Diary of a speculator Prev: cofounder @icoalert @pirata_capital
speculating
Joined February 2018
- Tweets10.5K
- Following310
- Followers1.4K
- Likes24.3K
Finch retweeted
An announcement that is as personal as professional:
18 months ago, we started collecting online prices in real time to track how Liberation Day tariffs were feeding into inflation.
A lot has happened since. The BLS commissioner was fired. The longest government shutdown in US history meant CPI wasn’t published for the first time. And the case for an independent measure of inflation became much stronger.
What started as a tariff inflation tracker has since grown into something much bigger:
The Bloomberg Price Project (BPP), our attempt to build the US CPI from the ground up using alternative data.
Today, BPP covers 87% of the CPI basket, across both goods and services (more new data will be added in coming weeks). Every month, we collect millions of prices from card transactions, e-commerce sites, grocery stores, pharmacies, rental listings, corporate data, Bloomberg Terminal market prices and government records. And more data are coming.
The payoff is detail. Official CPI data let you drill down to categories such as “uncooked beef steaks.” BPP goes much deeper — breaking the CPI basket into almost 700 line items, compared with roughly 200 in the official data.
Brought to you by: myself, Troy Durie, Daniel Ripperger-Suhler, Joshua Danial and Scott Johnson from across Bloomberg Economics.
Explore the Bloomberg Price Project at BECO MODELS BPP <GO>, or download the data at ECAN US BPI <GO>.
I believe the paywall on Bloomberg.com is coming down briefly so you can read this without a subscription (but you should subscribe!)
bloomberg.com/graphics/2026-…
With $ARM up big today, an S1 from Anthropic that was greeted with praise and excitement and bc OpenAI is aiming for their next fundraising round announcement this Friday, I think you can start building a Softbank long here. Remember, 70% of their portfolio is OpenAI and ARM. ARM is up today and the OpenAI round at $1.25T is still a big upround from their previous ~$860m-ish round in March.
I'm attaching to this Softbank long a 1:1 $CRBRS short because they've got insider shares unlocking on Wednesday, with more coming on Nov 9th and then earnings on the 12th. I wouldn't be surprised if price frontruns poor/sketchy earnigns expectations. Finally, only about 6-8% of the tradable total float is short compared to Coreweave at 18% and Blue Owl at 16%. This isn't a perfect hedge 1:1, they're not perfectly correlated, I just like both bets individually.
As CRBRS shorts fill I add to the Softbank long in equal size
Volatility is your friend, fren
Yooo... uhhh... hmmm
This AI debt debacle is really starting to stank the more I look into it. Its not just "circular economy lulz" anymore...
For example, take OpenAI alone. They're a big piece of the AI bubble puzzle bc they buy so much compute. Compute is scarce, so OpenAI and others like them promise payments for compute many years out. All good right? Well fast forward to today and they now have future payment promises that equate to TWENTY FIVE TIMES their current 2026 revenue. Their Oracle bill alone next year ($60B) is 20% larger than their annual revenue this year, thus far.
If revenue can't keep up with future payment promises to companies like Oracle, Coreweave, Microsoft, Nvidia, Broadcom (the list goes on) they have to keep fundraising at higher valuations from the big guys, like Softbank...
Well so, uh, Softbank's stock is crashing because they're actively seeing their debt (the vehicle they use to fundraise) get closer to being downgraded to junk. Its potentially being downgraded bc their debt:portfolio ratio is nearing S&Ps 35% LTV limit.
What's wrong with Softbank's portfolio?
Welp, 70% of Softbank's balance sheet is made up of only TWO companies: $ARM (look at that chart too) and OpenAI equity. Softbank has $40B in bond repayments/refinancing to complete by March alone, with another $15B sprinkled throughout 2027. Basically they borrowed 4.5x more between March 2025 and March 2026 compared to the TWO years prior in order to yeet OpenAI at higher and higher valuations. This type of investment concentration worked when they bought ARM at $32B in 2016 (now worth over $230B) but it's their leverage that is killing them now.
