I Made 3,388% on a Single Stock. Here's How I Found It.
My best individual stock picks have returned 10x, 20x, even 34x. And they're still not an argument for stock picking as a strategy.
In July 2022, I bought shares of a company called Applied Blockchain at about $1.10. In January 2026, I sold at $38.61. A 34x return.
The trade doesn’t change my core philosophy. The vast majority of my portfolio sits in low-cost index funds running on autopilot, and that boring system is the reason my financial life isn’t dependent on any individual pick going right.
In I Analyzed Stocks for a Living, I wrote about why even professionals with Bloomberg terminals struggle to beat the market. In Five Investing Beliefs That Sound Smart but Cost You Money, I called “do enough research and you can pick winners” one of the most expensive myths in individual investing.
Both of those things are true, and I still pick stocks anyway. The slice of my portfolio I run actively has produced enough good outcomes that the occasional bad one hasn’t mattered, and I think there’s value in walking through specific trades to show what the framework actually looks like in practice.
How I found it
Applied Blockchain in mid-2022 was a tiny company providing data center hosting for cryptocurrency miners. The stock traded around a dollar. The Fed had raised rates seven times that year, crypto was in a brutal winter, and Ethereum was months away from shifting to proof-of-stake, which would eliminate the economic basis for the kind of mining the company supported. By every conventional measure, you didn’t want to own this stock.
What caught my attention was the gap between what the stock priced in and what the company actually owned. The market was valuing Applied Blockchain as a crypto mining company in terminal decline, which is fair enough, because that’s what it was. But strip out the crypto business and what remained were data centers. Buildings in specific locations, with power contracts and cooling systems already operating, that had a use case completely independent of whether anyone ever mined another Bitcoin.
I had a thesis about what those data centers would be worth.
The thesis
In mid-2022, demand for high-performance computing was already accelerating, and the constraint wasn’t going to be silicon. It was going to be the physical layer underneath. Power capacity, cooling, the ability to actually plug something in at scale. You can’t just build a data center overnight. The permitting, the grid connection, the construction itself, all of it takes years. Anyone who already had operating capacity in 2022 owned something the market would eventually have to pay up for.
I didn’t know which specific use case would drive the demand. AI was an obvious candidate, but it could have been general cloud computing expansion, or scientific computing, or any number of other things. The point wasn’t to predict the application. It was that compute demand was a one-way bet, and the physical infrastructure to serve it was a constrained resource.
Applied Blockchain owned that infrastructure. The market was treating it as worthless because of what it was currently being used for. That was the gap.
My thesis was that the assets were worth multiples of where the market priced them, and that the path to realizing that value would come either from a strategic pivot, an acquisition by someone who wanted the capacity, or a re-rating as compute demand made the underlying assets visible. The asymmetry was clean. Big upside if any of those paths played out, and a downside floor set by the value of physical infrastructure that existed regardless of what happened to crypto.
What happened
The pivot happened in May 2023. The company (now renamed Applied Digital) launched specialized AI cloud services and announced its first major contract worth up to $180 million. The stock jumped 25% that week. Over the next two years they kept executing: a $5 billion leasing agreement with a hyperscaler, a direct investment from Nvidia, revenue growth of 250% year over year by fiscal Q2 2026. The stock went from around $1 to over $40 at its peak.
Why I sold
I sold in January 2026 at $38.61 because the stock had run too far, too fast. After a 34x move, it seemed more likely to go down than to keep going up. I took the profits and redirected the money to positions where I still saw a gap.
The trap with winners this big is that you start to feel like you have a special read on the company and should hold because you saw it first. The price doesn’t care that you bought at $1. It only cares about the next dollar of value the company creates, and at $38 the market was already paying for several years of that value in advance. The original trade was the gap between $1 and what the assets were actually worth. Holding past $38 is a different trade, on a different setup, and one I wouldn’t have opened from scratch.
How I think about the individual stock slice
A meaningful portion of my brokerage account, around 40%, is in individual stocks. The rest is in ETFs that provide the stable, diversified base. The ETFs are what make it possible to take real risk on the individual names without my financial life depending on any one of them.
This is the part of the personal finance internet that doesn’t get talked about much, because most of the content is calibrated for people who haven’t built the foundation yet. The advice for someone with no emergency fund and no automated investing is correct: stay away from individual stocks, just buy the index. The advice for someone who already has the foundation looks different. Once the base is solid, you can introduce asymmetric exposure on top of it, and that’s where the actual wealth creation happens for individual investors. I wrote about this framework in The Asymmetric Bets Framework, and the individual stock slice is one of the places that framework actually applies in my own life.
Within that slice, the principle I keep coming back to is buying when the consensus is against you. This part is uncomfortable by construction. If the trade felt obvious, the price would already reflect it. Most of the positions I’ve taken that worked involved buying something other people were selling. So did most of the positions I’ve taken that didn’t work, which is the part nobody mentions when they tell these stories. Contrarianism on its own isn’t a thesis, it’s a precondition for one.
One more thing
I bought Applied Blockchain in July 2022, four months before ChatGPT launched. The thesis required looking at a dying crypto miner and seeing a compute infrastructure play that the market wouldn’t be willing to price in for another year or two.
I’ve spent the last year using AI heavily for investment research, and I’m confident it would not have flagged APLD as a buy at $1 in 2022. The consensus view at the time was that the company was in terminal decline, and consensus is what AI returns when you ask it for an analysis. I wrote about this in AI for Investment Research, but the APLD trade is the cleanest example I have of why it matters. The setups that produce returns like this one live in the gap between what the public information looks like and what’s actually going to be true, and that gap is where AI is structurally weakest.
What to Read Next
📖 The Dhandho Investor by Mohnish Pabrai. The cleanest practical book on asymmetric thinking I’ve read. Pabrai’s “heads I win, tails I don’t lose much” framing is the operational version of what made the APLD trade work.
📖 You Can Be a Stock Market Genius by Joel Greenblatt. The title is awful and the book is excellent. Greenblatt walks through special situations where the market routinely misprices something for structural reasons, which is exactly the kind of gap APLD lived in. If you’re going to pick stocks, this is the playbook for the kind of setups that actually have a chance of working.
📖 Thinking in Bets by Annie Duke. The framework for evaluating a decision separately from its outcome. The fact that APLD worked doesn’t mean the process was good, and the fact that other trades have gone to zero doesn’t mean the process was bad. Duke is the clearest writer I know on holding that distinction.
🎧 All three are excellent on Audible. The free trial gives you one credit to start.
As an Amazon Associate, I earn from qualifying purchases.