저자: 미상 · 외국인
요약 보기
채널 진행자는 드러켄밀러의 비트코인 채굴주 투자(약 1.25억 달러)를 비트코인 베팅이 아니라 AI·데이터센터 전력 인프라 베팅으로 해석한다. 하루 동안 채굴주(IREN 약 +7%, HUT8 +6%, BITDEER +4%, RIOT +3%)가 비트코인(-2.4%)보다 크게 올라 약 10%p 괴리가 났고, 이를 전력 엔진 가치의 측정치로 본다. 마이크로스트래티지(-1.4%)는 대조군으로 제시. 진행자는 IREN 매수, HUT8 매수(소액), BITDEER 관망, RIOT 패스를 언급하며 비트코인 직접 보유보다 IREN 같은 인프라형 채굴주를 선호한다. 다만 고변동성 리스크와 재동조(비트코인 급락 시 동반 하락) 가능성을 명시한다.
다듬은 글
Everybody's telling you smart money is rotating out of AI and into Bitcoin. And I think they've got it exactly backwards. And today's tape actually proves it. Five names this actually matters for and by the end you know which one I'll own and which two I wouldn't touch. If this is helpful and this is useful, hit the like button and subscribe.
It's free and it generally help us grow the channel. Right, let's get into it. Here's my read plain. Druckenmiller's 125 million minor basket isn't a Bitcoin bet at all. It's an AI power and data center bet that happens to be wearing crypto tickers and the two trades have quietly divorced. Now, start with the thing that made me sit up because it isn't the following, it's today.
Four Bitcoin miners, the exact four in that basket, all closed green as I'm recording this on September 6th on a Sunday. I ran up more than 7%, Hut 8 up over 6%, Bit Deer up about 4% and Riot up around 3%. Same session, Bitcoin itself was down roughly about 2.4% as per Friday when the market closed.
Sit with that for a second. That's a 10-point spread between I ran and Bitcoin in a single day. And this is a group of stocks the entire market describes as a leverage bet on the coin. That's not supposed to happen. Miners are meant to be Bitcoin with the volume turned up. Coin goes down, miners go down harder.
Instead, the coin bled and the miners ripped. So, let's set that table properly and only with what actually matter on paper. Stanley Druckenmiller's recent 13F filing, his family office to Canaan, disclosed roughly 125 million spread across four Bitcoin mining stocks. We have Bit Deer, Hut 8, Riot, and I ran.
Not spot Bitcoin, not the ETF, the miners. Separately, Jane Street's filing show around 1 billion sitting in Bitcoin ETFs and Paul Tudor Jones bumped his ETF position around 19% and look at what that actually tells you because everyone reads it as one story. It's actually two story. Some of the smart money bought Bitcoin.
Yes, that's undeniable. Some of the smart money also bought miners. Those aren't the same trade and the crowd calling all of it the rotation into crypto is simply lumping together two things that have behaved completely differently, which is exactly what today's tape screamed. So, here's the one idea I want you walking out with because once you've got it, you will never look at these Bitcoin miners the same way again.
A miner's share price is two engines bolted into one ticker. Engine one is Bitcoin beta. That's the obvious one. They mine coins, they hold coins on the balance sheet, and the money they make per unit of computing power rises and falls with the coin prices. Bitcoin up, that engine revs. Engine two is power and compute beta, and this is the part most people skip.
These companies spent a decade doing one unglamorous thing, hunting cheap electricity. Substations, grid connections, land nobody wanted, sign contracts on megawatts. That's the single scarcest input in the whole entire AI build-out right now. You can order chips, you cannot order a grid connection and have it show up in 18 months.
So, a miner can take that same warehouse full of power and rent it out as AI and high-performance computing capacity. Steady contractor revenue paid in dollars and none of it depends on what Bitcoin does next. Now, here's why this matters for you rather than being a fun fact. On a normal day, those two engines move together and you can't tell them apart.
Bitcoin rallies, miners rally. Everyone nods and says leverage crypto play and moves on. But, on a day like today, when Bitcoin engine is clearly down, even ever so slightly, and the stock is up 7%, the market's handing you a clean measurement. The spread between those two numbers is the power engine value on its own with the crypto noise stripped out.
That's the whole method. You're not guessing which engine is driving, you're reading it off the tape. And when you run today's names through it, it's almost too tidy. I ran up over 7% against Bitcoin like down 2.4%. That's the widest spread in the group. Hut 8 up 6% right behind it. Those two carry the biggest contracted compute component of the bunch and they posted the biggest gaps.
Bit Deer up about 4% real spread but thinner. And that fits because its mix leans towards the hardware and the mining side. Riot up 3%. That one looked to me like it got dragged along for the ride rather than value on its own merits. And then the control experiment, MicroStrategy, pure Bitcoin exposure, a balance sheet stuffed with coin and precisely zero megawatts of contracted power, fell about 1.4% on the exact day the miners ripped.
Bitcoin engine only. Bitcoin engine was down. Stock was down. Exactly what the decomposition predicts. I had to check that lineup twice. Honestly, you rarely get a market this obliging about proving a point. Quick and necessary, none of this is obviously financial advice. I don't know your account. I don't know your timeline.
