GM. This is Milk Road PRO, where I finally learned how to rotate into a completely new sector and hope I didn’t wait too long to do it.
Today's edition features our latest PRO report on how M0xt rotated his portfolio out of crypto and into AI, fintech, and a few mispriced category leaders.
Here's what we've got for you today:
- 🔄 Why his crypto exposure dropped from 49% to 25% in 90 days, and where that capital actually went.
- 🎯 Why Sortino (not Sharpe) is the right scoreboard for asymmetric portfolios.
- 🛒 His last five buys, all sitting below 0.5 correlation with Bitcoin.
- 👀 The four signals being watched next, including what would trigger adding crypto back.
Nexo is back in the U.S. - and new clients get 30 days of Wealth Club Premier perks! Higher yields, lower borrowing rates, and crypto cashback - start here.
Prices as of 2:00 p.m. ET. Powered by Coingecko.
WHY I'M ONLY 25% CRYPTO NOW
Before we start, I (M0xt) want you to remember that these are my own views. Other analysts in Milk Road, like John Gillen, are very bullish and very allocated to crypto - you can check out their portfolios (and mine) here in Milk Road PRO, which is only a buck for a seven-day trial.
Now, let’s get to my portfolio:
- Open P&L: +35.9%.
- Max drawdown: 8.4%.
- Crypto exposure: down from 49% to 25%.
- Realized profit on closed positions: $5,781.
For the last few years, I was largely exposed to crypto. When Bitcoin moved, my alts would rip. This worked well.
Even just three months ago, I was still 49% allocated to crypto. My only AI exposure was Tesla.
My strategy has now changed.
In Q1, equities ran to fresh all-time highs while crypto sat flat. AI infrastructure companies compounded revenue faster than anything onchain, and I owned almost none of them.
So I rebuilt my book. Crypto is now 25%. Cash is 16%. The other 59% sits in AI infrastructure, fintech, and a few mispriced category leaders like Uber.
The rotation is still a work in progress. But 90 days in, the lessons I’ve learned are clear: stop optimizing for returns alone, stop betting on one market, and start diversifying without sacrificing the asymmetric upside.
The rest of this report is how I'm doing that, what's working, and the one number I use to track it.
THE SIGNAL I COULDN'T IGNORE
For most of the last cycle, crypto led risk assets up and down. Bitcoin moved first, equities followed. That relationship is what made an overweight crypto book defensible: owning crypto meant owning the whole risk complex by proxy.
The signal started breaking in Q4 last year and became clear in Q1.
Once the Iran-U.S. tension “eased” in early April, the Nasdaq and S&P 500 ran to fresh all-time highs over multiple weeks. Bitcoin stayed down across the same window. The textbook said crypto should have led the move, yet it didn't even participate.
Whatever you want to call this (regime change, decoupling, rotation), the practical consequence is the same: the asset class I was overweight on had stopped doing the job I was relying on it to do.
The AI side of the trade was even more direct.
I was using AI tools every day in my own work, the way I used Google ten years ago. I covered the macro view of this market in my last report on AI here. The point that mattered for my portfolio was simpler: the companies serving this market were compounding revenue faster than anything onchain. And I owned almost none of them.
So the question stopped being "is crypto going to rip?" and started being "what am I doing about the AI book I should already have?"
THEN VS NOW
Here's the simplest version of what changed: crypto, cash, and stocks, at the start of the portfolio versus today.
Source: Milk Road PRO
The headline rotation: crypto dropped from 49% to 25%, stocks rose from 48% to 59%, cash went from near zero to 16%. I'm holding the cash on purpose, not because I'm bearish or missed a bid (more on that in the drawdown section).
Moving capital from tokens into stocks doesn't always change my exposure. Sometimes it only changes the type of exposure. The "stocks" bucket includes Coinbase, Galaxy, and Robinhood, which still trade with crypto beta.
Source: Milk Road PRO
Combine tokens with crypto-adjacent equities and my real crypto exposure is 52.6%. Down from 78% at inception, but still the largest driver of portfolio performance.
The right question for every position is what it's exposed to, not what asset class label it carries.
WHAT CORRELATION ACTUALLY SHOWS
Galaxy (GLXY) and Coinbase (COIN) correlate more tightly with BTC than some of my native crypto tokens, including Sky (SKY) and Hyperliquid (HYPE).
Source: Milk Road PRO
That's the clearest proof that "stock" labels hide crypto exposure.
But the issue is that even after the rotation, my portfolio-level correlation with Bitcoin is 0.84. When BTC has a big day up, so do I. When BTC dumps, I dump with it. That's one risk factor driving most of the book.
Which is the problem I’m still trying to solve, but not by selling everything correlated.
CRYPTO SHOULD WORK HARDER FOR YOU
Most people hold crypto and hope.
