Don’t call it a comeback loan
Welcome to Cautious Optimism, a newsletter on tech, business, and power.
Friday! We made it. And with a little good news to kick off the last working day of the week: April inflation in the United States came modest. PCE was up 2.1% YoY, while core PCE was up a similarly tame 2.5%. Both figures were slight improvements on March data.
But the news is not all good. Recall that the Trump administration was handed stinging defeats in court over its use of certain emergency powers to levy tariffs without Congressional input. That got put on hold by another court, but no matter which way the Supreme Court eventually rules, the White House is fomenting alternative ways to keep trade barriers high. So, if you are a free trader, the outlook for lower trade taxes is cloudy.
Oh, and the China-US trade negotiations appear to be going about as well as those with the EU (or not?). — Alex
📈 Trending Up: GDP growth in India … inflation in Japan … hope that folks get help regarding their substance use … domestic green energy storage … Perplexity’s product mix … Foxconn …
📉 Trending Down: Embarrassments… Synopsys et al … the Zuck-Luckey rift … US GDP in Q1 … protected status for immigrants …
📒 Required reading: Bond’s May AI report. This is what I am reading over lunch. And this look at European defense hackathons!
Don’t call it a comeback loan
Grammarly is an interesting company. It’s not very young, and it’s pretty big. And the well-known grammar-checking service recently absorbed fellow General Catalyst portfolio company Coda, which competes with notion.
What does a database-notes-sheets-docs-collaboration service have to do with an AI-ish writing helper? Well, when the two companies linked up the gist — Jason and I interviewing the firm here about its new form — was that they were barking up the same tree. Fair enough, but any work to unify the pair would take a lot of capital, right?
Right. The good news there is that Grammarly is apparently already sitting on $700 million worth of “annual revenue” and per coverage is profitable. But if it wants to drop a pile of capital on product work, how could it also fund continued growth?
By snapping up $1 billion from General Catalyst’s Customer Value Fund at no equity cost. The catch? Reuters reports that GC gets a “capped return linked to revenue generated” from the billion dollars, “structured as a percentage of the revenue generated from the fund being used in customer acquisition.”
The GC CVF is designed for this sort of transaction, mind.
When Stripe emails CO offering us capital with repayment based on our revenue — to be clear, we read these notes with curiosity, not business interest — it’s a loan. I suppose in higher finance you can rebrand things, but borrowing money with set repayment terms sounds like a loan to me.
Regardless, the transaction is predicated on GC’s confidence that Grammarly’s GTM motion is so clearly defined today in results terms that capital in = capital out, so the arrangement is smart.
Or, you know, companies could fucking go public and not have to lease capital from private backers in order to pay for their ad spend.
Money flows
The way money flows through the investing world could be heading for several material changes that could wreck more than a few end-of-year bonuses. The NYPost reports that the White House’s pressure and outright hostility to the crown jewels of American education could force those same universities to drop their illiquid investments in hedge funds and similar, and instead turn to more liquid investment options like passive funds.
A provision in the House bill — the now famous ‘One Big Beautiful Bill Act’ — that would raise excise taxes on the wealthiest universities’ returns from 1.4% up to as much as 21%, the same as the corporate tax rate. Sure, the GOP wants to tear down higher education and foment plumber supply — the latter is a good idea, but is silly as a tradeoff — but it’s going to be interesting to see how much power the donor class has to push back.
Another thing to consider: Many VCs raise from university endowments. They too could take a hit here. It’s not just your bog-standard PE firms that we’re talking about. So, the tax change could wind up making it harder for venture investors to pool capital. Not good.
And then there’s this: Section 899.
Again from the OBBBA, this potential change to US policy that would “allow the US to impose additional taxes on companies and investors from countries that it deems to have punitive tax policies,” per a FT summary. The financial press is kicking up dust over the matter, because finance folks are not stoked. To wit:
And frankly, they have a point. Business folks and investors crave enough stability to make bets. If the underlying tax treatment of their American investments can change on a whim—and outside of their control—they will have to underwrite more risk. That means lower prices for US assets, eroding a valuation advantage that domestic stock markets have long enjoyed over those in other markets. That’s why, in part, so many global companies want to list in the States.
Mix in the rug-pulling we’re doing to international students at our schools, bright kids who often stay here and build, and we’re looking at the United States harming its own capital accretion flywheel, limiting the beauty of American assets, and clamping down on the reverse-brain-drain that has long been a boon.
I suppose all I can say is if you value free enterprise, and the United States as the world’s capital playground — as I certainly do — the above set of news is not entirely exciting.
The OpenAI IPO is going to make my year
But something that is exciting is the prospect of OpenAI going public. Its CFO Sarah Friar — not Sarah Frier, mind — says that her company’s planned reconstitution positions it for a future IPO. When the time is right.
What I care about is simply that the company is thinking about it. Hopefully seriously. Think of how much venture money would get recycled in such an outcome. Backers from Thrive to SoftBank to a16z to Sequoia to hell, even YC and Microsoft could make a mint. And if they do, LPs will get checks. Which everyone would appreciate.
What’s the over-under on OpenAI’s eventual IPO valuation? $500 billion?
While the Chime IPO is kinda going as expected
We’re still waiting on Chime to drop a new S-1, but The Information reports that the company is considering an $11 billion valuation when it does list. That would be a fraction of its prior, $25 billion valuation that it set in 2021.
With $518.7 million worth of Q1 revenue, Chime would be valued at 5.3x its current run rate at an $11 billion pricetag. Is that good? compared to Block’s 1.65x price/sales and PayPal’s own 2.27x price/sales multiples, yes, it’s incredible. (Keep in mind that how Block accounts for customer bitcoin transactions, its P/S is a little bit apples::organges, but all the same a greater than 2x differential is notable.)
One more thing
Today’s live taping didn’t make it across the time zone transom, so I’m sharing my notes from the now-defunct docket that went over two startups that I think are taking on the right incumbents:
Chalk raises $50M
Chalk is a software startup that helps companies use their data in AI tasks — with a twist. Instead of requiring customers upload all their data in a batch, Chalk pulls data only when needed.
Reuters OSS: “Chalk helps enterprises get their proprietary data into AI and machine learning models quickly, enabling companies to use AI to make up-to-date decisions.”
News: Chalk raised $50M at a $500M valuation, after raising $10M in 2023 in a Seed round; implication? That Chalk has hit on something that is growing very quickly.
Fun game: Guess the revenue? What’s Chalk’s ARR today?
It’s cool that Chalk has enabled real-time data processing for AI compute work — but what stops Databricks/Snowflake from adding the capability to their own platforms?
ClickHouse raises $350M
ClickHouse is an open-source startup building ClickHouse, an open-source database management tool that allows users to run complex data queries quickly.
News: ClickHouse, the startup behind the ClickHouse open-source project, has raised $350M per Bloomberg, at a valuation of $6.35B.
Bloomberg reports that Khosla led the round. Bond, IVP, Battery, Bessemer, (new investors) and Index, Lightspeed, Coatue, and Benchmark also contributed (existing investors).
ClickHouse is “nearing an annualized revenue run rate of $100 million per CEO Aaron Katz.
ClickHouse is good for data analysis, but also helps in an AI/ML context by supporting vector search and quick data preparation for model training.
ClickHouse directly compares itself to Snowflake, arguing that it is better for real-time analytics. [Claims 2x faster queries, and a 3-5x reducution in cost compared to $SNOW]
Keep an eye on those two!