I like the idea of outperforming through concentration, but without fragility. I want enough weight in my best ideas for being right to actually matter, while minimizing the chance that a single mistake permanently destroys capital. That’s why I like 15-20 positions, mostly in cash-rich, cash-generative, shareholder-friendly businesses. Recent FinTwit dynamics would probably call this basically a bond portfolio, but outperformance doesn’t have to come from extreme beta.

A few sold puts approaching expiration. Easy to say in hindsight, but I should've just bought $ONON outright. I'd already done the homework and had high conviction in the stock. Not a missed opportunity just yet, though. I still like it at its current valuation. As for $SMCI, if I get assigned, I'm genuinely unsure whether I want to hold the shares. Arguably one of the worst corporate governance profiles in the entire AI ecosystem.

Being forced to break the virtuous cycle of achievement → endorphins → adrenaline → energy to chase the next big thing is the absolute worst.

I'm seeing compelling value in EssilorLuxottica $EL after this massive selloff. Between 2021 and 2025, the company increasingly traded on the excitement surrounding AI-powered glasses and their potential to become as essential as smartphones. I'd never touch a stock trading at a massive premium on the vague promise of "capturing AI demand". That kind of narrative is the bullshit I try to avoid. However, long before AI glasses were even a thing, Essilor's legacy optical business was already trading at around 25–30x earnings. Has that underlying business fundamentally changed? I don't really see it. Its competitive advantages remain intact, and its long-term growth prospects look ok/optimistic. Meanwhile, the smart glasses opportunity is considerably more tangible today than it was a few years ago. Millions of units sold, improving technology, and real consumer adoption. Yet the market has gone from paying a premium for the possibility of smart glasses becoming mainstream to punishing the company just as that possibility starts becoming a reality. The question is whether the market has simply removed the AI premium or has also forgotten why the core business deserved a premium in the first place.

Equity analysts keep massively underestimating how violently stocks can move. Their valuation scenarios always look something like: Bear case: 6x EV/EBITDA, -20% downside Bull case: 9x EV/EBITDA, +35% upside Then earnings arrive. Actual EPS is within $0.02 of one of their scenarios, and the stock casually drops 40% or rallies 80%. We've been watching this happen over and over since 2021.

Do we really think that higher interest rates and a Barclays downgrade are going to stop the AI infrastructure buildout? $HPE down almost 25% since October 9.

Okay $HPE say no more. Time to go in

Keep seeing these “best places to live” charts ranking cities by cost of living and quality of life, with several European cities near the top. A friendly warning to non-Europeans considering moving here: you're entering one of the most air-conditioning hostile regions on Earth. Those 25°C summer nights will feel very different from what you're imagining. And I'm not talking only about commercial buildings, where almost anything goes under 26°C for six months. It's also residential. To give you an example, plenty of new residential buildings in Spain are being built with underfloor heating and cooling systems. Sounds great, right? Until you discover that it's the service provider, not you, who decides when summer begins and ends. And apparently, summer runs from June 1 to September 15. That leaves you all of May, half of September, and most of October to enjoy the wonderful experience of lying in your 24°C bedroom at 3 a.m.

Power Solutions International ($PSIX) could be sitting on significant upside if execution finally catches up with demand. The company expanded its Wisconsin manufacturing footprint from 150k to 800k sq ft to meet surging demand for data-center generators, but the ramp brought production inefficiencies, preventing it from fully capitalizing on the opportunity. Even if execution only partially improves, I see substantial upside at this valuation. I'm currently selling puts with some pretty hefty annualized returns.

EU–China trade talks are moving towards fewer Chinese hybrids entering Europe and better access to China for European olive oil and footwear. 20 years ago, you'd have expected the opposite. Europe is becoming the low-value end of the supply chain. Soon, the cheap one too.

When there’s a secular inflection point driving +250% YoY revenue growth and the index barely dips:

Why do so many people treat strong divergence within US indexes - or global indexes, for that matter - as inherently bearish? To me, that’s just data. If anything, I’d say it’s slightly bullish. A small number of companies driving most of the gains? OK. That just means there’s plenty of room for rotation in case the context changes. “Only a few stocks are carrying the index” isn’t a bear thesis by itself.

Is it too late to join the $HPE party? Doesn’t really look like it. The stock has skyrocketed, but you’re still looking at 15–16x FY27 earnings, with Juniper transforming the networking side and AI demand looking unstoppable. Still a hardware business, but most likely less cyclical than the old HPE some market participants remember. Just promise me you won’t look at the chart first.