Foxes, Hedgehogs, VC Alpha, Politics

Plus: The Middle-Income Trap, working capital, Lambda School, and empires

The Latest

You know the line about “Foxes and Hedgehogs.” The hedgehog has a Theory of Everything, the fox has a theory for every single thing. But is fox status the result of being a polymath, or just being so smart that the common thread between different theories is invisible? 

Politics in the Big City: I like talking about politics, which is not the same thing as having any useful opinions. For example, in the piece linked I talked about why people love Yang (who has since dropped out) and ignored Mike Bloomberg, the high bidder for what turned out to be an auction for frontrunner status. Bloomberg’s prospects are uncertain after last night—it’ll take a lot of 30-second spots to compensate for getting put on the spot like that—but he’s still a contender. (A while ago a news site gleefully assembled a supercut of Sanders referring to “millionaires and billionaires,” and noted that he switched to just “billionaires” once his net worth hit seven figures. Depending on how Super Tuesday goes, he might further refine things to “That one billionaire.”)

Judging VC investing skills is a surprisingly hard problem. Only a handful of venture-backed companies really matter for returns each decade, so luck is important. There are ways to bend luck your way, of course, but since a good venture investment is by definition contrarian, better information about a given VC’s process and views will lead venture LPs to make worse allocation decisions.

Finally: I’ve been fascinated by the “middle-income trap” for a while. It’s the phenomenon where a country grows its economy by competing on cost, but, as it gets richer, the cost advantage disappears. This applies to careers, too. If you get your foot in the door because you’re cheap to hire, your reward is getting raise after raise until you’re profitable to fire.

And stay tuned for future posts on Coindesk, where I’ll be contributing pieces to their new opinion section


Alex Danco is a frustrating writer, because he keeps publishing the definitive take on stuff I kept meaning to get to. I have a ten-year-old draft of a post about nepotism as a form of finance in my Dropbox. I’d been not-working-on-that-story for a decade, and he just… tweeted it out. That was a few weeks ago, and he did it again, writing a good take on how more venture-backed companies will choose debt over equity. 

The basic way I think about it is that a lot of companies, especially DTC and SaaS companies, raise their first really big round on the basis of a customer acquisition cost to customer lifetime value calculation: we spend $x to get one more customer, they produce 3x in expected gross profit, so give us 100,000x to repeat the process. That’s a trade expressed by selling equity, but it’s a credit decision. CAC is the face value of the loan, LTV is the current value of the collateral, and like any other loan the collateral’s value can vary—but in this case, it varies a whole lot.

The other useful point to add is on convexity: traditionally, you’d think of equity as having more upside than debt. A lender is the claimant on some finite amount of money, while equity owners get the rest, so the more money there is the more incremental profit accrues to the equity holders. But that assumes the lender just makes one loan. If a lender keeps lending more as the company grows, their share of each incremental dollar of profit looks like that of an equity holder, to a lender. 

NY Mag has a hit piece on Lambda School, making two points: 1) their placement rate is lower than they claim (bad news if so!) and 2) that while they claim to get paid a percentage of students’ additional income, they actually sell these income-share agreements to outside investors. Point 2 is a ridiculous point to argue about: if Lambda School is charging tuition the right way (i.e. only after they’ve delivered something of value to students), then they’ll wait a long time for cash—so they either need to raise a ton of equity or sell the proceeds in advance. Lambda’s incentives are still aligned with students, because it’s an iterated game: if they sold a bad batch of ISAs, the sale price on the next batch would be worse, which would hurt their equity valuation and potentially be a lot more expensive than just having bad debt on the books. 

Some of Lambda School’s problems sound like traditional scaling problems. I didn’t go to college in the 60s, but I did read some David Lodge novels, and Lodge has plenty of jokes about bureaucratic snafus driven by rapid enrollment growth back then. A lot of the coverage of Lambda School compares it to some abstract ideal, or to what Austen Allred’s absurdly sunny Twitter account. I compare them to the other alternatives, and despite some problems Lambda comes out ahead.

And a fun piece investigating whether empires promote trade. Does this matter today? Depends. Maybe it’s a coincidence that trade is largely denominated in US dollars, that parties to trade who don’t have the same first language will probably negotiate in English, and that transoceanic trade is protected by the US Navy. One future possibility is that the best US/China comparison involves Zheng He; that we’ll let our ships fall apart because we have plenty of nice things at home.