The Diff - Our Surprisingly Capital-Abundant World

In this issue:

• Our Surprisingly Capital-Abundant World—"Capital" is a broader concept than it gets credit for. This makes some things much harder to analyze, but also helps to illustrate that economic growth is not a purely transactional process.
• Audience Capture at Two Levels—Content creators are subject to a phenomenon called "audience capture," but it applies well to the decisions made by platforms.
• Does Crypto Add Value to Financial Markets?—A more purely speculative asset class can still be useful, until people figure out that kind of utility.
• Continuation Funds, Club Deals, and the New Equity Market—A single proposed transaction provides a nice summary of where private equity is heading.
• Implementation Details—It's easier for the US government to remove monetary constraints than institutional ones, but sometimes this illustrates how important the latter category is.
• Anti-Goldilocks Economy for Junk Loans—What happens to an asset class when bad news is correlated?

The Diff August 12th 2024 14:07 • Click to play audio

Today's issue of The Diff is brought to you by our sponsor, 2 Hour Learning.

Our Surprisingly Capital-Abundant World

We all have an intuition about what capital is. Financial assets are clearly capital, as are things like factories, equipment, planes, and ships. It's a load-bearing concept—advocates for socialism take pains to argue that what they're after is workers' control over that sort of capital, not making all assets fully communal. If you happen to own, say, Koch Industries, your net worth takes a hit under socialism. But they're not coming for your toothbrush.

It's fine to take them at their word on this, and it's an academic point; Jacobin-reading fanatics aren't going to seize control of the country any time soon, especially when the most appealing examples of socialism don't have top marginal tax rates very far off from what you'd pay in San Francisco or New York.^[1] On the other hand, it raises a question: is a toothbrush not really capital? It's a complement to labor that produces more value than that labor would on its own. Transaction costs and sheer weirdness preclude operationalizing this status, but that's not true of other possessions.

For example, a bike is probably personal property. At least, it's hard to imagine a regime coming to power by arguing that it would seize everybody's bikes and give them to the bikeless. But combine a bike with a delivery app, and suddenly it's a complement to labor that increases the income that labor can extract—you may think you're a gig economy courier, but you are, in fact, a rapacious capitalist!

As it turns out, there's a way of thinking about capital that simplifies the theory and wildly complicates the practice, albeit in useful ways: "capital" is just any way to defer consumption. A share of JPMorgan Chase is capital, because you can sell it to convert it into spending, or consume the dividends. A computer is capital, of course—not just when you're sending work emails, but when you're gaming, tweeting, or using Netflix, all of which turn your prior expense into present enjoyment. Home appliances are definitely capital, but so are your kitchen utensils—if you're on the receiving rather than delivering end of the gig economy transaction, it helps to have a fork.

The first thing this model does is to tell us that a huge fraction of capital in the economy is going to be impossible to measure accurately, especially when it produces a consumer surplus: if you buy products because you'd rather have the product than the money spent, you're implicitly valuing that product at higher than its cost. How much higher, though? There are some products that end up being disappointing, and some with an absolutely insane ROI, like cheap mementos of fun experiences. There are some whose value shows up in the form of deferred opportunity cost: mechanizing "laundry day" into laundry hour, making it equally tedious but much briefer and less backbreaking work, has a massive return on investment.

Households really are little firms, which like other firms are run with some combination of ruthless capitalist efficiency, bounded rationality, and preference functions that are a little more complex than the usual H. Economicus. They have balance sheets, and their economic balance sheets are not identical to the balance sheet an accountant would prepare for them.

This view that we're surrounded by capital is not the first intuition you'd have, but it's actually undeniable based on widely-available statistical data. look at the S&P 500, which trades at ~4.3x times book value. Clearly, most of what makes a firm worth something is not that it owns lots of assets, though those assets still contribute to the firm's ability to make money, in combination with labor and other intangibles. When you really think about it, a TSMC fab is just an unusually large and complicated sort of fork: a fixed physical product that, when combined with raw materials and skilled labor, leads to a more valuable output.