At a debt:portfolio ratio currently estimated to be somewhere in the low 30%s, it would only take another ~15% drop in ARM stock price to breach the S&P limit, opening up the possibility of a debt downgrade from "stable" to "junk". Fitch, another ratings agency, has already downgraded Softbank's debt to junk.
Softbank might have a quarter or two to fix the issue before the S&P officially downgrades it, after which a slew of consequences (read: mandated, forced selling from large institutions) would occur for Softbank and its stock price.
This doesn't mean "its over", as OpenAI is only one piece of the puzzle and only needs to raise $30B this quarter. Even without Softbank other companies like Nvidia/Microsoft/etc. will likely fill the allocation.
It's also not lost on me that many of these companies like ARM and Broadcom sit right at their 200d smma/sma and OpenAI has little to no debt to worry about themselves. Bearish at the bottom?
But when major components of the AI bubble equation start to violently crack it opens up the potential for one-way flows out of stocks: the thing holding up the entire global economy. We got a taste of that over the last few days.
Going to try to map out each player over the next few days, but damn, worse than I thought already wtf
Stopped on the CRBRS short. Still fishing for the Bloom one on any bad news (now Thursday), but in general I agree with some that we've got some reversion to happen on these neoclouds, etc. at minimum
Softbank long remains for the Friday/early Nov OpenAI funding announcement
If I'm buying a short leg on oil going higher and the Fed being forced to hike I think I like crypto more than stonks now
Stonk land has finally understood reality, but I don't think crypto has yet. Fair price closer to $65-70k now potentially. Let's see how this week shakes out
I'm shorting $TRUMP on this bounce at 1.75 and higher. Midterm FUD seems obvious. Short XRP at 1.25.
I like shorting SOL, TAO and WLD too. Each except TAO have some sort of unlock/overhead supply happening between now and EOY.
Risk is in squeezes. SOL breakpoint conference is mid month. TRUMP gala bullsh*t happening Nov 22nd, snapshot for holders happens on the 12th. WLD maybe mentioned and rips during any sort of OpenAI mention from Altman. Each event likely a green candle to slam lower imo.
General sentiment is still very bullish in crypto based on long/short positioning. Lazy longs post-Binance hack and during macro chaos, right before big earnings and midterms. Good r/r for capitulation as sentiment fully reverses and positioning gets more bearish
Imagine the Druck tweet about Hyperliquid was the "Trump launches a memecoin" moment for hyperliquid:native
What's bigger PR than Druck's first tweet ever?
...Dems sweep, tokenization slows, regulatory blackballing again in favor of incumbents
I don't think it happens but it's a great bear porn thought on a day like today
Further to my post about @NYSE working to bring @HyperliquidX onshore, major credit to @ChairmanSelig and the Trump administration for making all of this experimentation possible.
There’s only one way of finding out if something is useful to the marketplace and I’m excited for us to find out. I do have some questions:
1. What is the benefit of trading perps on-chain on a regulated exchange? Isn’t the main benefit of trading on-chain to operate in a trustless (aka non-kyc’d) environment in order to minimize friction, and/or to be able to enter into atomic transactions in order to receive what you are buying in realtime? Since a perp is not an asset, that obviously isn’t helpful.
2. @standuquesne - are you aware that if you trade on Hyperliquid, the world will be able to track your positions and trades in real time? Do you see that as a benefit?
3. Hyperliquid has priority fees - a mechanism that enables market participants to pay to see information first, or even prioritize orders in front of the queue. Is this consistent with CFTC core principles?
4. Will it be possible to combine the liquidity of the offshore order-book with the onshore order-book?
5. If not, wouldn’t it be better to list perps in a traditional (centralized) order-book on a regulated exchange and not have the inefficiencies that on-chain matching introduces?
Still lost 3x these gains in the passive stonk port today, but its the small things that count on days like this one lol
