1bsar.github.io/BTBT-WYFI-NAV/
So I've been doing a ton of research on AI data center plays and Ive come across this and I want to lay it all out because I genuinely think this is one of the better setups I've seen in a while. I also built a full interactive NAV model on this which I'll link above.
So WYFI (WhiteFiber) is an AI data center company that IPO'd last August. They basically take old industrial buildings with existing power infrastructure and convert them into AI data centers. Their current big project is a huge textile mill in Madison, North Carolina that they're converting for AI workloads. The genius of the model is they skip the hardest part of building a data center which is getting power: the building already has it!
Now BTBT (Bit Digital) owns about 27 million shares of WYFI (They actually were what WYFI is now but spin off the "AI" part of their company into creating WYFI). That's 70.5% of the whole company. And their CEO has publicly said they're not selling a single share through all of 2026. WYFI only has about 11.3 million shares actually trading in the public float. So when demand picks up, there's basically nothing to buy. That's a squeeze setup right there (not the main thing I'm looking at).
On top of the WYFI stake, BTBT also holds around 155,000 ETH. So you're getting two assets in one ticker.
May 14th is the date
Three things are happening basically simultaneously on May 14th:
WYFI reports earnings pre-market. Same day, BTBT reports after close. And Cerebras, which is literally WYFI's anchor customer at their Montreal data center, is pricing their IPO the night before and starts trading same week.
Cerebras has been paying WYFI roughly CAD 1.4 million a month since November. They just announced a massive partnership with OpenAI for inference capacity. So the company that's literally paying WYFI's bills is about to go public with a ton of hype around it the exact same day WYFI tells the market how much money they made. That timing is not a coincidence, or maybe it is, but either way it's a pretty clean setup.
Then on top of that, WYFI has a $865 million 10-year contract with a company called Nscale at the NC-1 facility. Billing starts June 2026. So earnings guidance should speak directly to that ramp and how it's tracking.
IREN just signed a crazy deal with Nvidia. AMD earnings were strong. CapEx across the board is growing. The AI infrastructure space is genuinely one of the hottest things going right now and WYFI sits right in the middle of it. I really don't see a scenario where WYFI goes back to 52-week lows given everything happening in this space. Worst realistic case if earnings disappoint is maybe we drift back toward IPO price around $15. That's kind of the floor in my head.
(Looking at it now, probably should take into account the US-IRAN deal being rejected. Could possibly mess with overall market sentiment.)
Now the fun stuff:
So I actually built out a full net asset value model for BTBT based on their SEC filings, BTBT 10-K, WYFI 10-K, and the WYFI 8-K from January where they issued $230M in convertible notes. You can play with it here: 1bsar.github.io/BTBT-WYFI-NAV/
The model basically says: take BTBT's WYFI stake at whatever WYFI's current price is, add their ETH treasury, add cash, subtract WYFI's debt, divide by BTBT's share count and you get the net NAV per share. Then you apply whatever discount the market typically gives holding companies like this.
Here's the interesting part. Looking back historically, when WYFI hit its 52-week low on March 27th at $10.51 and ETH was around $1,991, BTBT was trading at roughly $1.30. Plug those numbers into the model and BTBT was trading at basically zero discount to NAV. Fair value.
Then around October when WYFI was near its highs around $40 and ETH was around $4,250, BTBT was trading around $4. That implies about a 20% discount to NAV. So even at peak conditions the market was applying a 20% haircut, probably just from the complexity of the structure and normal holding company friction.
With WYFI at $21.58 (when I made the model) and ETH around $2,326, BTBT at $1.80 also implies roughly a 20% discount. So the discount isn't unusually wide right now, it's actually already at its historically tight level. What that means is the upside isn't really about the discount compressing, it's purely about WYFI's price going up on earnings and lifting the NAV that the 20% is applied to.
Scenarios:
Worst case, WYFI drops back to $15, ETH stays flat, discount tightens to around 10% because the stock is falling and the market historically prices it closer to fair value on the way down. You're looking at maybe a 14-20% loss. That requires basically everything going wrong at once.
Base case, WYFI hits $30 on good guidance, ETH maybe nudges up to $2,500, 20% discount holds. That's roughly a 40-42% gain on BTBT.
Bull case, WYFI pushes toward $35, maybe another contract gets announced, ETH stays around $2,500. At 20% discount that's about a 60% move. If the discount somehow compresses to 10% you're looking at 80%.
I'll be updating this model live after earnings drop on May 14 using the new 10-Q numbers, so if WYFI reacts big in either direction I can recalculate what BTBT should theoretically be worth in real time.
Now probably what most of you may be asking: Why BTBT over WYFI directly?
You could just buy WYFI. But BTBT gives you leverage to WYFI through the thin float dynamic plus you get the ETH treasury basically for free. WYFI has already moved 63% in the last month. BTBT hasn't caught up nearly as much. That gap is what I see as the opportunity.
The asymmetry here is what makes this interesting to me. Downside is capped by the fact that the whole sector is on fire and NAV support kicks in on the way down. Upside is a specific catalyst on a known date with three separate drivers stacked on top of each other. I'm not saying this is a guaranteed win, nothing is of course, but the setup is pretty clean and the research seems to back it up.
Posted by _ibsar