The smart money? They're earning interest on it, borrowing against it without selling, and trading it.
Where can you do the same all in one place? Nexo.
And right now, new U.S. clients get 30 days of Wealth Club Premier (benefits normally reserved for loyalty program members):
- Enhanced interest rates on your digital assets
- Lower borrowing costs against your crypto
- Up to 0.5% cashback on trades
No need to sell to access liquidity. No juggling 5 different platforms.
*Disclaimer: Geographic restrictions and terms apply.
THREE POSITIONS, THREE JOBS
The framework isn't here to minimize crypto correlation. It's to understand what each correlated position is actually paying me to hold. If the BTC factor is doing work I want, I keep it. If it's doing work I don't want, I cut it.
The aggressive version of this report dumps anything above 0.5 correlation and calls it a day. That version also sells positions about to decorrelate mechanically (Galaxy) and positions that pay me for the exposure (BTC itself). My top three explain why neither got cut.
#1: Galaxy (GLXY) - 17.4% of my portfolio
Galaxy trades like a crypto stock today because most of its revenue is still crypto.
The AI data center business comes online this quarter, and once $200M of annualized revenue starts hitting the income statement, the cash flow profile will stop looking like a crypto proxy and start looking more like a modern business.
A 90-day correlation reflects past revenue mix, not next quarter's. The market is pricing the current 0.84 as permanent. I'm holding because I expect it to break and re-rate aggressively.
#2: Bitcoin (BTC) - 14.5% of my portfolio
Why is BTC still #2 after a whole report on hidden crypto exposure? Because hidden exposure is what I'm cutting, not crypto itself.
I'm carrying crypto beta on purpose, and I want it clean: direct, on the balance sheet, in the asset I actually want exposure to. Not laundered through equities that stack execution risk, dilution risk, and regulatory overhang on top of crypto risk. Not in tokens with open questions on their tokenomics.
There's also a benchmark argument. Every correlation in the chart above is measured against BTC, so holding it means I own the literal index of the asset class I'm trying to be deliberate about. BTC earned the slot. Everything that didn't earn its slot is gone.
#3: SK Square (402340.KS) - 11.4% of my portfolio, -0.09 correlation
SK Square does a job none of my other positions can: it doesn't move with Bitcoin. Every dollar in it is a dollar of portfolio risk that doesn't depend on what happens to crypto. When BTC dumped 15% twice this year, this position was unaffected.
To lower portfolio correlation, I have two choices: sell crypto-correlated names I still believe in, or add genuinely decorrelated winners on the other side. The second move is almost always cheaper, because it doesn't force me to exit a thesis just to lower a number.
The 16% cash position does the same work and earns 4% doing it. SK Square earns a return and decorrelates at the same time.
THE RIGHT SCOREBOARD
Two portfolios that both did 30% in a year aren't the same portfolio. One halved on the way and clawed back. The other ground higher with a 10% worst month. Same number on the scoreboard, different lives.
For most of my career, I optimized for the top number: annual return, year-to-date. The rotation, the three positions, the 16% cash level. None of those decisions make sense in that old system. I needed a different scoreboard.
The institutional default is the Sharpe ratio: return divided by total volatility. The problem is that Sharpe treats a 40% rip and a 40% drawdown as equally bad. For a book whose whole game is asymmetric upside in crypto and AI, that's backwards. I want the upside vol. I just don't want to give it back.
What I use instead is the Sortino ratio. Same idea as Sharpe, with one change that matters: Sortino only counts downside volatility. The upside swings don't count against you. So a portfolio with five +20% months and one -10% month gets punished by Sharpe and rewarded by Sortino. That's the asymmetry I actually care about.
Quants tend to prefer Sortino for asymmetric strategies and rarely put it on a tear sheet. Sharpe sells funds - whereas Sortino runs them.
My book on Sortino
The portfolio is sitting at a Sortino of 5.45. For context, above 3.0 over a long window is rare, and above 5.0 over a short window usually means a great regime, a lucky one, or a small sample. That number should make you skeptical.
90 days is a short window, most of it has been an uptrend, and the 8.4% max drawdown means this book hasn't been hit by a real BTC dump yet. At 0.84 correlation to BTC, when one comes, it will sting. I expect the number to come down as the window lengthens, and the honest version of this report 12 months from now will tell us how much of the 5.45 was real and how much was conditions.
The number that does hold up is the vol split: 41.7% total volatility against 23.2% downside volatility. Almost half of my portfolio's swing is upside, which is exactly the swing I'm not trying to suppress.
Why this changes the whole report
Once Sortino is on the scoreboard, every position in this report is the same trade: cut downside vol, keep asymmetric upside. The crypto rotation does that. The 16% cash does it as the highest-Sortino line in the book. SK Square does it by lowering the downside number without forcing me to exit a thesis. Galaxy does it on a forward basis, because its correlation is about to drop and its Sortino contribution is about to rise.
