Working in Public and the Economics of Free

Plus! Delevering, Alternative Lending, Time To Build in Britain, Pinduoduo, Executive Orders, More...

This is the once-a-week free edition of The Diff, the newsletter  about inflections in finance and technology. The free edition goes out  to 10,366 subscribers, up 977 week-over-week. This week’s  subscribers-only posts:

In this issue:

Programming note: The Diff is relocating to Austin. I still like  New York, at least New York circa 2019, so think Elba, not St. Helena. I  don’t expect any disruption to the usual publication schedule.

Working in Public and the Economics of Free

It’s the fate of every essential part of the economy to become  ubiquitous, then invisible, then superficially boring. You don’t think  about plumbing, electricity, gasoline, or staple food products (mostly)  because a handful of specialists work very hard to make them available  all the time. As software eats the economy, both in terms of market  capitalization and in terms of how much behavior it mediates,  open-source software increasingly fits into that category.

But we don’t actually know how to pay for it. We don’t know how to ensure that it exists. Sometimes, chunks of it disappear in a fit of pique, albeit for a brief period.

That’s an important problem. We’re in a strange in-between state  where software is increasingly essential, the most important bits of it  are built by dedicated volunteers, and there’s no great way to encourage  them to a) keep doing this, and b) stay motivated as the work gets less  and less fun. That was my main takeaway from Nadia Eghbal’s Working in Public, published  this week.

The book is a tour of the open source phenomenon: who builds  projects, how they get made and updated, and why. This is an important  topic! Almost all of the software I use to write this newsletter is  either built from open source products or relies on them: the text is  composed in Emacs, on a computer running Linux; I browse the web using Chrome, which mostly consists of open-source Chromium  code; my various i-devices run Unix-based OSes (iOS is not open-source,  but there are as many open-source implementations of Unix as you want);   etc.[1] All of this software Just Works, and I don’t directly pay for any of it. Clearly there are some incentives at work, or it wouldn’t  exist in the first place, but generally the more complex an activity is,  the harder it is to coordinate without pricing signals. Somehow,  open-source projects don’t fall victim to this.

As Working in Public shows, they do face a demoralizing  lifecycle. When a project starts, it’s one coder trying to solve a  problem they face, and trying to solve it their way. Sometimes, there’s a  good reason their preferred solution doesn’t exist yet, and the project  languishes. Sometimes, the project catches on, acquiring new users, new  contributors, and new bug reports and feature requests.

The book points out that a project is not just the literal code:

There’s a reason why open source projects are called  “projects,” rather than just code. While code is the final output of a  project, the term “project” refers to the entire bundle of community,  code, and communication and developer tools that support its underlying  production.

That’s an important point: a software project is a loop for  determining what changes to make, making them, and then double-checking  to make sure they were the right call. Over time, a project moves like  an amoeba, extending a single-feature pseudopod and then squelching in  the direction of highest user demand.

And, over time, the outside requests soak up most developers' time.  Maybe you write a nice library for doing interesting things with text,  which is great, until someone who uses it points out that it can’t  handle Cyrillic or Hangul characters. You create a way to do something  fancy in browsers, and it turns out to be broken in Internet Explorer.  Your library is great for analyzing what you think of as “big data,” but  somebody else’s idea of “big data” is orders of magnitude bigger, and  they’re asking you for help.

Basically every successful software project runs through the classic  story of economic development, at warp speed. The cycle many countries  go through is that early in their economic growth is driven by  manufacturing, and workers' output per hour rises fast (a country that  builds its first few factories can soon justify building a bigger port,  better roads, and more power plants, all of which boost workers' output.  Meanwhile, those workers are getting better at their jobs). As a  country grows, more of its internal consumption takes the form of  services, rather than goods, and it’s very hard to raise productivity in  services. And since the service sector of the economy is bidding for  workers against the manufacturing sector, higher manufacturing  productivity drives up service worker wages, even if those workers  aren’t producing any more. As Baumol  famously observed, it takes just as many musician-hours to perform a  string quartet as it did in the 17th century, but since the other things  those musicians could do have gotten more lucrative, the price of their  time has risen. Moore’s Law makes computers faster, which gives  programmers higher output per hour, so if your fellow musicians aren’t  going to put away their instruments and enroll in Lambda School, their  wages have to rise.

