On Friday I closed my Intel position.
Bought at $20.24 last August. Sold at $99 — a gain of 389% in 269 days, or roughly 528% annualised. Close to five times my money in nine months. I won’t pretend that’s typical. It isn’t. But it didn’t happen by accident either.
This is a worked example of how I pick individual stocks, and — just as importantly — how I sell them.
The model
Every individual stock I buy goes through the same checklist. It’s not original. It borrows heavily from Warren Buffett. Six things have to line up:
A good company with a long, credible track record.
A share price beaten down well below what the fundamentals justify.
A turnaround already in progress — new management, restructuring, a real strategic shift.
A specific catalyst, in place or visible on the horizon.
Macro or cyclical tailwinds the market hasn’t fully priced in.
Realistic potential for at least a 2-3x return in 2-3 years.
That last point is the threshold. If a stock can’t reach it, I’m not interested. The whole point of this part of the portfolio is excess returns — not 10% or 20%.
Last summer, Intel hit every box.
Why Intel was the buy
To most people, Intel looked finished.
The stock was below $20, well off its recent highs. The company had reported multi-billion-dollar losses. Its manufacturing nodes were a generation behind TSMC. The new CEO, Lip-Bu Tan, three months into the job, was laying off 15% of the workforce just to keep the lights on. The financial press wrote about Intel the way it writes about businesses that don’t survive.
What I saw was different.
A 50-year-old American semiconductor giant, trading at the price of a struggling start-up. Already promised billions of dollars in CHIPS Act money for domestic chip manufacturing. A new CEO with a serious turnaround track record. And — the part most investors weren’t pricing in — a US administration that had decided, openly, that semiconductor sovereignty was a matter of national security.
You don’t let your only domestic leading-edge chipmaker fail. Not on Trump’s watch.
The question wasn’t whether Washington would backstop Intel. It was how, and how visibly.
I bought on 5 August at $20.24.
The catalyst, when it came, was bigger than I’d modelled. In late August, the Trump administration announced it had taken a 9.9% stake in Intel — buying 433 million shares at $20.47, $8.9 billion in total. SoftBank followed with a $2 billion investment. Nvidia announced a $5 billion collaboration. The stock did what stocks do when the market suddenly realises the floor is concrete, not sand.
By the end of April 2026, with Q1 results beating expectations and AI revenue up 22% year-on-year, Intel was trading around $99.
Why I sold
Here’s the part that matters more than the entry.
I didn’t sell because I think Intel is finished going up. It probably isn’t.
The President was on Truth Social last week claiming a $45 billion paper gain on the government’s stake, and showing no sign of letting the story end there. Tan’s turnaround is starting to deliver real numbers. The fundamentals are catching up with the share price for the first time in years.
But — and this is the discipline the financial industry hopes you’ll forget — the position had hit my exit price. About two years ahead of schedule. The thesis I bought has now played out. From here, Intel is a different bet at a different valuation.
It might double again. It might not. What I know is that the easy money — the gap between a beaten-down share price and a fundamentals-justified one — has been closed. Anything from here is harder-earned.
I’d rather take the profit and put it to work somewhere that gap is still wide open.
Where the cash goes
I’m not opening any new positions. The proceeds are going into existing positions in the portfolio — specifically the ones that haven’t run yet.
Some are still close to where I bought them. A handful have drifted lower since I opened them, which means I get to pick up more shares at a lower average cost.
This is how the strategy compounds. Take a profit when one position resolves quickly. Use it to lower the cost base on the ones that haven’t.
Why this matters
A 389% return in nine months wasn’t what I modelled when I bought Intel. The model called for 2-3x in 2-3 years. We got nearly 5x in nine months. Sometimes the catalyst lands harder and faster than expected.
But the return is downstream of the process. You don’t pick stocks like Intel by accident. You don’t sell them at the right time without a plan. The checklist is the foundation. The discipline to take profit when the thesis is done is what turns the foundation into a result.
The full process — the actual positions, the buy and sell decisions, the reasoning behind each — is what my paid subscription will cover. It’s launching shortly.
Free readers get the framework. Paid subscribers see the portfolio and the trades.
I am not a financial adviser. Nothing here is personal financial advice. The Intel position described is one of mine, shared as an illustration of how I think — not as a recommendation. Individual stock picks can go wrong as easily as they go right; a 389% gain in nine months is not typical, and you should not assume similar results from any other stock. Please do your own research before making any investment decisions.