MY LAST FIVE BUYS
And a quick reminder before I get into it: You can see my full portfolio and all my moves in Milk Road PRO, which is only a buck to check out for a week. You can even join the PRO Discord and ask me questions about today’s report!
The framework is only useful if it changes what I actually do. Here are the five most recent positions I've added or sized up:
Source: Milk Road PRO
1. SK Square (402340.KS), -0.09 correlation with BTC. AI-memory leader, completely unrelated to crypto. Holds a major stake in SK Hynix, which gives me HBM exposure without paying the Nvidia multiple.
2. Klarna (KLAR), 0.36 correlation. Recently IPO'd, finally profitable, and trading like the market still thinks it's a 2021 fintech. The re-rate happens when the next two quarters confirm the model works.
3. Figure (FIGR), 0.46 correlation. Expanding from HELOC issuer into a loan marketplace. That moves the business from a balance-sheet originator to a higher-multiple platform, and the market is only starting to price it in.
4. Nu Holdings (NU), 0.39 correlation. Brazil's dominant digital bank is expanding into Mexico and Colombia with the same playbook. Earnings still compounding at 40%+, and the stock doesn’t trade like a U.S. fintech multiple in a market that should be priced higher.
5. Uber (UBER), 0.23 correlation. Cash-flow machine and consumer platform with a free option on autonomous vehicles. The base business compounds. The AV layer is the asymmetric upside that no one's paying for yet.
All five sit below 0.5 correlation with BTC. Three are non-U.S.-listed. None are crypto equities or tokens. Six months ago, this list would have been three tokens and two crypto equities. Today it's the opposite mix, on purpose.
I don't size positions because of where the ratio sits. I size them by what each one is exposed to, what role it plays, and whether I can find something genuinely uncorrelated to add next.
What I'm watching, and what each signal would do to the portfolio:
- Crypto vs equities correlation. If crypto recouples and starts leading risk-on again, the Sortino case for rotating weakens and I add back to crypto.
- Galaxy's AI revenue. If the revenue lands and the correlation stays stuck at 0.84, the thesis was wrong and I cut.
- The next BTC drawdown. If it takes the book down more than 12-15%, the construction isn't doing what I claim and the framework needs another pass.
- The ongoing screen for uncorrelated names. As long as it keeps surfacing real businesses on the other side of the BTC map, I keep rotating. When the supply dries up, the rotation slows.
Three months ago, this portfolio was heavily overexposed to crypto. It is not the same today. The job from here is to keep the upside intact and stop the downside from compounding.
The question for your own book
A lot of readers feel this rotation is something they should do, but they hesitate for one reason: what if crypto finally picks up?
It can absolutely happen. But the other question rarely gets asked: what if it doesn't?
Picture a portfolio where 30% of performance is driven by crypto, 30% by AI, and the rest by everything else. What if crypto stays flat and the rest of the market doubles? How are you going to feel about that? And if you're sure crypto outperforms from here, what's the actual reason?
One green candle on BTC and crypto is in the spotlight again. If that happens, great, 30% of your holdings are right there with it. And if it doesn't happen, you are still fine. That's the whole point.
I'm open to adding crypto exposure back when the reasons show up. Right now, the music is playing somewhere else, and I need to adapt. Markets are moving faster than ever. The investors who are slow to adapt will have huge opportunity costs.
I'm not that kind of investor. So I react.
AI-GENERATED PODCAST 🤖
We’ve turned this PRO report into an AI-generated podcast to make it even easier to digest. You'll find the audio player below. 👇️🎧️
How I beat crypto while rotating away from it
Disclaimer: This podcast was created using AI and is based on the research report above. While we've done our best to ensure accuracy, the audio may contain minor errors, technical glitches, or mispronunciations. Please note that this podcast provides an overview of the report and is not a comprehensive or definitive take on the topic.
BITE-SIZED COOKIES FOR THE ROAD 🍪
Pre-internet money rails meet the agentic economy? Yikes. Arc is Circle’s open Layer-1 built for the internet-native economy. Humans and AI welcome.*
NEAR co-founder: The crypto market is entering its consolidation phase, and only a few will survive.
Huge if true / approved: U.S. and Iranian negotiators have reached an agreement on a 60-day memorandum of understanding to extend the ceasefire.
$166B in tariff refunds: Who actually gets paid?
Special offer for Milk Road readers. Kalshi is giving Milk Road readers $10 free when you trade $10 on their platform.*
*this is sponsored content.
MILKY MEMES 🤣
ROADIE REVIEW OF THE DAY 🥛
VITALIK PIC OF THE DAY
