The Chairman's Ledger

New account, so here’s the short version. I started at JPM. Built a company. Sold it for $100M+. Served in places where risk was not theoretical. Now I spend my time looking for public companies sitting at the intersection of capital, technology, and strategic necessity. Defense tech. AI infrastructure. Space. Bitcoin. Critical infrastructure. I am not here to post 50 tickers and celebrate the ones that work. I am here to find the few names where the market is using the wrong frame. Old category. New asset. Messy transition. Right team. Massive demand pull. That is where the asymmetry usually lives.

@Spykex02 Absolutely not.

@ChairmansLedger Did you trim at all after the rips with onds and nuai today?

$IREN just signed $2.8bn of new AI contracts, raised its YE26 AI Cloud ARR target to $4bn+ and now has ~85% contracted. Customers include $MSFT, $NVDA, Perplexity, Figure AI, Together AI, Fluidstack, Fireworks AI, Fal AI, Hume AI and another leading AI developer. They’re prepaying ~45% of associated GPU capex while demand still exceeds available and planned capacity. Tell me again compute isn’t scarce. CC: @JaradCapital for the intel.

@jiahanjimliu $IREN D-Day. 🫡

$IREN: New Enterprise Customers and Raised 2026 ARR to 4B+ Increase ARR Explained Previously, for 2026 GPUs, PG and H1-4 were signed while Mackenzie and Aircooled Childress was unsigned. IREN has signed partial or all of Mackenzie and the contract values came in higher than expected so 2026 ARR increased. Previously expected contribution from Mackenzie to 2026 ARR was 1.3B and but now with just in time contracting, Mackenzie has 1.6B+ 2026 ARR contribution. IREN's strategy of signing contracts as late as possible when GPUs arrive allow them to have the highest GPU price possible as GPU price has been increasing. Mackenzie is 80MW and the first shipments of GPUs to Mackenzie have started to arrive. IREN has only signed 85% of it's 2026 GPUs with 85% skewed by an overweighted amount going to MSFT. The remaining 15%, once signed, will likely result in another upgrade in 2026 ARR of ~200m. Combined with Nvidia contract next year, 4.4B ARR translates to 4.7B ARR for early next year contracted. With only 480MW buildout this year, the 730MW next year will be mostly VR which will be contracted at 50%+ higher per MW. With IREN's 2027 total active power of 1210MW and the majority being VR, ARR will likely reach 15B+ by end of 2027. Enterprise Customers - Figure AI Critics have wrongly assumed IREN's bare metal could only serve hyperscalers and AI Platforms but this is false. Even before Mirantis acquisition closing, IREN is now able to land Perplexity and Figure AI. Figure AI is a very promising startup working on robotics competing with Tesla Optimus. Figure AI had previously poached engineers from Tesla and I personally know people who moved from Tesla to Figure AI when Figure was first founded in 2022. Figure AI is now at $39B private valuation. Many Figure AI engineers are from Tesla - they like to do everything in house including the full training/inference stack. Thus, Figure AI has chosen bare metal GPU with both IREN and hyperscalers. When I say bare metal it include include Kubernetes aka bare metal+ which IREN has without Mirantis. Having the full software stack in house allowed Tesla to iterate faster than it's competitors. If AI software is publicly available on a cloud then it's available to all the competitors and is undifferentiated. Enterprise Customers - Fal AI Fal AI is a AI Platform similar to FireworksAI, TogetherAI, Baseten, Modal, and Nebius' new business line. It specializes in LLMs for image, video and audio (1). It allows for development and fine tuning on managed/orchestrated GPUs so it's considered a full AI Platform (1). It's customers include Shopify, Adobe, Canva, Quora and has over 1.5m developers (1). Fal AI is known for having the fastest diffusion model inference outside the hyperscalers (1). I envision the future to be where each AI Platform has a specialized segment where they are the best. IREN is positioned to serve as the IaaS backbone for these AI Platforms. Enterprise Customers - New Leading AI Developer IREN has left this up to speculation. I will not mix speculation into this concrete only post so I'll address it else where. Prepayment at 45% Across the industry, GPU prepayments have been rising so IREN benefits from this sector uplift. 45% is a great prepayment ratio! Cash in Hand 7.6B Everyone has been disappointed by IREN's ATM. However good news revealed today is that IREN is not ATMing just to get by. They are stockpiling all the way to 7.6B for a larger purpose. IREN's Robinhood shares is 358m but with the ATM and RSU package, it's likely at 410m which puts IREN's market cap at ~16.4B. Subtracting out cash on hand, you are buying IREN's business at 8.8B! Microsoft H1 Handover July 19 was the internal H1 handover date and word on the street is that there's alot of heat coming out of H1 so Microsoft is likely running everything at full throttle for acceptance testing. For a large heterogenous DC, this can take 1-2 weeks but Horizon 1 is homogenous GB300 so I expect acceptance testing to be between 3 days to 2 weeks with the higher end of the estimate if post-handover adjustments are needed like networking configuration. Looking Ahead IREN's shorts likely got squeeze today resulting in a +20% move back to $40. I believe this is the start of the positive dominos for IREN aka IREN D-Day. 1. H1 official acceptance will be a huge derisk. 2. H2 GPUs are coming in and expected ready in late September or October. 3. H3-4 in Nov/Dec. 4. Rest of Mackenzie, Aircool Childress for 2026 ARR while CF is signed for end of 2026 or 2027 billing. 5. Sweetwater Signed in Fall Most important for stock price is each quarterly earnings to reflect IREN's execution which market is not pricing in right now because of H1 delays. I expect Q2 earnings to be overall neutral getting back on track. I expect Q3 and Q4 to get successively more impressive.

