- Noah Jacobs Blog
- Posts
- On Bubbles
CIX
[Behavioral Edge, Price Vs Value, The Perversity of Risk, FOMO, 5-10x Productivity Gains, Shared Delusion, The Contrarian's Creed]
Thesis: Precisely because it is so challenging to maintain when it matters most, disciplined behavior is the most desirable way to generate alpha.
This post is highly influenced by one of Howard Mark's books, "The Most Important Thing." If I told you it was pedantic in the past, I've started re reading it and have realized that the fault was mine for not appreciating the nuance.
[Behavioral Edge]
Last week, we discussed that the goal of investing is to generate alpha, or get a better return than you should at a given level of risk. There are three ways you can do this:
Information
Analysis
Psychology / Behavior
I claimed that Psychology was the most important edge, but I did not go into it nearly enough. So, today we'll get further insight into behavioral edges by examining the frenzied & destructive, herd like behavior of investors who find themselves in a bubble.
The biggest investing errors come not from factors that are informational or analytical, but from those that are psychological.
Luckily for us, we don't have to look very far - we're in an AI Bubble right now!
[Price Vs Value]
One of the most critical concepts to understand in investing is the concept of Price vs Value. In short:
Price is what you pay, value is what you get.
This sounds straightforward enough - value is something intrinsic about some investment, price is what someone is willing to sell it to you at.
In theory, markets exist to facilitate trade between two parties who have disagreements on the price & value of an asset, and the current price of the market therefore represents some average perceived value.
Then, your job as an investor is to make investments where you disagree with the market's price. A good investment is one where you believe the value is lower than the price.
If someone is selling Ford stock at the price of $5 a share, that implicitly means that their perceived value of Ford is less than $5*. But if I think the real value is $10 a share, then I will gladly buy their Ford stock. Eventually, the market will see that I am right about the value and they were wrong. Then, I will make money!
All things considered, the bigger the gap between your perceived value and the price you can get something at, the better the investment is for you.
In other words, if I could buy Ford stock at $3/share rather than $5/share, and I still think it’s worth $10, it is now a better investment. The gap between the price I can get and what I think it's worth is even higher now!
However, one of the chief distortions in markets is that people tend to reverse this: they often perceive the value of investments to be greater as the price increase, and they tend to perceive the value as lower as the price decreases.
*This is naive in that it ignores the fact that you’re likely balancing risk across a portfolio.
[The Perversity of Risk]
The Perversity of Risk is a term coined by Howard Marks to discuss this strange correlated relationship between price and perceived value:
When everyone believes something is risky, their unwillingness to buy usually reduces its price to the point where it's not risky at all...
…when everyone believes something embodies no risk, they usually bid it up to the point where it's enormously risky.
On one hand, if something is very, very expensive (like tech stocks in the dotcom bubble, or ai rigth now), a lot of people actually start to view it as less risky and more valuable!
Really, all things considered, if the price of a stock goes up, by definition, it becomes a worse investment. Nvidia is NOT a better investment now because it's worth more than it ever was... as a matter of fact, unless the business is 25% more valuable than at the start of the year, it is definitively a worse investment!
This is especially hard to conceptualize when the investment does have high intrinsic value. If, for instance, the cure for cancer came out tomorrow, it would be REALLY valuable... but, there would still be some price at which it was not worth buying stock in the company that was producing the cure.
If everyone is on cloud is there alpha in being partly on prem?
Something can truly be worth a lot and still be egregiously over priced.
And then on the other hand, if something is very, very cheap, people tend to believe that it's not worth investment at all. As we discussed with the Ford example, if value is constant, a lower price makes it a better investment. In reality, lower prices make us second guess are valuation of the asset in question.
In short, the "very obvious notion" that the bigger the gap between value and price, the better the investment, is not at all held to in practice.
The good news is that this gives us a chance to generate alpha, by looking at these distortions. Since we are in a bubble right now, it is much more interesting to examine the situation where people are overpricing investments rather than underpricing them.
[FOMO]
The hallmark of a bubble is when people give up rationale thought in favor of fitting in with the crowd and out of FOMO (fear of missing out).
Howard Marks talks about the psychological pull of group think & herd mentality. He actually references a study by Solomon Asch, where Asch more or less gaslights experimental subjects into saying something that was markedly untrue by planting other subjects that were all paid to shill the belief in the untrue thing.
That intelligent, well-meaning young people are willing to call white black is a matter of concern.
To make matters worse, in financial markets, not only do you see the crowd doing the same thing as each other, but you see them doing the same thing AND making money!
