What We’re Reading (Week Ending 14 September 2025) - 14 Sep 2025
Reading helps us learn about the world and it is a really important aspect of investing. The late Charlie Munger even went 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 14 September 2025):
1. Secular Bull Market Peaks – Are We There Yet? – Tyler Grason
The concentration and valuation of the market today often draws parallels to the tech bubble. We analyzed the Tech Bubble and the nifty fifty period (late 1960s to early 1970s), both of which marked the end of secular bull markets, to assess the similarities. As to the end of this secular bull market, as Mark Twain said, “The reports of my death are greatly exaggerated.”…
…Both the nifty fifty and tech bubble periods both coincided with a Fed hiking cycle. In January 1973, the market peaked 3 days prior to the first hike, and the Fed did not cut until December 1974. During the tech bubble, the Fed began hiking rates in June 1999, or about 9 months prior to the market peak. While the Fed continues to be on pause, fed funds futures prices suggest a 0% probability that the Fed hikes by year-end 2026. Odds now show the Fed is 88% likely to cut rates in September, with 2 expected rate cuts this fall…
…Market today is cheaper than the tech bubble despite better fundamentals. The S&P 500 at 22x forward twelve-month earnings is ~15% cheaper than the peak of the tech bubble at 25.5x despite having 60% higher profit margins and 10% better ROE. When compared to the 10yr which traded at 15.9x at the height of the tech bubble, equities were 10x turns more expensive vs 1x turn less expensive today. On a justified P/E basis, fundamentals and bond yields would suggest the market today should trade at 24x, or slightly above the current multiple of 22x…
…Market concentration today looks much more aligned with fundamentals. During the tech bubble, the concentration of the top 10 largest stocks at 27% was nearly 2x above its earnings contribution. The expected earnings growth that was priced in failed to materialize. Today, the weight of the top 10 stocks relative to their earnings contribution is much more aligned at 35% and 32%, respectively. While the top 10 stocks in the S&P at 38.3x is above the tech bubble at 34.4x, Tesla at 145x is meaningfully skewing the data. Excluding Tesla, the top 10 today trade at 26.5x, or ~25% below the tech bubble peak despite returns on capital that are >2x higher.
2. This Is Why America Is Losing to China – Ross Douthat, Sophia Alvarez Boyd, and Dan Wang
Wang: I decided to take two friends and go on a lengthy bike ride in China’s southwestern province of Guizhou. This is a land where a local said, “Not three feet of land is flat, not three days go by without rain and not a family has three silver coins.”
China’s fourth-poorest province, I was surprised to see, had much better levels of infrastructure than one could find in much wealthier places in the United States, like New York State or California.
We saw very tall bridges all around us. We saw a guitar-making hub. We saw a lot of fancy new roads that were a cyclist’s dream. And it was only afterward when I realized how bizarre it was that China’s fourth-poorest province — about the level of G.D.P. per capita of Botswana, much less than Shanghai or Guangdong — was able to build all of these things.
It is a province with 11 airports, 50 of the highest bridges in the world and brand-new, spiffy highways — and that’s because China was just building a lot in its equivalent of a South Dakota or West Virginia…
…Wang: I think that the first and most important part of China’s technological success has to do with something I call process knowledge.
Process knowledge is also known as tacit knowledge, also known as industrial expertise. In a kitchen analogy, it is something like the recipe, and the hardware is something like the stoves and the pots and the pans.
But let’s say, Ross, we give someone who’s never cooked a day in his life the most well-equipped kitchen, as well as the most exquisitely detailed recipe. Are we sure that this person will be able to do something as simple as frying an egg for breakfast?
I’m not sure if that person will burn the kitchen down in some big way.
Douthat: My children have often given evidence for that hypothesis.
Wang: Yes. And I think the crucial part of technology is actually all of this tacit knowledge, process knowledge that we can’t really write down.
That is the core part of what has been driving China’s technological advantage. It started when China started making pretty simple things — socks, T-shirts, all these things that we think and know are not terribly important — before they get to slightly more complex things, like shoes.
Then they get to everything that now includes iPhones and electric vehicle batteries, and they are really good at climbing this ladder.
China’s hardware capital, Shenzhen, was mostly a backwater — making textiles all the way up until 2008, when Shenzhen started producing Steve Jobs’s iPhones.
iPhones started rolling off the line and you had this enormous work force, hundreds of thousands of people making the most sophisticated consumer electronics in the world, making the next consumer drones, more sophisticated electronics. And I think that is really the basis of China’s technology advantage: It’s just these gigantic investments and work force.