Software projects go through this exact cycle: at first, they’re  mostly producing and building on “software capital,” code that executes  the underlying logic of the process. Over time, more of the work  involves integration and compatibility, and as all the easy edge cases  get identified, the remaining bugs are a) disproportionately rare (or  they would have been spotted by now) and b) disproportionately hard to  fix (because they have to be the result of a rare and thus complicated  confluence of circumstances). So, over time, a software project falls  into the Baumol Trap, where high productivity in the fun stuff produces  more and more un-fun work where productivity gains are hard to come by.

This is depressing for project creators. Early on, they have the  triumphant experience of building exactly what they want, and solving  their nagging problem. And the result is that they’re cleaning up after a  bunch of requests from people who are either annoyed that it doesn’t  work for them or have unsolicited feedback on how it ought to work. No  wonder Linus Torvalds gets so mad. Running a successful open source project is just Good Will Hunting in reverse, where you start out as a respected genius and end up being a janitor who gets into fights.

Somehow, though, the important open source projects keep on going. The book suggests a few reasons for this:

In the end, it’s a mystery that this system works at all. The best  explanation is that we’re still very early in the software story:  there’s so much low-hanging fruit that we can tolerate vast economic  inefficiencies as long as the work being done is on approximately the  right projects. The “sponsorship” model is the one that most closely  matches the underlying economics, because a sponsor will pay more to  back a project if it’s more essential. On the other hand, the sponsor’s  economic interest is not in building the best software for the world,  but the best software for their purposes; it’s all about commoditizing  the complement. So this isn’t a great option, but it’s a least-bad one.

Disclosure: I know Nadia, I got a free copy of the book from  Stripe press, and I was delighted to discover—midway through reading the  book and quoting the review—that the book quotes my modest proposal about COBOL.

[1] I also use Excel, which is definitely not open-source, but for heavier data analysis I prefer pandas.

A Word From Our Sponsors

Here’s a dirty secret: part of equity research consists of being one  of the world’s best-paid data-entry professionals. It’s a pain—and a  rite of passage—to build a financial model by painstakingly transcribing  information from 10-Qs, 10-Ks, presentations, and transcripts. Or, at  least, it was: Daloopa uses machine  learning and human validation to automatically parse financial  statements and other disclosures, creating a continuously-updated,  detailed, and accurate model.

If you’ve ever fired up Excel at 8pm and realized you’ll be doing  ctrl-c alt-tab alt-e-es-v until well past midnight, you owe it to  yourself to check this out.


Household Delevering

The Fed’s quarterly report on household debt and credit is out, and debt has declined for the first time since 2014.  Delinquencies are also down. The combination of large transfer payments  and limited spending opportunities has basically forced households to  either pay down debt (if they have any) or save more (if they don’t). So  this is evidence that the recovery in consumption will be fierce when  it arrives.

More Alternative Lending

A few days ago I linked to this interview with the CEO of Pipe, on turning SaaS cash flows into a financial product. There’s more competition ($), from a surprising source:

Japan’s MUFG Bank will use artificial intelligence to  screen Asian startups for financing, looking beyond balance sheets to  forge relationships with promising businesses who otherwise may not yet  qualify for bank loans… The fund will finance e-commerce, education and  health care startups. Using AI, the fund will analyze companies' bank  accounts, subscription numbers, accounting data and other information to  screen them for loans.

This thesis has a long way to run: interest rates are low, and  subscription products turn one-time expenses into an asset with longer  duration. Getting that asset off startups' balance sheets and onto the  balance sheets of traditional lenders is a vast arbitrage, and as it  gets more common we should see even more new subscription-based  startups, especially selling to mid-sized and larger companies.

It’s Time to Build in Britain

The UK is reducing regulatory barriers to new housing.  At a small scale, this is a policy tweak in a country that doesn’t  build much housing (as Boris Johnson put it: “In 2018 we built 2.25  homes per 1000 people. Germany managed 3.6, the Netherlands 3.8, France  6.8. I tell you why—because time is money, and the newt-counting  delays in our system are a massive drag on the productivity and the  prosperity of this country.” I don’t know what “newt-counting” is, but  I’m opposed.) At a larger scale, the big story is that populists usually  side with current homeowners over future homeowners (i.e. they try to  drive housing prices up rather than make new homes more affordable).  There’s increasing evidence in the UK and US that this is changing.


