This is exactly what I mean by previous post about CRWV and why we dont want to short CRWV here

$CRWV is the most obvious short in the market and that is why i'm not currently shorting it: For full disclosure, in the short section of Fraxinus Capital’s Q2 letter, I wrote that we were short the neoclouds. We are no longer short any of them. We may have covered a little early, but at the time we felt we'd captured most of the downside and r/r was not as good as it previously was. CoreWeave and the other neoclouds, but especially CoreWeave, look to me like the most obvious shorts in the AI market. I see very little foreseeable upside in the underlying business economics. And that is exactly why I'm not shorting them right now. If you're long Nvidia or other names in the AI ecosystem, it's easy to see why these companies look like the natural short against that exposure. The recent selloff has shown us that the market expects the Neoclouds to take the first to get hit if there's a crack in the AI trade. While this may have been obvious to anyone who has looked at the AI market, until the market sees confirmation, it is slow to react on a trade like this. If you’re a net neutral long/short fund, You could adjust the position sizes for beta and look at the pairing as a potentially attractive hedge. I think that's part of what the market is looking at and why these companies could have a poor equity reaction over the near term. Then there's the debt. CoreWeave has a lot of it. With Corweave’s debt trading between 14-15% Yield to Maturity, and somewhere around a 20% discount to par there's another trade to consider: going long the debt while shorting the equity. In my view, that pairing could offer a favorable skew for R/R, So there are two reasons an investor might be short these shares. Going back to why we were short in the first place, the assumptions needed to make these businesses economically viable just didn't look tenable to us. Fabio and I looked at CoreWeave, and he did most of the work on this name and other Neoclouds. From my memory, you had to assume GPU rental rates would rise over the next five years, and even then the result was something like a 1% return after debt costs. I don't have the exact calculation in front of me, but that was roughly the outcome. To us, it didn't make much sense. When you compared those economics with the profitability of companies such as Nvidia and memory names (we are long both NVDA and SK Hynix) elsewhere in the AI ecosystem, the short made sense to us as a hedge for a portfolio with AI exposure. We liked the economics in other parts of the trade, and we could use these companies to hedge some of that exposure. Fast forward to today, and these companies have been slammed. Most of the neoclouds are down 50-60% from their recent highs and down 20-30% in the last week. To me, the bearish argument now seems much more widely understood. Other investors have had time to do the work, look at the economics, and reach the same conclusion. That's actually a scary setup for me as someone looking to go short. If weak economics are already widely recognized, and the shares have already fallen that much on recognition of this, the next major downside catalyst starts to feel like something closer to bankruptcy risk. I don't personally see these companies going bankrupt in the next couple of years. I could be wrong on that. But given that view, I don't think the risk and reward favor putting the shorts back on here. At some point, I could see looking at them as a trade in the other direction. If AI sentiment turns upward, you could get a violent rally. For an opportunistic trader, there may be a technical long setup. I don't like them as fundamental longs, but that doesn't stop the shares from having that kind of move. They've been in the headlines a lot lately, and I wanted to explain where I stand. Because of this though, we are not wanting to take either side of the trade at the moment. This could change at any time in the near future and all of this is just a thought process and meant for entertainment, but we currently think there are more attractive R/R setups in the market and would rather target those for shorts.

After the IPO we shorted: $CBRS (closed) $SPCX (closed) $FRMI (still short) Please, please keep having VC pump up the pre ipo valuations of companies in this system.

Looking at the timeline this weekend and a lot of people got creamed over the past few weeks. If this is you I strongly suggest you @Crowded_Mkt_Rpt hes been noting failed news events for what feels like the past 3-4 weeks and he's said that it felt like something was coming. Obviously hes not 100% correct, but very high quality content in my opinion and worth a follow/subscription

Come you all, come here and get your fresh copium. You don't need to worry about that margin call.