The state sometimes gets in the way; the state sometimes harnesses this work force. You also have a lot of entrepreneurial energy. I’m not sure if I wanted to define it as state capitalism with Chinese characteristics, but I just view it as technological catch-up.
Douthat: Right, but what is the difference, then, between that model and ours? Part of your argument is that America has lost a lot of that knowledge through the process of outsourcing and allowing factories to move overseas and allowing deindustrialization to happen, and becoming an information and financial services and service economy — a very rich one, but not an industrial economy in the way that China is.
I want to understand how much of this is saying there are engineering minds in the Politburo who made these choices that maybe you can only make in an authoritarian society, or maybe we could have made different choices ourselves in the U.S.?
How much of it is that versus some other element of competition or culture in China right now?
Wang: I think the crucial mistake in the U.S. was that it wasn’t even a choice that the U.S. made to outsource a lot of manufacturing. Now, there is this line that politicians like to trot out that China stole all the jobs — and sure, that’s one framing of it.
But I think a more accurate framing is that since the 1990s, big American manufacturers had been actively moving their production to China, and the U.S. government did almost nothing to restrain them.
I’m not sure whether that was actually a really deliberate choice plotted out by the Council of Economic Advisers advising Bill Clinton. Maybe it was, but I think this was just a process of business lobbying saying: Well, we need to tap into this market and produce at these cheaper places.
And something that the Communist Party actively decided was that they were going to import big American manufacturers in the 1990s and 2000s, Apple, Tesla.
If they want to build their products here, we are going to completely welcome Steve Jobs and Elon Musk to train our workers and make them as good as they can be.
That was a more conscious decision, I think, made by engineers who realized they had to catch up to the global frontier. They couldn’t do it with China’s existing level of technology, and they were going to have Americans help them…
…Wang: I think you’re absolutely right that America is highly dynamic, and I don’t want to count out America in this stage of competition. I think at various points the U.S. will look weak. At various points it will look strong.
But what are the stakes here? Because I think there is still a broad view in the U.S. that deindustrialization has been pretty bad — not just for regions like Pennsylvania or Michigan, where the deindustrialization has been felt pretty badly.
There’s also a pretty clear loss of manufacturing expertise that is represented in the declining fortunes of American apex manufacturers. Companies like Intel, Boeing, Detroit automakers and now, increasingly, Tesla.
They’ve had mostly bad news over the last few quarters, last few years. In the case of Detroit, the last few decades. Apex manufacturers are not working very well.
If we take a look at the early days of the Covid pandemic, the U.S. manufacturers were not very good at making simple products either — necessary products, like cotton swabs and cotton masks. And they weren’t able to really rejig their supply lines in order to build out critical materials.
If we take a look at the U.S. defense industrial base, after the U.S. shipped a lot of munitions to Ukraine for its self-defense against Russia, the U.S. hasn’t really been able to rebuild its munition stockpiles.
If we take a look at naval ships with the U.S. Navy, every class of ships is now behind schedule…
…Douthat: As a potential scenario for Chinese success. How could China, how could this model fail? What do engineers get wrong?
Wang: Engineers are meddling extensively in the economy. And maybe we will wake up and find one day that central planning is a ginormous failure and the Chinese will not be able to fundamentally overcome these contradictions in the model of state capitalism with Chinese characteristics.
That is a potential scenario in which the extensive meddling that has scared the living daylights out of a lot of venture capital investors in China, as well as a lot of entrepreneurs who would really prefer not to suffer through a lot of the edicts of the Politburo — they decide to not contribute so much to the great rejuvenation of the Chinese people.
I think that a lot of people have been pretty extensively burned out by the mistakes and some of the foibles of the Communist Party. A lot of what I have seen is that many young Chinese are willing to take leave of the great rejuvenation that is conducted in their name.
We have a lot of data on Chinese entrepreneurs, a lot of wealthy Chinese people who would much rather live their lives in Chinese communities like Irvine, Calif., by buying some property and just having their businesses be established in Singapore, and still not really quite trusting the Communist Party to respect everything that they want to do.
Young Chinese creative types are interested in smoking dope, just as young California types may be. They are smoking dope in Chiang Mai. I’ve spent a little bit of time seeing these people who are just as into marijuana, as well as cryptocurrencies, as folks are in Silicon Valley.
We also see a lot of Chinese migrants who are not necessarily rich, who are not necessarily the creative types, dare to fly to Ecuador, which has been visa-free for a period of time to the Chinese, and try to walk across the Darién Gap — a perilous journey to cross to the southwestern border of the United States.