I’ve been planning to write about Pinduoduo for a few weeks, but this Turner Novak post says all that needs to be said. The company has built an amazing user acquisition machine, and it’s worth $100bn five years after founding.

One thing that stands out is how well Pinduoduo uses existing  companies' infrastructure to market itself. Their first big growth  channel was WeChat, and it uses WeChat Pay for transactions. I’ve  pointed out a few times that the fastest growth stories are not from the  best companies, but from companies that have a temporary cost advantage  selling a commodity adjacent to a company with a durable advantage. In  any given year in the 2000s, a new SEO-driven e-commerce site would grow  faster than Google did; early Zynga put up better growth numbers than  early Facebook; Compaq grew faster than Microsoft. Pinduoduo actually  seems to be aware of this, and is moving more of its business to its own  channels, but the growth model is more fragile than it looks.

Industrial Espionage in Taiwan

A long-running hacking campaign stole “source code, software development kits, and chip designs”  from Taiwanese companies, probably on behalf of China. China has gone  to great lengths to develop a national chip industry, and has stepped up  its efforts in the last few years in light of US sanctions. So far, it  hasn’t succeeded. Chip fabrication benefits from the same  “geoepistemological advantage” that Shenzehn does: the right people,  with the right relationships, colocated with a huge amount of capital  equipment. Some of that is easy to copy, but since all the components  are complementary copying half means getting less than half the benefit.  And the half-life of this knowledge is short, since chip companies keep  rolling out newer processes. Still, this underscores how aggressive  competition in this space is.

In what may or may not be related news, SMIC, China’s largest competitor to Taiwan’s semi fab industry, has raised its capital expenditures plans for the year.

Executive Orders

Donald Trump has signed an executive order mandating domestic sourcing for some pharmaceuticals.  This is a backwards-looking plan, but prudent: one of the drawbacks to  complex global supply chains is that a disruption anywhere becomes a  disruption everywhere. China and India have a comparative advantage  compared to the US in producing many low-priced inputs for drugs, so  making them in the United States is not economically rational—unless the  plan is to have a stockpile and the ability to scale up production in case imports get disrupted.

Domestic sourcing has another advantage: quality control for cheap  drugs is important. The low margins mean an increased incentive to cut  corners. There was a long-running scandal at Ranbaxy  a few years ago, and what stood out about that scandal was not so much  the quality control issues themselves as the fact that the problems—and  fines—went on for years. A domestic supply chain is easier to monitor,  and at least makes it possible that bad actors in pharma eventually have to stop.

Executive Orders, Con’t.

The plan to extend unemployment benefits by executive order relies on some tricky accounting games:  essentially, some money has been allocated for aid to states, but not  technically spent, and the executive order plans to spend it on  unemployment instead. Timing matters: if Trump acts and then gets sued  to reverse his actions, the money still gets spent and Congress has to  deal with the consequences. It’s a very politically savvy move, since a)  it’s a vast overreach of Congress' traditional role, which Congress  doesn’t like at all, but b) the only way to stop it is for  House Democrats to sue the President to stop him from cutting large  checks to the unemployed at a time when the unemployment rate is above  10% and the election is 87 days away.

Executive Orders, Pt. 3

Yet another set of executive orders, this time banning any transaction with Bytedance or WeChat after 45 days. The WeChat addition is interesting. As Jordan Schneider has pointed out,  WeChat is a much less important issue because a) it’s smaller (3m  users), b) it’s mostly used to keep in touch with friends and family  back in China, and c) because it’s so dominant in China, a US ban  wouldn’t hurt the app much, although it would inconvenience millions of  people.

In the words of one constitutional law lecturer,  “[The executive branch is] not just going to be waiting for  legislation. [The President has] a pen and…  a phone, and [whoever is  President at any particular moment] can use that pen to sign executive  orders and take executive actions and administrative actions.”