Gonna draw some parallels just so everyone can see the picture: - 10 year yield went from 3.5% in march to a peak of 4.92% on October 16. - Questions were being raised about the profitability of AI Sequoia put out an article in September stating "GPU Capacity is getting overbuilt." sequoiacap.com/article/follo… - CCS called for a "cold shower" on ai demand in 2024. Their reason was that the rollout for businesses would be very expensive - To the surprise of no one, people were already asking "how will openai and this new upstart anthropic eventually pay for the investment that companies are putting into them?" Looking back, some of the narratives look crazy now like the GPU overbuild in 2023 or the lack of buisnesses being able to roll out AI due to expenses. The questions today revolve around "how will ai frontier labs be able to continue to support their funding agreements with open weight/open source taking share" "we are overbuilding" "what happens if the politicians in the US hinder capex growth?" My personal opinion is there is still a lot of growth for AI needed (all my friends and I who use it are still wanting more usage and the things we are building with it are incredible). the good thing is we can monitor the situation and just see how it develops. AI stocks are already discounting a slowing down in 27/28. If that doesn't happen, you get upside, if it does happen, well, it's not fully discounted so you get more downside haha.

Ive been using Q4 of 2023 as a guide to help decide our actions- especially with our short exposure and any thoughts about increasing exposure to ai names. It's been a really good guide so far

Sounds a lot like October 2022

Marting broke SALP blowing up so there could be a lot of credability to this. Also, based on the overlaps and strategy....could be possible lol

As of right now, Scenario 1 is playing out which is the most bullish for AI names and for the economy in general (we also have a bull steepener today and shipping rates down). I guess the PMs woke up and liked the number.

Market just told you what the easiest barbell pair trade is: Long AI names and long bonds/bond proxies. (I am not doing this trade, probably because i'm dumb)

I get that some things sounded crazy before that are just what they are now. Like if you told me that i'd have multiple ai agents who are able to do the work of several employees 4 years ago that'd sound crazy. With that being said.... This sounds absolutely insane, stuff like this reinforces the view that bulls are just walking around with blinders on. Obviously i'm still very bullish on AI, but come on lol. I will be very happy to eat crow if i'm wrong tho, come back in a year and I guess we will see.

FT saying Anthropic said they did 16.8B in Q3 (~70B annualized and 74B exit ARR) and that they are FCF positive for the quarter, hitting 11% adjusted operating margins. What I would watch for to see how the market perceives this: 1. Sell off into a rally tomorrow -> Very bullish, probably means algos sold on headline ARR and then PMs woke up and liked the numbers 2. Sell off tomorrow into no rally -> TTFN TATA FOR NOW 3. Muted price action -> no one knows what is going on bears say anthropic is lying about FCF positive (as they do with all incrementally good news), bulls say "if you adjust to only count ARR on the last day of September then it's actually 110B ARR" and round we go.

This is most likely why $ORCL is up 4.5% today. OpenAI gaining share at a rapid pace per Openrouter. Good news for the OAI ecosystem if this holds. Anthropic's loss in share probably explains why they decided to turn off distilation (both them and OAI can do this whenever they want).

Lack of risk management is the #1 killer, it's bred by bull markets, but it will always rear its head and punish you if you don't follow it. I try not to dunk on people or anything, but I really think this is a great lesson in taking profits. If this ever happens to you, you take profits. I don't know this individual and my guess by how cavalier they are being that they are very wealthy outside of this. But let's assume that is not the case and they had 1.2M and had it run up to 27M. I tell people you can't really blame this person for not taking profits on the way up, but when SALP blew up and this went down, they had a chance to close out at ~6-7M. They obviously have some kind of leverage, at a minimum, they should have closed out the leverage and gone to unlevered commons (not my advice, but just saying at a minimum what they should have done). IF they did, I think they would be around ~8M in port value (they are clearly in Memory names, so just kind of guessing based off that). Instead, they are left with one of the most embarrassing charts on X right now. Maybe they didn't realize how much money $20M or even $10M is. With that, you could put half into treasuries and half into spy or just into stocks. The treasury yields are above 5% on longer term bonds, you could be getting ~$250k-$500k/year just in the interest on those to fund your lifestyle and additional investments. Truly a sad result.

Oil bulls sound a lot now like memory bulls did in June of this year...

Literally bought my wife's wedding ring in December of 21'. Maybe the worst market timing in my life lol

it's really not that complicated, if you followed outlets like nitter.siddz.com/tickerplus/status/2104… Then you knew what the actual ARR was/is