At its peak in 2024, the U.S. was apprehending something like 30,000 to 40,000 Chinese who were trying to cross over into Texas. It still blows my mind that many people would try to do that to escape the regime…
…Douthat: Let’s end with advice for the United States. What are the actual implications of your analysis — and especially the bull’s case that we started with, the Chinese century case for what the U.S. should do right now? What should we be doing differently if China is poised to be as powerful as you think it might be?
Wang: I think that the U.S. should first and foremost rebuild its manufacturing base. That follows quite naturally from a lot of my analysis of China’s greatest strength, which is that China is a manufacturing superpower and China is poised to further deindustrialize Europe and it is poised to further deindustrialize the United States as well.
I am skeptical that President Trump’s efforts to reindustrialize America through the tariffs have been very effective. I am more positive about the Biden administration’s policies on efforts to reshore through industrial policy. But we can still see a lot of flaws with that approach as well.
Douthat: Do you think tariffs — essentially trade war — can’t work, in your view, because China has become too strong and resilient?
Wang: I think that the trade war, as prosecuted right now through the tariffs, is not going to be very effective. If we just take a look at the manufacturing employment data since Liberation Day in April — with the next jobs release, I’m not sure if we’ll get that data probity back — the U.S. has lost about 40,000 manufacturing workers.
It is not a natural fit if the U.S. is to become a technological, scientific superpower to advance its science by denying a lot of funding to scientific agencies like the National Science Foundation and the National Institutes of Health.
I think that universities, flawed as they are, are still driving a lot of American innovation and scientific advancements, and it also doesn’t make a lot of sense to attack universities in order to save the scientific base.
And it really doesn’t make sense to try to deport a lot of workers who may be working in the construction industry or the manufacturing industry, or to frighten away a lot of high-skilled researchers who may want to be in the U.S. from Europe or Asia to do a lot of their work here. So I think that as prosecuted, the trade war is not making a lot of sense.
The industrial push in the U.S. is not making a lot of sense. Maybe there’s something positive to be said about Trump’s energy agenda in terms of building more nuclear power, in terms of building more facilities online. Maybe there’s something positive about the deregulatory agenda. I can certainly see that case, but I certainly see more headwinds than tailwinds.
3. Are We at Bubble-Level Valuations? – Ben Carlson
Here’s the monkey wrench — Bernstein also wrote about why regression to the mean can be so tricky outside of science:
There are three reasons why regression to the mean can be such a frustrating guide to decision-making. First, it sometimes proceeds at so slow a pace that a shock will disrupt the process. Second, the regression may be so strong that matters do not come to rest once they reach the mean. Rather, they fluctuate around the mean, with repeated, irregular deviations on either side. Finally, the mean itself may be unstable, so that yesterday’s normality may be supplanted today by a new normality that we know nothing about…
…This is the CAPE ratio going all the way back to a time when Francis Galton was still alive: [Average of 17.6x since 1881, and average of 28.3x over past 30 years]
What’s more relevant here — the 150+ year full history or the past 30 years? Which average is more relevant?…
…Last week I wrote A Short History of the S&P 500 which looked at the composition change to the index over time in terms of the types of stocks. The S&P 500 was full of capital-intensive industrials and railroad stocks for much of its history. These were relatively low-margin businesses that required a large number of employees and lots of physical assets that needed to be replaced over time.
Today’s companies have more intangible assets and are far more efficient.
Take a look at average margins by decade going back to the 1990s and you can see this shift happening:
Every decade the average moves a little higher.
This was supposed to be the most mean-reverting series in all of finance. Market historians have been shouting it from the rooftops for the past 15 years. And they were wrong…
…It’s interesting to note that the biggest crash on this list–the Great Financial Crisis–started at relatively muted valuation levels. Stocks were not insanely overvalued heading into the fall of 2007. It’s just that no one saw earnings were about to fall off a cliff.
Picking tops is not easy.
4. Finding Fraud – Farrer 36 Asset Management
One of the first things I do when reading an annual report is search the PDF for the term “Material Weakness” – you’d be surprised how often you get a positive hit. A material weakness is a flaw or combination of flaws in a company’s internal controls over financial reporting that creates a “reasonable possibility” of a significant error occurring in the financial statements. For example, take Evolv Technologies that declared a material weakness in its 2024 annual report.
The discovery of the accounting mishap (it turns out an employee was overstating sales) sent the stock tumbling 50%…
…Many ‘material weakness’ declarations get remedied, or don’t turn out to be much, but their existence is cause for more work…
…Swedish small cap Intellego has been on a tear recently – with the stock up more than 300% this calendar year. The stock is being driven by impressive revenue (+152% yoy in Q12025) and profit growth (+162%). Given this, you would expect that operating cash flow would have also exploded. But would it surprise you that it has instead decreased over the same time?
