21 Building my portfolio in public 🇨🇦
Hunting underfollowed stocks before they hit everyone’s timeline.
Real money. AI-assisted research. Every win & loss documented.
Goal: beat the market.
Let’s see if this shit actually works.
Here’s the X post I’d make from this whole research rabbit hole:
I’ve been screening small caps for one thing: businesses where the next 3–5 years could look NOTHING like the last 3–5. Not penny-stock garbage. Real revenue. Real customers. Inflecting growth. Manageable dilution. A realistic path to 3–5x. My current finalists: $AADX — defense/precision strike, $RDW — space/autonomy, $BLZE — AI storage, $ALLT — cybersecurity, $CGNT — intelligence software. The goal isn’t to predict the next $PLTR. It’s to find companies BEFORE everyone starts calling them the next $PLTR.
I spent way too long building scanners before realizing the real edge isn’t “finding more stocks.”
It’s having a process that stops you from doing dumb shit when emotions kick in.
My system now forces every decision through:
portfolio truth → thesis → challenger → risk → sizing → decision → execution → review
No buy because “it’s down.”
No sell because “it’s up.”
No changing the thesis after the fact.
No pretending a pending order is a position.
No rotation unless the replacement is actually better.
The goal isn’t to predict everything.
It’s to make fewer stupid decisions.
$PNG is getting absolutely smoked right now.
Down ~26% from its August highs while the actual business has been putting up strong numbers.
Q2 revenue: C$27.3M
Adj. EBITDA: C$5.0M
Announced 2026 orders with Covelya: C$355M
This is where I separate a falling STOCK from a broken THESIS.
So far, I see the first. Not the second.
If Kraken keeps executing, this selloff is starting to look less like something to fear and more like an opportunity.
I’m watching $PNG very fucking closely here.
@drayinvests $ORCL is one of the few names where I think the analyst targets might actually be underselling the long-term story. If cloud + AI infrastructure execution keeps delivering, $250 won’t look crazy at all.
@probablyGF That actually makes the move way more interesting. If tariffs are the main reason $KITT got hit rather than deterioration in the underlying business, the question is how much of that 50% cost exposure is permanent vs something management can mitigate.
$NOW is the one I’d separate from this group. The market is treating “AI kills SaaS” like every software company has the same problem. ServiceNow sits directly inside enterprise workflows, data and automation. If AI agents actually become useful, somebody still needs to govern what they’re allowed to do and connect them to the business. That could make $NOW more important, not less. AI disruption is real. I just think the market might be shooting the wrong hostage
Calling Adobe the cockroach of SaaS is disrespectful to cockroaches 😂 $ADBE keeps surviving every “this is the end” narrative, but surviving isn’t the same as compounding. I’m not interested in catching the trendline just because it bounced. Show me that AI can restart durable growth instead of slowly commoditizing the exact products Adobe built its moat around.
This is the part people underestimate after a 20% earnings dump. You don’t get bonus points for being the first guy to “buy the dip” 😂 I’d rather miss the first bounce and buy once the valuation + fundamentals actually give me an asymmetric setup than catch a falling knife just so I can tweet the fill.
Mostly agree, but complexity gets blamed for people using it badly. Simple decisions are great. Simple thinking isn’t always. I want the research underneath my decisions to be ridiculously deep, then the actual decision at the end should be stupidly simple: buy, hold, trim, sell, or do nothing.
This is the kind of setup I actually like. Everyone loves a stock when it’s vertical, then it gets smoked and suddenly nobody wants to touch it 😂 If $AAOI is genuinely capacity-constrained through mid-2027 and that >$200M hyperscaler order converts into shipments, the business could look very different while the stock is sitting near the 200-day. I care way more about that order converting than RSI though. Fundamentals first, chart second.
@QualCompounders This is underrated. People think investing means you need to be making moves every week. Sometimes the best trade is realizing nothing changed and leaving the portfolio alone. I’m getting way more comfortable doing absolutely nothing until the evidence gives me a reason to move.
My portfolio review gave me a funny answer today: I found stocks I want to buy, but my best use of the next $500 is still paying down margin. That’s investing. Every dollar has to compete against stocks, cash AND debt. If I have to invent a reason to deploy money just because I got paid, I’m probably already making a bad decision.
$FLT finally did the thing I was waiting for. The defence thesis went from “Canada wants more drones” → qualified supplier → actual initial 100-system order. Now comes the part people love skipping: deliveries, margins, follow-on orders and cash flow. The other 4,900 systems are options, not guaranteed revenue. Story is getting real. Now prove the economics.