When some people make money doing things, and then tell other people that they made money doing that thing, and then those people start making money doing that thing, and then tell other people that they will make money doing that thing... well, it's hard to ignore it! Especially when it starts with legitimate value and some way to actually make money.
But, at some point, the money stops coming from the value and starts coming from the person who is told they will get value.
What makes for belief in silver bullets? First, there's usually a germ of truth... it produces profits for a while... eventually, the appearance that (a) there's a path to sure wealth and (b) it's working turns into a mania... after it has popped--a mania is called a bubble.
Since everyone seems to be making money except you, you stop asking questions and just do what everyone else is doing... until the music stops.
But, if you maintain rationality throughout the mania, you can start to see where some of the claims about value maybe don't actually justify the price.
[5-10x Productivity Gains]
You don't have to look very far to see some extreme claims about AI, and you don't have to look much further to see that they contain a lot of hot air.
Check this excerpt from a CNBC piece on Garry Tan, the leader of Y Combinator:
The sheer power of LLMs is allowing startups to stay leaner, Tan said. What would've once taken "50 or 100" engineers to build, he believes can now be accomplished by a team of 10, "when they are fully vibe coders."
While this is actually kind of a tame claim in terms of AI outlandishness, I still see big problems with it.
For starters, if we assume he's comparing fast engineers at a startup to 50-100 engineers at an established company, the claim is vacuous and doesn't say anything. Silicon Valley is already built on teams of 10 or less that can run circles around teams of 50 or 100 or even 1000!
Now, if we assume that he's comparing a team of 10 vibe coders at one startup to a team of 50-100 engineers at another startup... then, all things considered, you would expect the YC companies to be well, about 5x-10x better than they were a couple of years ago. Yet, in the same interview, the best stat that Tan can give is that YC companies are "growing 10% week over week," which is pretty vague. Growing what? Revenue? Users? Customers? He also doesn't give us a point of reference other than to say that it's never happened before.
All of this is besides the point, because the alleged productivity gains are very likely bullshit - studies are starting to come out showing that using ai actually causes people to code more slowly! For Tan to be right, you’d think that these studies would at least see some positive improvement.
But, if you hear a smart man shouting that we'll get unheard of productivity gains and the value’s been barely realized yet... well, that distorts the perceived value and makes you want to invest in ai, too, and use it as much as possible. Which is super convenient for the guy who just paid some non zero price to invest in hundreds of more ai tools that are all supposed to increase your productivity!
One guy, even one as influential as Garry Tan, is not enough to make a bubble.
Moving past just this one quote from one guy, Ed Zitron recently wrote a very compelling piece that emphasizes how big the gap between reality (value) and narrative (price) is. Some highlights:
$39B in total revenue generated by ai companies from… $370B in total investment
The leaders (OAI, Anthropic) hemorrhaging money to make it work
A Salesforce Study (yes, the company that has all but rebranded to “Agentforce”) claiming: “These findings suggest a significant gap between current LLM capabilities and the multifaceted demands of real-world enterprise scenarios.”
Satya Nadella giving a circular answer about how GenAI will generate $130B in rev for MSFT by saying “we’ll know it’s working when the GDP is growing".”
In other words, across the board, there is a huge gap between the price being paid and the value being created. This is what a bubble looks like.
[The Contrarian's Creed]
I interchangeably speak about public markets, private markets, and even individual use of time as place where you make investments. In a lot of ways, it all blends together—whether everybody is investing in the same thing in public markets or with their own time, I think very similar rules dictate how to generate alpha.
And to be clear, I'm not against AI; I use some form of it every day, and it does make me go faster. BirdDog is dependent on AI in a number of different ways. There will be incredible winners coming out of the AI craze, just like there were incredible winners coming out of the internet craze. Many of my friends are building AI companies, and I hope they are some of the winners.
What I am against is blindly following anything, especially when everyone believes it just because everyone believes it. If you make any investment, it should be because you seriously believe that the price is lower than the value, not because somebody else told you that it was. If you’re going to use or invest in AI, be prudent and do so in a way that YOU believe you will get a positive return on, not in a way that someone else is claiming you will.
Financially, investing in a company “because AI” is a terrible thesis, just like investing in arbitrary companies “because .com” was 25 years ago. Personally, “learning ai tools” rather than learning how to do something is also quite foolish.
The distance between price and value always collapse. In this case, a lot of people who made investments or built businesses on or made life decision on outlandish claims of 5-10x productivity gains or that you “don’t need to learn anything but ai” will lose their shirts.
It’s AI & agents now, it’s been crypto or real estate or tulips or ‘the internet’ in the past. It will be something else later.
Resisting FOMO & herd mentality is, in the long term, one of the surest ways to generate alpha.
Live Deeply,