This is because much of Intellego’s revenue, while recorded, has not actually been received by the company. Receivables have increased 6x over the same period.
The above begs the obvious question – are the revenues real? Let me be clear, I am not stating that this is fraud – the company has explained that some of their older contracts gave too loose of terms to their clients, and newer contracts have stricter terms. However, such a large mismatch between profits and cash should give any investor pause…
…Many of Enron’s troubles lay with CFO Andy Fastow’s creation of SPVs which he and his family owned. These vehicles had the dual purpose of raising billions for Enron (and thus allowing the consolidated balance sheet to appear debt-free) and paying himself millions of dollars…
…Going back to the Enron example, even though they showed positive operating cash flow in three annual reports prior to declaring bankruptcy, their working capital assumptions raised alarms. You can see from the above table that from 1998 to 2000 (read right to left) that both receivables jumped (see the previous example for what that implies), but to compensate, there was also a significant jump in payables…
…For years Yes Bank had posted numbers too good to be true. Their loan book grew much faster than peers, margins and profits were higher than its comparable set, and all this despite exposure to troubled sectors like real estate, airlines, and telecoms. It turns out that Yes Bank was underreporting stressed loans (they reported NPAs under 1%, whereas the RBI showed a 400-500bp difference). When the truth was revealed we saw a 96%+ drop in stock price and jail for the founder.
5. Why retention is so hard for new tech products – Andrew Chen
Just as there’s the laws of physics, weirdly there are some constant patterns that keep cropping up over time. Here are a few that I’ll share:
- You can’t fix bad retention. No, adding more notifications will not fix your retention curve. You can’t A/B test your way to good retention
- Retention goes down, it doesn’t go up. And weirdly, it decays (oh, does it decay) at a predictable half life. Early retention predicts later retention.
- Revenue retention expands, while usage retention shrinks. Good news: You lose people over over time, but the ones that remain sometimes spend more more money!
- Retention is relative to your product category. There’s nature, and there’s nurture. Sorry, you’ll never make a hotel booking app a daily use product
- Retention gets worse as users expand and grow. The best users are early and organic. The worst users come after that
- Churn is asymmetric. It’s far easier to lose a user forever than to re-win them back
- Retention is weirdly hard to measure. Seasonality is a real thing. New tests throw things off. Bugs happen. D365 is a real metric but you can’t wait
- Crazy viral growth with shitty retention fails. We’ve run this experiment many many times already, across multiple platforms and categories
- Great retention is magic. When you see it out in the wild, it’s amazing…
…You might read all of this and still have a big question: So wait, how do you get to great retention? (If I knew the answer in a deterministic way, my job as a startup investor would be so much easier, wouldn’t it?)
But let’s try our best. In my points above, there’s a few clues:
- The idea really matters.
- If you want a high retention product, you need to pick a category that is high retention already.
- You need to pick a product category where you already use an existing product every day.
- You’re going to build something that directly competes against that.
- If you win, then you’ll stop using that other product and use your product instead.
That’s a high bar, but I think it’s a good start…
…The natural counterpoint is that new markets are often more exciting than existing ones. Isn’t tech about building brand new things rather than innovating 20% on old stuff? Of course this is true, but I think this is the tiny tiny minority of products.
My counterpoint to this counterpoint is that most products actually have some kind of prior lineage, even if those prior products are quickly forgotten.
Before Instagram there was Hipstamatic, which had become the #1 paid photo app in the early App Store. It demonstrated the success of photo filters. Of course Google was not the first search engine, it was actually #10 or whatever, after Lycos, Excite, Infoseek, etc., which demonstrated consumers wanted search but that it was impossible to monetize. Tesla was not the first electric car, nor iPhone the first smartphone. Sometimes it’s the 10th iteration that matters. Some call this “last mover advantage” rather than first mover. I think an important point.
Yet sometimes new things do happen. Uber was created to turn an existing offline action — calling a cab — into an app, not because there was already a hugely successful ridehailing app. (And no, not Lyft — it was a weird bus booking thing at the time). Of course a lot of ChatGPT, with OpenAI’s 5 year journey between inception and v3 which really took off, and without any real blueprints for what it might replace. These types of journeys are remarkable, and the tech industry is better off for it, because they involve real risk as part of new category creation.
Disclaimer: None of the information or analysis presented is intended to form the basis for any offer or recommendation. We currently have a vested interest in Alphabet (parent of Google), Apple, Meta Platforms (parent of Instagram), and Tesla. Holdings are subject to change at any time.
