Really enjoyed the SoFi example from 2023, the balance sheet risk vs growth potental dynamic is textbook. One nuance worth flagging: lateral thinking is crucial but the barbell approach can get tricky when you're sizing the speculative vs stable split, especially when a tournamnet-effect candidate suddenly runs into execution stumbles before network effects lock in.
In the old days, value investing was like wandering through a quiet flea market with a flashlight. If you were willing to dig, you could find a perfectly good “dollar” selling for 50 cents because nobody else had the time or the data to notice it. Today? That same flea market looks like Times Square. Algorithms scan every table, private equity bids on anything remotely interesting, and the only things left in the bargain bin are broken, dusty, or missing half their parts. Graham’s world is gone.
So the game shifted to Buffett’s world: buy great businesses, let them compound, stay patient. Still works—but now everybody’s read the same playbook. Moats, margins, management… it’s all mainstream. The alpha isn’t in the accounting anymore; it’s in seeing around the corner. The modern investor’s job is predicting which companies can keep printing cash when the world changes, not hunting for statistical mispricings that machines arbitrage away before breakfast.
Meanwhile, the “Chinese Buffett” crowd—think Li Lu—plays a slightly different sport. They invest where policy, demographics, and national strategy practically choose the winners in advance. Instead of guessing who survives the next decade, they look for companies the system itself wants to survive. Western value investing is about understanding businesses better than the market. The Chinese version adds: understand the country better than the market. Different arenas, same quest—find the compounding machine everyone else overlooked.
Good write up and I agree with the vast majority of it. A couple of things I have a different opinion on though.
In the mid 2000s, Buffett went back to the basics when Korean stocks were undervalued, focusing on statistical bargains without even being able to pronounce the names of the businesses he was buying. Stocks selling below working capital and a PE of around 2 he said. So he still loves the classic Graham approach, but it was just never scalable and the opportunities went away when markets became more efficient, so he was forced out of them, quicker with Munger’s influence.
That efficiency however is less prominent in tiny illiquid stocks that wouldn’t make sense for people to use algorithmic trading in. Some so small it takes a week to buy 10-20k worth of shares. Tiny obscure stocks are where the biggest inefficiencies lay. Incredibly time consuming and hard to hunt for, but they are still there and any time Buffett was asked what he’d do starting over, he’s always said he’d focus on these tiny obscure mispriced opportunities.
So while things have changed a lot over time, if someone uses the right hunting ground, I think these opportunities are still relevant today.
Really enjoyed the SoFi example from 2023, the balance sheet risk vs growth potental dynamic is textbook. One nuance worth flagging: lateral thinking is crucial but the barbell approach can get tricky when you're sizing the speculative vs stable split, especially when a tournamnet-effect candidate suddenly runs into execution stumbles before network effects lock in.
Yeah this is why I qualified it saying that it shouldn’t be speculative actually, it should just be higher risk higher reward.
They shouldn’t be Oklo type bets, they should be more similar to things like Sofi in 2023.
Great article!
I'd even add two steps before:
1. Select the macrotrend you believe is worth investing in.
2. In that macrotrend, who are the dominating or fast-growing players?
Once you selected a subset of those, you can perform a valuation as you described.
In the old days, value investing was like wandering through a quiet flea market with a flashlight. If you were willing to dig, you could find a perfectly good “dollar” selling for 50 cents because nobody else had the time or the data to notice it. Today? That same flea market looks like Times Square. Algorithms scan every table, private equity bids on anything remotely interesting, and the only things left in the bargain bin are broken, dusty, or missing half their parts. Graham’s world is gone.
So the game shifted to Buffett’s world: buy great businesses, let them compound, stay patient. Still works—but now everybody’s read the same playbook. Moats, margins, management… it’s all mainstream. The alpha isn’t in the accounting anymore; it’s in seeing around the corner. The modern investor’s job is predicting which companies can keep printing cash when the world changes, not hunting for statistical mispricings that machines arbitrage away before breakfast.
Meanwhile, the “Chinese Buffett” crowd—think Li Lu—plays a slightly different sport. They invest where policy, demographics, and national strategy practically choose the winners in advance. Instead of guessing who survives the next decade, they look for companies the system itself wants to survive. Western value investing is about understanding businesses better than the market. The Chinese version adds: understand the country better than the market. Different arenas, same quest—find the compounding machine everyone else overlooked.
Good write up and I agree with the vast majority of it. A couple of things I have a different opinion on though.
In the mid 2000s, Buffett went back to the basics when Korean stocks were undervalued, focusing on statistical bargains without even being able to pronounce the names of the businesses he was buying. Stocks selling below working capital and a PE of around 2 he said. So he still loves the classic Graham approach, but it was just never scalable and the opportunities went away when markets became more efficient, so he was forced out of them, quicker with Munger’s influence.
That efficiency however is less prominent in tiny illiquid stocks that wouldn’t make sense for people to use algorithmic trading in. Some so small it takes a week to buy 10-20k worth of shares. Tiny obscure stocks are where the biggest inefficiencies lay. Incredibly time consuming and hard to hunt for, but they are still there and any time Buffett was asked what he’d do starting over, he’s always said he’d focus on these tiny obscure mispriced opportunities.
So while things have changed a lot over time, if someone uses the right hunting ground, I think these opportunities are still relevant today.