spewed into the atmosphere, which will stay there for decades. We have to find a way to vacuum the skies to avoid the worst effects of global warming…
…What do you want to know about the economy that we can’t know?
What is the right level and distribution of income to maximize total national happiness, both now and in the future? The time element of the question is important. If you waved a magic wand and made it so that everybody had equalish income today, that would clearly eliminate a lot of misery. But if you enforced equal incomes permanently, you’d create a lot of new problems. Where are the rewards for effort? Where are the incentives for hard work, or invention, or problem-solving? How do you fix the issue of free-loading, or resentment between workers and loafers in the utopia of pure and permanent equality? Mandating perfect and permanent equality doesn’t work. But it’s really, really hard to determine what level of inequality is “right.”
Disclaimer: None of the information or analysis presented is intended to form the basis for any offer or recommendation. Of all the companies mentioned, we currently have a vested interest in Alphabet (parent of Google), Apple, Meta Platforms (parent of Facebook), Okta, Shopify, Tencent, and Twilio. Holdings are subject to change at any time.