$ORCL might be the most interesting stock in my portfolio right now. Revenue +30%. Cloud +62%. OCI +121%. RPO hit $664B. And the stock closed around $150. The market is basically saying “cool, now prove you can make money on this insane AI buildout.” Fair. That’s also exactly why I’m interested 👀
1/ I think AI is going to make the gap between people MASSIVE. Not because some people will have AI and others won’t. Everyone will have it. The difference will be what you choose to spend your mind on before AI ever enters the equation. 🧠
2/ I know guys who know EVERYTHING about hockey. Prospects. Contracts. Trades. Stats. Salary caps. They can argue for an hour about why a coach made the wrong decision. These guys aren’t stupid. That’s the interesting part. They clearly know how to learn. They just chose what to become experts in.
3/ Your attention is basically capital. You only have so much of it. Sports can own it. Dating can own it. Drama can own it. Gambling can own it. Social media can own it. I’ve wasted stupid amounts of time on some of this shit too. The question is: what is your brain actually compounding?
4/ A 2-hour argument with your girl isn’t always a 2-hour argument. You think about it afterward. You wake up thinking about it. You’re thinking about it at work. Some shit quietly rents space in your head for DAYS. That mental real estate has a cost.
5/ I work construction. I know exactly what an hour of my life is worth because someone literally pays me by the hour. So I’ve started looking differently at the few hours that actually belong to me. Those hours might eventually be what gets me into a position where I don’t HAVE to sell more hours.
6/ Last night I stayed home working with ChatGPT on an investing system I’m building. I started with one normal portfolio question. Hours later I had completely changed how I wanted the system to understand the world. The important part? I DIDN’T HAVE THAT IDEA WHEN I SAT DOWN.
7/ That’s when something clicked for me: thoughts compound too. One thing you learn changes the next question you’re capable of asking. That answer creates another question. Do this for years and you don’t just know more shit. You literally start noticing things other people don’t notice.
8/ That’s why an experienced hockey fan sees a defensive mistake before I even understand what happened. Thousands of hours built the pattern recognition. Now imagine directing that same obsession toward business, programming, engineering, investing, your trade, whatever actually moves YOUR life forward.
9/ And now give that person AI. That’s where this starts feeling almost unfair. AI removes so many places where curiosity used to die. Random question at 11PM → explanation → another question → research → code → actual project by 2AM. A thought doesn’t have to disappear anymore.
10/ But AI still can’t give a shit for you. It can’t decide what you care about. It can’t make you curious. It can’t force you to sit alone and follow an idea nobody told you to follow. If your entire brain is occupied by other people’s shit, there’s nothing interesting for AI to amplify.
11/ So I don’t think the future is “people with AI vs people without AI.” It’s people who use AI to consume more vs people who use AI to create more. People who use it to stop thinking faster vs people who use it to keep thinking longer.
12/ I’m 21. I could spend 2 years building my investing system and some guy buying the S&P 500 while watching every Canadiens game could still beat my ass 😂 But that’s almost beside the point. I’m betting that spending my 20s learning how to think, build and use these tools compounds into something.
13/ Money is leverage. Ownership is leverage. Code is leverage. The internet is leverage. Now intelligence has leverage too. AI isn’t automatically replacing your brain. It’s becoming a multiplier on whatever you choose to do with it.
14/ Protect some empty space in your head. Get obsessed with something that belongs to YOU. Follow stupid questions further than they deserve to go. Build shit. Learn shit. Because I think AI is about to make what you choose to do with your attention matter more than ever.
I think investors spend way too much time staring at P/E ratios and not enough time looking out the fucking window. Where are people moving? Where is electricity demand going? What are governments spending on? What is becoming scarce? What is getting automated? What are cities being forced to upgrade? Financial statements tell you what already happened. I’m trying to figure out what causes the NEXT financial statements before everyone else sees it. Then I’ll worry about the multiple 👀
Did my full portfolio review this morning.
The most interesting setup in my portfolio right now is $ORCL.
The stock closed around $150 even after:
• Revenue +30%
• Cloud +62%
• OCI +121%
• RPO hit $664B
• $30B+ of new AI cloud contracts
This is exactly what I look for:
BUSINESS EVIDENCE ↑
FORWARD THESIS ↑
PRICE ↓
The market is still worried about the insane capex, financing and negative FCF.
And it should be.
But that’s what creates the opportunity.
The question isn’t whether Oracle has AI demand anymore.
It’s whether they can turn that demand into attractive returns on all this capital.
$ORCL just moved higher on my buy list.
I started looking into $SHOP because an underwear company somehow managed to follow me across the internet with ads.
Ended up finding a much bigger thesis.
Shopify isn’t just a website builder anymore.
Payments. Checkout. Merchant data. Audiences. Product discovery. AI agents.
The part that really interests me: as shopping moves toward AI, Shopify doesn’t need to build the winning AI.
It just needs ChatGPT, Gemini, Meta and whoever comes next to want access to Shopify merchants.
AI can own the storefront.
$SHOP wants to own the rails underneath it.
Going down this rabbit hole hard.