I closed a four-month trade with a +124.5% gain, banking ~ยฃ24k tax-free.
Fake crisis, real opportunity
On April 2, 2025, Trump announced the most dramatic tariff package in years: a 10% baseline tariff on nearly all imports effective April 5, along with much higher “reciprocal tariffs” on certain countries starting April 9. Within a week, after the market crashed, he declared a 90-day pause on most of the reciprocal tariffs. That pause, which began on April 9, was later extended beyond July 8.
The first thing that was obvious to me was that the tariffs were not really happening.
Do you remember what happened 90 days ago?
Nobody does. Trump knew that too. Calling it a 90-day pause looked like negotiation instead of admitting a loss; it was a face-saving tactic, not policy. That’s why I believed the fear was temporary, and it turns out I was right. Back in April, tech was healthy, GPT-5 was on the horizon, and the panic would pass. So I bought the dip.
The death zone of valuation
Nvidia and the other AI infrastructure leaders are now priced for perfection. Nvidia in particular gained about $1.42 trillion from April lows to June - under 3 months! They are all great companies, but valuations have overshot, fueled by FOMO money and institutional capital with nowhere else to go. Can you name one other sector growing as fast as AI right now? Probably not. Of the top 10 companies in the S&P 500, all but one (Berkshire Hathaway) are big tech firms. Everything else is flat, which forces capital to concentrate in AI. Institutional investors and fund managers cannot just sit on cash! They need to justify their existence with annual growth numbers and “active” allocations to collect fees, so they pile into big tech. Their size alone pushes valuations past fundamentals.
It’s like climbing Everest. The summit looks glorious from the base camp, but above 8,000 metres you enter the death zone. Oxygen is thin, margins for error disappear, and most deaths happen on the way down, not up. That’s where Nvidia and AI stocks are now: valuations stretched so high that one slip, one gust of bad news, one earnings miss, and the drop will be a freefall.
Cisco in 2000 is the perfect analogue. Dominant company with real technology, but so overvalued that shareholders waited 20 years to break even. I don’t need to catch the exact peak - after a rally this long, and zealous investors believing in perfection, even a small dip can spark panic selling. I was more than happy to lock in +124.5%.
AI vs bubble
I don’t think AI itself is a bubble. The shovel-sellers like Nvidia, TSMC, and ASML, and the frontier labs like OpenAI, Anthropic, DeepMind, and xAI are going to be fine. Top talent is flowing to these companies, funding is enormous, and they will keep building. The bubbles are in the copycat SaaS products, wrappers, wrappers of wrappers, and “pre-revenue” B2B startup hype. The biggest bubble isn’t in the core technology but in sentiment and valuations. Just like Cisco, the company and technology didn’t fail when the dot-com bubble burst. It was simply overvalued.
From a technology angle, most of the low-hanging fruit from the Transformer era is gone. I sold after GPT-5 launched because compared to the leap from GPT-3 to GPT-4, the step to GPT-5 was underwhelming. Scaling laws are running into diminishing returns. Clean data is running out, and throwing more money at GPUs won’t deliver exponential gains forever. The market still acts like GPU demand will rise forever, but that’s a risky assumption.
Rules > emotions
I felt no euphoria when I sold, nor when I checked the position every couple of weeks and saw it was in profit. For me this was just cause and effect: I caught a dip, understood the tariff noise for what it was, and knew tech was inherently strong.
I executed this trade cleanly because of the rules I set for myself. One of my rules is that I only speculate with less than 10% of my portfolio. Even if the money had been wiped out overnight, it would have stung for a day or two, but I wouldn’t have been in a bad place financially. That detachment let me act rationally. There’s only so much you can improve on the emotional side of investing. When stakes are high, it doesn’t matter how stoic you are; the best safeguard is setting rules up front and following them.
If this were tuition money or essential savings, I’d be second-guessing myself constantly and couldn’t make conscious decisions. Because I didn’t need the starting money and had nowhere else to spend it, I wasn’t emotionally attached. In this trade, timing the bottom was partly luck, but respecting my rules and taking profit without getting greedy was pure discipline.
Did I sell the exact peak? Maybe not, but you only know a peak in hindsight. Timing the peak is impossible because you can only know something peaked once it has started to fall. The last 5-10% is a massive gamble that usually ends badly. What matters is sticking to the plan and not getting greedy.