Equity Research @Rebellionaire. Primarily covering Upstart, but deep diving other companies too. Posts are not financial advice. #UTPI
Wisconsin, USA
Joined June 2021
- Tweets19.4K
- Following789
- Followers6.5K
- Likes79.2K
Pinned Tweet
@evansall as a top 3 shareholder in $UMAC, you have our permission @Rebellionaire to order the equipment for the future automated lines today
At $4M/line - why not order 4 right now ($16M) so they are at your door in 4-6 months or however long it takes? Let them sit in a warehouse if you don't yet have the space - at least they are then in the country and ready to be deployed when able.
We understand proof of concept on the first line is important, but we are also okay with a $16M mistake in the pursuit of something much greater - absolute dominating scale in this market vaccum.
Just a thought, we know you're cooking like crazy over there 🧑🍳💜
Henry retweeted
I was talking with @HenryInvests this morning about the market's concerns around debt funding the AI buildout. I literally called it Finance 101 that these companies should be using debt to lever their equity returns. The market is freaking out about one of the most basic tenets of finance.
Goldman posted this note about $CRWV this morning, sounds like we weren't the only ones to notice the obvious!
Henry retweeted
SpaceX's earnings are inflecting now. Tesla's still depend on Robotaxi and Optimus.
Tesla says Robotaxi probably becomes noticeable in its results in Q2 2027.
Matt's view: Holding Tesla potentially has an opportunity cost while AI companies grow.
But AI isn't without its risks. Public pushback on data centers is significant.
What do you think?
Rebellionaire clients and employees own $SPCX and $TSLA stock.
*Timestamps*
00:00 SpaceX's Strong Quarter Needs Some Nuance
00:50 Elon: Highly Confident of 10 Gigawatts
09:58 Tesla Still Needs Robotaxi or Optimus
11:58 Tesla's Robotaxi Payoff Moves to 2027
16:01 Optimus Could Be Two Years Out
21:54 SpaceX Has the Larger Catalysts
23:08 SpaceX and Tesla Both Must Execute
@nerdalert
$SPCX $TSLA
Opinions needed please 🙏
Have been working on the $UPST 3Q26 Earnings Preview, should have this out next week Friday. TLDR: I think this quarter will be fine, the guidance is more dependent on 4Q26.
Trying to decide how I'm going to do coverage this time with the change in format. Earnings release 7am ET (that's 6am for me!) followed by a video webcast at 8am ET.
Currently, I'm thinking about doing a live X space for when the results drop at 7am ET. However, I don't think we should listen to the call together anymore :( - it makes the most sense for everyone to tune into the video webcast Upstart is hosting directly. (X Spaces are audio only, video is the clear draw this quarter).
Which of the following options would you prefer for coverage?
1) X Space for the Live Earnings Release Only (6:55-7:35am ET)
2) X Space for the Release (6:55-7:35am ET) plus a 2nd Space AFTER the webcast (9:05-9:40am ET)
3) No X Space - coverage via posts on X & traditional earnings review, plus a potential YouTube video.
Henry retweeted
Unusual Machines $UMAC lands $16.9M drone component order from PowerUS
stocktitan.net/sec-filings/U…
Henry retweeted
This video has aged very well. We will always be willing to change our mind when new information comes to light.
But the harder thing can be to actually make a change in the portfolio when the relative opportunities have shifted. Rebellionaire helps clients make those tough decisions in real time.
What are Rebellionaire's SpaceX vs Tesla Growth Expectations?
SpaceX could be generating 1.5x Tesla's profits by Q1 2027, with 6x the revenue growth.
And by Q1 2028, SpaceX could see ~600% profit growth.
This is bull-case Tesla vs base-case SpaceX and it changes everything about the merger.
The numbers shocked us.
What do you think?
*Timestamps*
0:00 Intro
1:10 Tesla Q1 2027 Bull Case
5:03 SpaceX Q1 2027 Projections
9:28 Tesla vs. SpaceX: The Problem
11:58 Wall Street's SpaceX Blind Spot
15:42 SemiAnalysis & 2028 Outlook
17:19 Changing Views on the Merger
21:21 Still Tesla Fans
Data centers & #AI have a massive PR problem🛑
Memphis, Tennessee is the latest battleground. The City Council just passed a 135-day PAUSE on new data center development.
The catch? Locals wanted a 12–18-month moratorium - which City Council struck down.
Catch all this and more in under 6 minutes @Rebellionaire 👇
Local opposition is becoming a real bottleneck for AI data centers.
Memphis shows why.
On October 20 its city council tied 6-6 on an extended moratorium but passed a 135-day pause on new construction.
Henry's view: if city councils side with the communities, it could be a real risk to the data center build-out.
What do you think?
*Timestamps*
0:00 Local Opposition Is a Major Bottleneck
0:40 SpaceX's Colossus Cluster in Memphis
1:27 Memphis Weighs a Moratorium
2:10 Chaos at City Hall
2:43 Residents' Concerns Are Reasonable
3:40 The Moratorium Fails on a Tie
4:23 Memphis Passes a 135-Day Pause
4:46 Local Opposition Is a Risk
@HenryInvests
$SPCX
Rebellionaire clients and employees own SPCX and NVDA stock.