I don't know how much your net worth is already sitting in crypto adjacent stuff. This is how I'm reading the tape and you do your own work from here. So, why is the power engine getting bid right now this week? Because the news driving these names today isn't Bitcoin news. It's infrastructure news.
Recent coverage of the move points at AI and data center tie-ups at the minor level, Hut 8 and Iris specifically, as the reason the group caught a bid. That's the part that flips the popular story on its head. The headline everyone's writing is smart money rotating out of AI into Bitcoin. What the tape says is completely opposite.
They're rotated further into AI. They just did it through tickers that still have the word mining attached. And you'll see a lot of takes this week built on a whale wallet charts and ETF flow screenshots. I don't trade off those. I trade off price. And price handed me a 10-point spread today. That's a measurement, not a vibe.
Now, you might be sitting there thinking, "I don't own crypto. Why is this my problem?" Fair enough. Here's the bridge. Forget the coin entirely. What you're looking at is a group of companies sitting on one thing that AI build out is generally short of, power that's already connected. That's an industrial story in a crypto costume and it's available in a normal brokerage account.
Before the verdicts, the onus on the side because it's a good argument. One session spread is noise, not a divorce. Miners have detached from Bitcoin for a few days plenty of times and always snap back. And if Bitcoin generally breaks down hard, the mar ket has a habit of remembering these are crypto companies with crypto balance sheet and crypto financing costs.
AI deals or not. Plus, that Druckenmiller filing is a snapshot from a quarter ago. He could be out already. All true. Here's why I still land where I land. I'm not trading the filing. I'm not trading one candle. I'm trading the mechanism the filing pointed me at and the tip that measure. Contracted power is a real data dollar denominated asset and it doesn't stop existing because the coin has a bad week.
The 13F filing told me where to look. The spread told me is real. So, I ran is the one I'm a buyer of. Chart looks pretty good if you can have a good evaluation point and it's the cleanest expression of this whole idea while the spread in the group and the work is being done by contracted capacity, not by what a coin did overnight.
But size it like what it is. This is an AI infrastructure position, not a Bitcoin proxy. And companies like Iron when it comes to trading can be very risky. It is a high beta move. So, make sure you know exactly where to put your stop. If you size it like a crypto punt, you'll manage it like one and you will get shaken out of a perfectly good industrial trade.
I've made that mistake before with a stock right thesis, wrong label in my own head and I sold it for the wrong reason, of course, taking a loss at the same time. Okay, now let's focus on the names. The other name I want to bring up is HUT 8, also under Stanley Druckenmiller's portfolio. Up 6% on a data center tied up news, second cleanest power engine in the group and it is also in the Duquesne basket.
So, it's confirming the thesis rather than just riding the group higher. I'm a buyer there as well. Smaller size than Iris. Obviously it's a lot more riskier. Bit deer up about 4%. It's in the basket and the spread is real, but the mix is heavier on hardware and mining, which means more of the stock is still tied to the coin.
That one I'm watching rather than buying and it's one I'm tracking closely for the trading desk, which is my free Telegram channel, is Riot. Up about 3% is the controlled name in this experiment. That looks like sympathy to me. It's It went up because its neighbor went up. Not because anyone took a fresh look at its power book.
I'll pass at these levels. And MicroStrategy down 1.4% while the group ripped. There's nothing wrong with that. It's just doing exactly what it's built to do, which is be Bitcoin with extra steps. If you want Bitcoin exposure, I would rather just buy Bitcoin. I'll pass on the wrapper. If you want the exact level I'm acting on for Iron and the size I'll carry it at given its infrastructure position and not a crypto one.
That's in the trading desk. My Telegram channel link is in the description. Go ahead and join. Cost you nothing, right? And I'll let you know if I enter any of these positions because there are a few that I'm looking at in terms of the chart. Uh Bit deer looks good as I'm actually recording this. It looks like there's an inverse head and shoulder going on.
And then we have Iron, which is which is uh forming quite a nice pattern. We have a higher highs and a higher low. And then we have uh you know, the other ones The chart is also looking quite significant. If you guys want me to do a live charting session as well, comment in the description. Let me know.
Just type in live chart or live charting and I'll be more than happy to tune on live on YouTube and do a live charting together. And now, here's what would make me wrong and I want you watching for it as hard as I am. The recoupling test. If we get a week where Bitcoin sells off properly and Iron falls along with it at a multiple of the move, the spread's gone.
Back to being leveraged crypto, then the power engine was never being valued separately and I'm just long Bitcoin without admitting it. That's the tell and it's binary. I'm also wrong if those AI and data center deals stall, get marked down, or need a big equity raise to fund because then they're diluting me to build the very capacity I'm paying for.
So, the walk away here, the rotation everyone's describing isn't AI money going into Bitcoin. I think that we we we identify that in this video. It's Bitcoin infrastructure quietly getting treated as AI infrastructure and the market told us today by sending four miners up while the coin went down. And the rule of thumb, the one worth repeating to a friend, when a stock has two engines, wait for the day they disagree.
That's the only day you'll find out which one is really driving. Subscribe to this channel if this was worth the time, hit the like and the levels and the sizing are all in my Telegram channel. That's the desk of today. I'll see you in the next video.
Comments 0