@masayoshisson “LTAs >> no longer cyclical >> DD multiple.” No sir, LTAs neuter the bull case for pods, who mainly buy for revisions, and retail (thesis=spot price go up). The two main constituencies that buy your stock
One of the biggest lessons from my buy-side days is that fund managers love buying simple stories, driving the multiple higher.
The opposite is also true, complexity doesn’t sell and sum-of-the-parts stories stay at a discount.
$VITL is higher torque than $CALM. It’s $HIMS at $15. It’s a 14 dtc high short interest consumer growth story with a temporary hangover passing through the I/S. Leading indicators e.g. hatching flock have just turned negative. Question is whether you want to bounce along the bottom or you prefer to pay up once it’s clearer we are out the other side
Put together some longer-form thoughts on the AI Data Center Developers (aka Powered Shell Providers).. Key takeaways from Q2 earnings, what's driving recent weakness, and why I like the setup here with a number of catalysts that might reignite interest in these names.
$APLD $CIFR $CLSK $CORZ $GLXY $HUT $RIOT $WULF
Great take. Too contrarian and your return profile is the turkey-Thanksgiving lifespan chart … bad for most people’s psychology and impossible in seats where LP’s (or Izzy) demand monthly excellence. Crowding has obvious downsides, yet if there were no returns to crowding we wouldn’t all have been having dinner and buying memory for 9 of the past 12mo
With some notable exceptions e.g. Tiger style TMT western investors don’t play much in Tw. Part of this is market structure- TW cash settlement reducing capital efficiency for large customers / MM of global primes that trade on swap. Another part of it the disposition rules, challenges to short selling etc. So you end up having a global AI trade in the U.S. and an AI trade in TW mostly driven by Asian instos and Taiwanese retail, the latter having just reached record margin levels last week
@Klaudnin3 Just means price going to go down, HF are underwriting OE margins of 30%+ into 2030 vs. EM bringing on capacity in 12-24 months. Terminal Value is going to get impaired.
Agreed, HSAI not without hair though. An abnormal 40% gross margins as tier 2 auto supplier with 1-year contract about to roll off and Robosense bidding at half px
@tomicki Little bit of window dressing. Hyperscaler revenues benefitted from the reclassification of one customer as Hyperscaler due to a change in their business. Inventory provision down YoY, otherwise margin would’ve been worse. Extended payment terms. Room for more dressing next time
1/ $RMD ResMed global leader in sleep apnea devices; the higher-quality half of a near-duopoly (with Philips still recovering from its recall mess). ~22% ROIC, 60%+ gross margins, net cash balance sheet, and >100% free-cash-flow conversion. This is a high-quality compounder the market has abandoned on a single narrative.
Long ago, I was a "bottleneck bro" myself, churning out wildly bullish "buys" on data center "titans."
Having witnessed the subsequent bust, count me as a nonbeliever this time around.
It’s definitely a difficult situation for $CLBT today. The CEO change and lower revenue guidance has hit the stock hard. But I don’t think the situation is as bad as the 30% drop suggests.
The big issue is that 26’ was supposed to be a strong year driven by new products. Instead, we are seeing a setback. The new guidance is roughly where Street consensus was before Q4, so this is a reset, but not an huge one.
$CLBT said the miss was mainly due to 4 large cloud deals being delayed by new foreign-entity permit requirements at US federal agencies and in EU.
The uncertainty comes from the Inseyets product. Conversions are on track at 65% of the installed base, but the price and footprint uplift is below plan. This has pushed US state and local growth from the mid-20s to just under 20%.
I listened to an expert call from early August that highlighted some concerns:
- Sales quota attainment of only ~50-70%
- Slow adoption of Guardian & Pathfinder
- More competition from Magnet Forensics & PenLink
- Increasing discounting
- Customer budgets remain tight
- Long-term risk if Apple/Android security advances faster than Cellebrite’s ability to extract data
The CEO change looks bad, but management says the transition was already planned. Ramji was hired in May with the understanding that he would eventually succeed Hogan.
I think the market reaction is an overreaction. 15% would have been a more than reasonable.
Planned to launch with $1B (May 25). Downsized launch target twice to $500 (Aug). Launched with a little under $200. Of which >0 spent on canapés at JuiceCap Miami investor launch parties and stocking the fridge with Celsius. Past disclosure of number of LPs implies $1-3m avg ticket, presumably HNW, and current AUM implies healthy redemptions