@RonenHayempour 🫡

A lot of investors forget to do this. Everyone focuses on the bull case, but what about the bear case? You should be your harshest critic when it comes to investing in a stock. Check out this post on $ONDS 👇

@kevinxu The $IREN of Australia is $IREN pal.

The $IREN of Australia just got a price target raise to $110. That's an 83% upside. Leopold owns 19.9%, Anthropic deal rumors swirling, even NVIDIA might acquire a stake. Best risk/reward on potential catalysts imo. Why aren't you all in $SHAZ ?

@BlackPantherCap Welcome pal. Wild ride ahead.

Started a position in $NUAI today. -BP Not financial advice.

Every stock has a bear, base, and bull case. If you can’t argue all 3, you don’t have a position. You have a hope. $ONDS: Bear: The roll-up breaks. 13 acquisitions never gel, backlog stalls, dilution outruns revenue. Real tech, failed per-share economics. Base: Integration works. Backlog converts. The stack spans stratosphere to street: detect, decide, defeat. Bull: A next-gen defense prime, built in public, while nobody priced it. Now the numbers: Guidance just raised to $525M+ for 2026. Cyberhawk hasn’t even closed yet. Mgmt. pointing at $1B for 2027. That’s ~8x this year’s sales. ~4x next year’s. For a 60-80% grower turning EBITDA positive with $1.5B (less $200m for DZYNE) cash and a $PLTR partnership. $AVAV gets 10x on a fraction of the growth. Legacy primes get 3x growing single digits. 4x isn’t a valuation. It’s a clerical error. The error has a cause: 33% of the float is short, funding the bear case at full size. The price is the bear case. The filings keep printing the bull case. Aug 11, the acquisitions report as one company for the first time. If the number lands, that short interest isn’t the risk. It’s the fuel. Markets pay you for one thing: seeing which case is unfolding before the tape agrees. Not instruction, just a view.

@mathlonning I would say the dilution is perfectly controlled.

Why do people love $ONDS ? Isn’t the dilution out of control?

@kevinxu Normal people talk about other people.

Dinner with friends. Not one mention of Korean margin calls, neoclouds, or Jevons paradox. What do normal people even talk about?

@moninvestor Make it make sense.

$IREN - Even after today's 15% rally, the stock is still trading below $40, with a market cap of roughly $14 billion. The company also holds approximately $7.6 billion in cash and cash equivalents on its balance sheet. That is more than 50% of its current market cap.

@DragonUK677777 I'm only in 6 names: $IREN $NBIS $NUAI $ONDS $KRKNF $ASTS

@ChairmansLedger what are your top 3 overall stock picks at current valuations

Unstoppable. $KRKNF

@LandoInvests Yes but this doesn’t mean you are not.

@ChairmansLedger Brother this is absolute cinema, they called me crazy

@Erik96754 You’re correct. People don’t change. But now you have a new (artificial) variable.

@ChairmansLedger After reading Jesse Livermore many times, and as a Christian who has read and studied the bible many times. I believe this time is not different, and there is nothing new under the sun. Why? People don’t change…that’s the constant

Every cycle screams "this time is different." Usually it's cope. AI might be the exception. Boomers see a bubble because they've seen bubbles before. They've never seen intelligence get priced. This isn't the internet. AI is the cost of thinking going to zero. And that's just the software layer. Wait until intelligence gets a body. Nothing in market history has priced that. The compute debate is over. 3 questions are left: who supplies it, who captures the economics, who gets crushed. Plenty of companies will get decimated. Fine. The theme doesn't care. It's happening whether you own it, fight it or protest it outside a data center. Depression, wars, dot-com, GFC, COVID. All were "the end" and they all passed. Capital compounded through every single one. 5 years ago I wouldn't have chosen to be an investor. 5 years from now I won't be one. This is the window. The bears, boomers and doomers? Some will save you money. Most will talk you out of the trade of a lifetime. Learn the difference or stay poor.

@stocks4545 Fair point.

@ChairmansLedger What's with ageist crap? All Boomers I know, including myself, are long term term investors, not traders. Been thru bear markets & know how to ride things out in co's we believe in. And they aren't on Fintwit😝 Most doomers here are younger/clueless RE: long term investing

@MarkosAAIG You have a spare engine by chance?

@ChairmansLedger Im a passenger princess so pick me up boo

I just picked up the latest Aston Martin for cheap. Looks incredible. Didn’t come with an engine though. Apparently that part’s on me.

@gfujett I’m long $IREN $NBIS $NUAI. All compute.

@ChairmansLedger What are your three top picks?

@noaminvests Interesting that this was your conclusion.

@ChairmansLedger seems like you’re expecting AI to make investing futile….I’d argue, how does capital not concentrate to the companies creating the most value 5 years from now? I find a lot of value in your posts and am invested similarly. I expect $onds to grow massively 5 years from now :)

@Sandeman52 🫡

Good stuff

@noaminvests Use your imagination.

@ChairmansLedger what do you mean 5 years from now you won’t be an investor?