Legibility and Liquidity
a 2x2 for risk, my first mailbag, contact vs power hitters, age of inbound
Last month I wrote in Legible and Illegible Risk:
Startups can be either legible (clear, easily understood) or illegible (hard to read). […]
Companies are either extrinsically legible (founder came from Stripe/Anduril/OpenAI, did YC, sold a company before, etc.) or intrinsically legible (the business is turning over cards and crushing). […]
Multistage firms will always back companies at the first sign of legibility. So the job of seed specialists is to invest early and help illegible founders/companies become undeniable.
I think this is a correct and useful frame but there’s more to do here. Vaishnav pushed me to think about this as a 2x2 of legibility (can you underwrite) and liquidity (can you transact) rather than a line of legibility.
If legibility is driven by internal and external marks of quality, then liquidity is driven by the supply of opportunities: search and transaction costs.
These two factors are largely correlated (the easier to underwrite, the easier to transact):
Liquid and legible:
Consensus deals that are easy to underwrite and highly transactable (tight spreads, quick bids). These are known, knowable, and sought after. Eventually they are liquid enough that quite literally anyone can buy (secondary).
lliquid and illegible:
The opportunities that are challenging to underwrite because they reject easy categorization (have out there teams, markets, business models, etc.) and are hard to transact in/compete for, often because they are just so early as a company or narrative.
At the limit these are “just people” and not even companies. There are high tnx costs depressing liquidity and they are highly illegible because there’s nothing yet to underwrite. I love backing people before they are companies (instrumental to how I approach pre-seed).
So what are the edge cases where liquidity and legibility are not correlated? What does it mean to be legible but illiquid or liquid but eligible? Where can you underwrite but not transact and where can you transact but not underwrite?
Legible but illiquid:
Stuck companies (h/t Brad Hargreaves) that have raised too much but aren’t growing enough (the 2021 growth round class). They’re highly legible (I know what they’re worth) but totally illiquid (the spreads are inverted and there is no market), at least until they run out of cash which may never happen. The “transaction cost” here is accepting a downround or acquisition that doesn’t clear the pref stack and founders are loath to pay it.
Opportunities can be narratively legible but not financially liquid. That is, there’s a lot of enthusiasm but there are no transactions to do: AI circa October 2022, Growth Buyouts before summer 2024, and “saving” or buying the stuck companies today. If you can figure out how to transact, there are huge rewards for doing so.
Illegible but liquid:
Memes, broadly defined. When there’s no pretense of underwriting but transactions are high velocity the market is illegible but liquid. In the extreme that is literal meme coins (intrinsically worthless) but closer in on the risk curve it’s purely speculative narrative companies.
Sometimes these are vehicles designed to capitalize on hype, as Dempsey calls them “narrative receptacles.” Collective belief creates liquidity despite minimal legibility. In so doing, companies or categories can even flip from being legible and illiquid (expertise-driven) to illegible and highly liquid (memes). Enough noise and hype overwhelms clear heuristics about intrinsic quality as velocity overwhelms fundamental opportunity.
It’s clear here that the best place for a seed fund to spend time is in the lower left quadrant (illiquid and illegible). That's where our ability to operate subscale is an advantage (we can take our time figuring out novel ideas in illiquid markets).
It also seems clear that legible but illiquid ideas are good fits for true risk capital (irrespective of stage); they require being early with novel underwriting, and a motion to “catch” opportunities to make them liquid.
Our aim should be to fund opportunities to the point of intrinsic legibility/quality and, eventually, significant liquidity - narrative and financial.
Mailbag
I’m doing my first ever mailbag and reader survey. You can submit questions here.
I’ll be taking questions (anon or named) here and answering them in a mailbag post in the coming weeks. AMA - I’m a pretty open book and definitionally you care what I have to say if you’re reading this.
You can of course just directly email me your Qs, tweet them at me, or leave them as a comment here. LMK if you want to be anon or named. No snitching.
Power vs Contact Hitters
Speed to proof/value is the most important thing and there’s two ways to get it right early.
And there’s basically two ways to get things right when setting up and backing a team:
High velocity, technical teams who iterate their way to the right answer through brute force. This is the prototypical young gun founder who may have built or worked in the market before but has a bias to action and the technical skills to back it up.
Experienced leaders who know the problem/customer/market well enough to need fewer swings to get a hit. This is the “old hand” domain expert who has the breadth and depth within the category to skip a bunch of steps that others would fall into.
Obviously the best answer is “do both”. Duh.
The best early teams will have the technical ability to run fast and the earned insight/conviction to do so in the right direction.
But in either case, it’s really important to honestly know which you are and you how compliment yourself accordingly. Lean into your strengths, fortify your weakness.
I Read
Nice to get a shoutout from the GOAT Fred Wilson. Maybe now he’ll respond to my emails.
When the world is full of big-brained robots built on public data, “proprietary data”—personal experience, domain expertise, what we know rather than what we can do—is our only edge.
AI.com Is for Sale. Asking Price? $100 Million — The Information
Snack Makers Are Removing Fake Colors From Processed Foods - Bloomberg
If You're So Smart, Why Can't You Die?
Human beings are, broadly speaking, just some silly warm meat running on not particularly much energy. No one can convincingly define any sort of “secret sauce” we got. So isn’t it inevitable we get machines that are like humans but on more powerful hardware?
The way through this dilemma is to notice that the thing you’re looking for is not a skill that’s literally impossible for machines. You’re looking for knowledge that needs to be gathered the slow, iterative human way and not the fast, recursive, copied with perfect fidelity machine way. Allow your humility to say that a machine living in the world and subject to the same selection pressures as you could eventually reach a better place than you. The material question is how well the selection pressures can be simulated by sufficient compute and compressed into a much smaller time frame than the millions of years it took us.
Age Of Inbound
Join me and Nico Ferreya from Default.com for dinner with sales and marketing leaders to talk about how growth will work in a world without outbound.