2025 Recap
Solid year, feel confident going into 2026
In 2025, I had 3 career goals, having finally shipped my game studios last game, FORGED, for the Meta Quest. Unfortunately it was not a financial success, and so after 14 years away from playing money games for a living (I was a high stakes poker pro until the end of 2011, when I had a big enough year that I was able to make the decision to focus full time on music) I have returned to the arena full time (I have always invested pretty seriously, but nowhere near full time).
That said, deep down, I like making stuff, so rather than just try to beat $SPY, I came up with a more complicated plan with 3 main goals
Outperform $SPY
Grow this substack, but keep it free.
Build my app GTOPort (now ALPHAPORT.AI) into a world class InvestTech app.
All three are going well, despite the feeling at all times that no matter what I am working on, it is always the wrong thing. That is just the nature of doing a lot. The strength to my approach is all 3 things should feed each other - if I only invested full time, writing about investing would still be a strong way to improve at it. And although I could potentially just pay for other software to help my process, I think 2026 will be the year that using my own app is actually better, even ignoring the goal of having other people use it (plenty of people are already using it for portfolio tracking now, but the Research product is about to start getting really good too).
Goal One - Outperform SPY
My returns in 2025 were very good, a little over 34%. Although a lot of the investors I talk to about investing crushed me (multiple people I respect as investors managed to achieved >60% returns), I am going to maintain a levelheaded point of view and say that any time you more than double SPY’s return, you are doing well. On top of this, since most my winners hit long term gains before I sold and I cut losers fairly aggressively (actually too aggressively in the end, even adjusted for tax) I think on a tax adjusted basis I have done quite well. I haven’t done all the numbers yet, but I am pretty sure that I only had limited long term gains and almost no short term ones.
The IB long/short account returned around 62% but my US long only account returned only ~5%. Although a lot of this is variance, I do think it validates continuing my decision to step away from a 10x10 portfolio and work with a more varied long/short portfolio which I think also benefits from AI research products significantly.
That said, I am now running a large unrealized gain in the portfolio. For now, I don’t see anything that desperately needs to be closed out. I reduced Softbank a decent amount in ‘25 but still sizeable, the DAVE position is moderately small now after multiple times trimming it (and I have basically committed to holding it forever, going for the fabled 100 bagger) and the GOOG/INTC/TSM/AMZN positions are all positioned well to gain from the AI theme, which has legs until humanity is superseded by whatever comes next. Perhaps I will trim $GOOG in ‘26 if i have something really amazing to buy though. And I do occassionally get anxious about Xi’s 2027 Taiwan plans re: TSM, so its not impossible that I just balk at some point and trade into something with less macro risk.
Here were the biggest winners on IB
And the biggest losers
ISSC
As you can see from ISSC on IB (which I also owned at one point on Chase), I lost 22k which is sort of whatever. However, this was a $100,000 (10%) position with an average cost basis of ~$10 a share that i sold in the mid 8s. So, I lost 22k, but if I had simply held into earnings rather than getting cute, I would have probably made 80k for a 100k swing, just on IB.
Amazingly, much earlier in the year I did the exact same thing, buying ISSC at $8 and selling it down over the tariffs period in the $5.80s. The missed gain here if I had just held on those positions was $63554. Combined, the number is very large.
Is there a big lesson here? I’m not sure. People say cut losers, double down on winners, momentum is your friend and you pay less taxes cutting losers. Or am I meant to be a value investor, and double down on losers, and sell into momentum? Who knows.
I think the real lesson here is that if you do a bunch of work studying a name and still like it, you should probably just hold it and not get cute. I wrote a SubStack post on ISSC, I spoke to other investors about it, I liked that it was a small cap company capitalizing on themes that other investors i respected were betting big on (see Duquesne funds investment in Woodward, who makes the fuel injection system for a Pratt and Whitney engine used in the F-35). It felt like a much more leveraged version of that, that bigger players couldn’t buy into. I knew it felt like a venture position where I needed to just close my eyes and wait for the management team to execute, but I treated it too much like a stock.
Overall, it’s not the end of the world, but this one decision dramatically changed my results for the year.
Similarly in the unforced errors category, when I bought PRME, i told myself “Do not trade this, it is too cheap for something that will almost certainly become thematic and the science is legit. Just let it sit there and do nothing, prob works pretty often". Then I rotated into other stuff in a sell off. Although this was a smaller decision than some of my bigger offenders, it actually annoys me significantly more because of how clearly I could frame the long to myself before making it. Then I watched it become thematic, would have been 100% return from my cost basis or a triple from where I sold it.
Other results would have gone better if I hadn’t sold, or had bought more, etc. For example, as you can see from the table, I had a great return on $NXSN - but would have made another 60-70% and incurred significantly less tax (many of my sales didn’t even hit long term capital gains) if I had simply held it into 2026. But unlike $ISSC, I lose 0 sleep on this.
I thought about the sale quite a lot before making it. The core investment thesis was that this was an under-covered player in the drone space, with the best financials of all of those companies, and even an Anduril supplier (I noticed this much earlier than the market, watching all of Andurils video content to try to track down what cameras they used). But all of this became known and consensus.
