Shares of VCX were so in demand for a chance to own a piece of Anthropic that the shares reached 25x the NAV. Who in their right mind would buy a share of a company at 25x NAV?
The value of the holdings in $VCX are held at estimated fair value. They value their holdings by:
- primary input: last funding round valuation
- secondary input: comps in the public market
The marks are review quarterly by the valuation committee, and sometimes with third party input.
The last round for Anthropic was in Feb at $380B valuation. Databricks last funding round was also in Feb at $134B in valuation.
Relative to the rounds before that, it was growth of 107.7% and 34%, respectively. Or in $ terms, $183B and $100B.
However, a 25x valuation implies that Anthropic and Databricks will be worth, collectively, 7.1T.
Yes, you read that correctly.
$VCX NAV as of the end of March was ~$18.90. The last price of VCX was $121, for a premium of 540%. Who would have thought that opening up private equity to public investors would result in carelessness?
We wrote about this back in December.
I actually think more interval funds should convert to CEFs - but not when they are seeing MASSIVE outflows and use the conversion as a mechanism to retain their fee-generating AUM. They should do it BEFORE that happens and prevent a significant discount.
https://t.co/xF97PLTybx
@WarrenPies@LesliePicker@carlquintanilla That was a great interview. I believe you are 100% correct and think the drop in token prices will slow or even reverse as the data center backlash gains steam. That could change the equation on the ROI on the AI trade.
Understood.
One thing Jon Hatzius pointed out that the Fed has never hiked ONCE in a cycle. "We" did a bunch of cuts in '24 and now are going to hike once?
I think the thinking is strange. Are we prepared for multiple hikes? Or does the Fed just stay put until the oil issue is resolved?
I've got to say, this is one of the dumbest articles I've ever read.
“There’s no other reason to do it than avoid paying taxes,” said Sheila Bair, former chair of the US Federal Deposit Insurance Corp. who helped clean up the banking industry in the aftermath of the 2008 financial crisis. “There’s risk for the firms offering this.”
Tax avoidance or tax mitigation. Tax avoidance implies almost an illegal action. This is tax mitigation.
https://t.co/CgAagE6T5i
Remember, you are focusing on the liabilities. All of this is matched by income-producing assets and similar contracted customer revenue. This is a BIG NOTHING BURGER and perhaps, clickbait.
The narrative of "hidden," "don't even show up," "crazier than you think," is overblown. The genuine risk here is reflexivity, not concealment. All of the liability commitments are in their 10-Ks. There is no "don't even show up."
The WSJ did not uncover a secret; it added up the footnotes. "Off-balance-sheet" is an accounting-location term, not a synonym for hidden.
@GlobalMktObserv This is a big SWING AND A MISS. There is nothing here that hasn't been here before. The notion of off-balance sheet commitments, basically leases and some GPU/CPU contractual buying, is not some red-herring or Enron-style of event.
This is nonsensical. Capex is ALWAYS someone else's revenue. Spending equals income by construction, so "capex is being counted as revenue" is true of all capex in every era. Stated that baldly it is a truism, not a scandal. Every capital-goods cycle in history worked this way.
Railroads, telecom, and shale all booked supplier revenue now against buyer depreciation later. What is unusual today is the scale, the concentration (one dominant seller, a handful of dominant buyers), and the circular financing, which is a difference of degree, not a new accounting phenomenon.
"Expense capex at the same rate and earnings go negative" misapplies the matching principle. Capex is supposed to be capitalized and depreciated precisely because it produces multi-year assets; expensing it all immediately would understate earnings, not reveal a hidden truth.
"Eronification." Really? Enron was fraud: fabricated mark-to-market gains and sham off-balance-sheet vehicles hiding losses. Capitalize-and-depreciate is standard, fully disclosed GAAP.
The reason why S&P 500 earnings estimates keep going up is because for the first time in history, capex expense is now being counted as revenue in Wall Street's forward earnings estimates.
If Wall Street was counting AI capex expense at the same rate as they are counting the capex revenue, then earnings would be negative.
AI is the Enronification of the entire stock market.
At 80 bps, it implies just a 6% chance of default assuming a 40% recovery rate. I would also note that the Nvidia CDS contract only began trading last November so the "record high" on the chart spans a whole 10 months. Additionally, there is a very thin dealer base so this is likely a few small trades whipsawing the CDS levels.
A nothingburger....
From JP Morgan: "•Hyperscaler earnings drove a clear shift in sentiment over the past week, lifting the S&P 500 to new highs as Tech rebounded, and putting the Momentum Unwind on pause. Retail flows remained muted (4.4%ile vs 1Y, up from 0.8%ile last week): single stocks were net sold at roughly the same pace (3.6%ile, from 3.2%ile), while ETF inflows were healthier (38.5%ile, from 1.6%ile). The key difference is that single-stock selling reflected profit-taking in hyperscalers and software as their prices surged post-earnings; net buying otherwise resumed in Memory and Semis, along with Tech ETFs—signaling that retail investors may be ready to re-engage with the AI trade. "
This was inevitable. The narrative had become some dour and basically implied a 100% chance that there was no return on hyperscaler capex.
Andy Jassy, CEO of Amazon on the conference call about capex spending.
“At this level of spend and higher, we have clear line-of-sight to strong financial returns. I'll explain why. There are two major parts of the investment, the data centers and the servers and networking equipment that go into them. These have different capital cycles. Data center capital is spent starting two years before we can put servers into them to start monetizing. Once a data center opens with servers plugged in, we start generating significant revenue right away and then get to monetize these data centers for 30-plus years without having to spend that start-up capital again. Servers and networking equipment operate on a shorter cycle. We typically purchase these a few months before putting them into service, so we have strong visibility into customer demand before we trigger the spend. If the demand isn't there, we won't spend the capital."
And then there's META CEO Mark Zuckerberg:
“At this level of spend and higher, we have clear line-of-sight to strong financial returns. I'll explain why. There are two major parts of the investment, the data centers and the servers and networking equipment that go into them. These have different capital cycles. Data center capital is spent starting two years before we can put servers into them to start monetizing. Once a data center opens with servers plugged in, we start generating significant revenue right away and then get to monetize these data centers for 30-plus years without having to spend that start-up capital again. Servers and networking equipment operate on a shorter cycle. We typically purchase these a few months before putting them into service, so we have strong visibility into customer demand before we trigger the spend. If the demand isn't there, we won't spend the capital."
I'm just not sure where the hyperscaler capex bears are coming from. Yes, they're issuing a bunch of debt but these companies also generate a ton of cash flow and have a ton of cash. MSFT is a AAA borrower - a better credit rating than the US Government!!
This is making the rounds on the FinXverse (tm!).
The main memory players - Micron, Sk Hynix, and Samsung - have apparently sold out ALL of their DRAM production through 2027. Incredible.
https://t.co/YA0gBYIaPQ