Henry retweeted
Local opposition is becoming a real bottleneck for AI data centers.
Memphis shows why.
On October 20 its city council tied 6-6 on an extended moratorium but passed a 135-day pause on new construction.
Henry's view: if city councils side with the communities, it could be a real risk to the data center build-out.
What do you think?
*Timestamps*
0:00 Local Opposition Is a Major Bottleneck
0:40 SpaceX's Colossus Cluster in Memphis
1:27 Memphis Weighs a Moratorium
2:10 Chaos at City Hall
2:43 Residents' Concerns Are Reasonable
3:40 The Moratorium Fails on a Tie
4:23 Memphis Passes a 135-Day Pause
4:46 Local Opposition Is a Risk
@HenryInvests
$SPCX
Rebellionaire clients and employees own SPCX and NVDA stock.
Henry retweeted
Four of the nine Gauntlet II winners are Unusual Machines $UMAC customers
Drone Dominance Program Gauntlet II orders are in the works.
Perennial Autonomy received the top award for the Deep Strike mission area, and Neros Technologies took top marks in the Close Quarters Battle mission area. Each received the maximum initial order of 8,000 units.
Other top performers included Hyperscale, Skycutter, Swarm Defense Technologies, ORQA US, XTEND Reality, Vector and ModalAI. Each received orders from the 60,000 pool based on performance.
Congratulations to all the winners. Now, it’s time to deliver.
View the full leaderboard here: dronedominance.mil/leaderboa…
$SPCX
Let's say SpaceX truly has ~15GW online end of 2027.
Pricing: External "slightly above high end of $30-$50 per Watt" - call it $52 per Watt per Year on external compute today, ~$55 per Watt YE27 (strong demand, compute constrained).
External compute is ~50% of the mix, other half used for Internal which let's assume monetizes at $25.0 per Watt per Year as Grok / Grok Bot scale by YE27 (Still less than half the return of selling external compute).
That's a blended ~$40.0 per Watt per Year at YE27 (($55 x 0.5) + ($25 x 0.5)). In other words- $40B in revenue per GW.
15 GW x $40B per GW = $600B annual revenue.
🤷♂️ the math works, don't even ask me about 20GW!
How much demand for AI hardware / optics / networking does SpaceX $SPCX create alone if they bring online 13 GW of compute next year?
Elon is highly confident of this. OpenAI and Anthropic are expected to add 14 to 15 GW each in 2027. SpaceX right up there.
Demand is insane.
ELON MUSK TARGETS 13 GW OF AI COMPUTE BY 2027, WITH 440,000 NVIDIA VERA RUBIN GPUS IN A SINGLE TRAINING CLUSTER
Musk says SpaceX is on track to bring roughly 200,000 $NVDA GB300 GPUs online per month for the remainder of 2026, with deployments continuing at a similar pace next year, primarily using Vera Rubin.
The company expects to have 2.5-3 GW of compute online by the end of 2026, with another 10 GW potentially coming online in 2027.
Musk also plans to have a 440,000-GPU Vera Rubin training cluster operational by the end of Q2 2027.
To support the expansion, SpaceX is targeting an additional 20 GW of power and cooling infrastructure next year.
Musk said there is a “non-zero possibility” of exceeding 20 GW of operational AI compute in 2027, although he acknowledged the probability remains small.
Upstart $UPST 8-K: Up to $200M Revolver to fund initial capitalization of Upstart Bank.
~$200M or so will be used to open Upstart Bank, purchase loans from Upstart Network, who will then take that $ and pay off the revolver. Priced at SOFR + 300bps.
@paulxgu
ir.upstart.com/static-files/…
Rumble's $RUM pivot to AI Compute is (surprisingly) very real - they've signed a $13.7B deal with Anthropic.
@Rebellionaire took a deep look at Rumble back in August and thought they had a great chance at landing a large deal with a hyperscaler.
We didn't end up investing, but the research process here was fascinating and thorough. What an interesting business!
Rumble's pivot to AI compute looks real.
So does the bill for building it.
In our best case, its 2027 revenue could reach about three times Wall Street's August estimate.
We still didn't invest.
Henry's view: Rumble may need to raise billions to build it out, a lot for a company worth about $3 billion.
What do you think?
*Timestamps*
0:00 Rumble Is Pivoting to AI Compute
1:22 Free Speech Meant Owning Data Centers
2:45 Tether Paired Rumble With Northern Data
5:46 Quake AI's Revenue Math Is Aggressive
7:19 Best Case Near 3x Wall Street
8:35 If Rumble Traded Like a Neocloud
9:30 Why We Passed: Funding the Build-Out
@HenryInvests
$RUM
@chrispavlovski check us out... this is pretty good coverage on $RUM. Always open to a Q&A chat!