What remained was a solid stock, but hardly RIDICULOUSLY cheap, and very opaque. A company making cameras for drones based in Israel is always going to have a bit of risk, and you should get paid a premium for that risk. Overall, I am comfortable with the sale, even though as I said, I would have had a major windfall going into 2026 if i had just done nothing.
Similarly, I did a decent amount of work on $TEVA and built it into a core position, then promptly rotated it into something else over the Trump tariff period. Although it would have been nice to get a large return on a diversified biotech stock (that I could pretty much hold forever after it rerated), i don’t lose any sleep on this one either. Even though I had the conviction to recommend it to my mother in law at $13 when she relentlessly asked me for a stock pick (“AMZN?” “I own it”, “GOOG”, “Own that too, give me something more interesting!” she interrogated), I still did not feel especially mentally committed. It just seemed like solid diversified value.
This is how i’m fairly confident that the $ISSC sales were legitimately bad decision making. I am fairly clear headed about the difference here. I wrote no substack posts on TEVA, NXSN. My research was thorough but less comprehensive. I never had the thought “close your eyes and wait for management to execute on their very clear plans”. I just thought the stocks looked pretty good at the prices the market was offering.
Mistakes made shorting
On the short side, I did quite well overall. Many of my shorts were profitable, and given the solid performance of my portfolio (and the market) overall, this helped my returns. Especially seeing my software shorts do well while my software longs also did well was nice as this leads to extremely good performance. I will try to do more of this ‘absolutely dominate the long and short side of a sector’ in 2026. Trading around the Softbank ecosystem (ORCL, ARM, SFTBY, $9984 etc) on both the long and short sides was also quite profitable.
By FAR the biggest mistake i made was in quantum and nuclear sizing. When my quantum shorts turned against me, i actually increased the positions (on top of them already increasing themselves), something i will never do again. Overall, i got very close to breaking even in Quantum shorts, with QUBT making up for losses in RGTI and QBTS. But I closed some RGTI and QBTS out at the absolute top because the positions had just become too big. I never lost conviction in the shorts, and actually reshorted the same stocks later in the year when i had realized the losses, but it would have been much easier to just keep them smaller.
I had a similar situation with OKLO, where i shut out a short while i was in hospital because the department of energy was talking about significantly increasing nuclear production in the US and I didn’t want to be thinking about this while I was in hospital - reasonable, but the position was just too big. If it had been smaller I would have just let it ride and done extremely well. Instead, I waited a while, took the L, and reshorted it lower. This was profitable but way less profitable than if i had just done nothing.
So the key learning there is to never get greedy shorting stuff. This requires more work though, always finding big groupings of stocks to short. My hope/plan is for AlphaPort Research to really help me here in 2026.
Grow this Substack
It seems like i’ve been trying to find the time to write this Substack for years, but actually it’s only been 6 months and already close to 500 subscribers. On a comparative basis, its not that many, but to me, 500 people is still a very large amount of people, so I hope you are enjoying reading about my investing journey.
How did the Substack picks go? Well, first let me say that this Substack is not financial advice, and the goal is not to make posts for people to clone my trades. Seriously, it’s not. It is really hard to copy someone’s entrances, exits, conviction when there is a sell off, etc.
That said, if you had bought and held all of them, you definitely beat the market. But if you picked a random sample, it is quite likely you did quite poorly. That is the nature of investing unfortunately, with the winners being much larger in magnitude than the losers. The bulk of the returns were made in 1 or 2 positions.
I still hold all the losing companies however and believe they will all do well going into 2026.
Build out my InvestTech app
ALPHAPORT.AI is my investing app. Generate alpha for your portfolio using AI. It used to be called GTOPort. Why the name change? Well, GTO is a poker term that means “Game Theory Optimal”, eg find out what the computer would do and use that to inform your play. Unfortunately, we live in a complicated world with only so much time to explain things to people. And so we will go with Alpha, which resonates more with investors and requires no explanation.
ALPHAPORT.AI has come along very nicely for portfolio tracking and reached a level of stability there that is comforting. Every day it aggregates all my positions across all my accounts (Interactive Brokers, Chase, Robinhood, Schwab, etc) then takes a time stamp. So I can jump back and forward to any point in time and see what I held at that point.
You can chat with the portfolio now using AI, which is getting pretty good (the AI having the historical snapshots provides an offering you just can’t get from chatGPT). And it’s great to see people are using the app regularly without my really pushing it super hard. I’ve been focused on building out a product I use daily over really ramping up marketing, so I’m happy to see this.
The next big step that is coming for ALPHAPORT is ALPHAPORT RESEARCH. It’s coming along really nicely, but a few things have taken a little extra time to get really good. I’ve said it before on X, but I am highly confident I am going to create a world class app in the InvestTech space at this point, so if you’re interested in talking about that, or trying out the app, now’s a great time to try ALPHAPORT PORTFOLIO. It is easy to setup, and if you have any issues, you can always join our discord channel here: https://discord.gg/ttuUhBnF6S. Or free to just join the channel to talk about investing.
Happy new year!
Martin
Congrats on a great year! Love the analysis.
Great work - just signed up to your pfo tracking tool.
Any views on Snap & Four? Think both great plays for 26