The other data-based way to argue this is through the Solow Model, a simplified view of economic growth that starts by noting that a society will have more output if it has more labor, more capital, or both. If you actually try to use that model to predict things, your numbers are roughly right, but start to lag the measured outputs, and after a while that lag is quite substantial. We label it Total Factor Productivity, not so much because it's a measure of productivity but because productivity is a good term for what we're seeing but not measuring. Over sufficiently long periods, productivity can double, and has done so many times. (A bit more in the US, Japan, and Western Europe than in other places, but just about everybody on earth is richer than they would have been a generation ago.) One way to look at this is that if there's a period where total factor productivity has doubled, most of the world's output by the end of that period consists of returns on intangible capital.

Some of that capital is narrow tacit knowledge: type a lot, and you'll get better at typing more words per minute. Pitch the same product or service to dozens or hundreds of people, and you a) learn how to compress the pitch into something that really works, and b) can let your mind wander a little bit while you rattle off the spiel. Dropping your keys and wallet off in the same place every time you get home, instead of choosing somewhere each time and then hunting them down when you leave, is its own little contribution to your economic balance sheet.^[2] Language ends up being a form of capital, in several ways: it's a medium of communication, and lifetime earnings tend to be higher for people who learn a language spoken by people who generate a large share of global GDP. And new terminology is also a form of capital: Aggregation Theory and "Everything is securities fraud" are both convenient mental models that let you quickly fit new observations into well-understood categories so you can do more with them. Relationships are also a form of capital, enabling all sorts of useful coordination, whether that's delegating something at work, splitting up chores at home, or having something in common with alumni of the same school or company whether or not you actually overlapped.

So, going back to our original question, we have a fun answer to the socialist debate on personal property and capital. As it turns out, there's more capital than we thought, and some of the most valuable capital assets—math, languages, social norms, units of weight and measurement—are already under common ownership.^[3] There's still inequality, of course, and sometimes quite a lot of it. Capital is nice to have and it's a good idea to accumulate a lot of it if you can, but different people using the same capital inputs can get very different returns.

A Word From Our Sponsors

Can this really be true?

Elsewhere

Audience Capture at Two Levels

"Audience capture," is the general phenomenon where social media users slowly adjust their persona to maximize whatever sort of feedback they're most interested in getting—money, kudos, hatred, whatever. It affects individuals, but it also happens at the level of the platforms themselves: LinkedIn has slowly shifted its content towards glurge and conventional social updates rather than purely professional content. One way to look at it is that this sort of content entropy happens in real-world professional contexts, too; work with people long enough, and you'll end up with at least a few office friends, and it would be pretty bizarre to only talk to friends about work just because that's where you happened to meet them. Offices have a natural immune system, both because people have deadlines and because it's easier to enforce social norms in-person. But online, there's always someone who's using LinkedIn as a source of work-like leisure rather than actual work, so there are fewer constraints against this kind of content getting more engagement and rising to the top of feeds.

Does Crypto Add Value to Financial Markets?

Setting aside questions of real-world utility, and setting aside the tradeoff between actually-useful things (some categories of remittances), potentially-useful ones (smart contracts for purposes other than liquidity provision for crypto speculation), and illegal uses (gambling, money laundering), there's an argument that crypto adds value merely by being a new asset class that investors can diversify into, and can sell out of when they need capital. It's a fun and novel claim—suppose we assume that the social utility of crypto is exactly zero, but it still has a large market cap. That could make it worth something! But that value is a function of how poorly it correlates with other asset classes. A product with a truly random price, even if it doesn't have a high expected return, can still add value to an overall portfolio: when it rises, you diversify into something else, and since the correlation is low those increases can coincide with drops in the value of other assets. But to the extent that that's the main value-add of crypto, you run into a paradox: the marginal buyer will be someone diversifying their assets, and that marginal buyer can use more leverage than they otherwise would because the variance of their portfolio has gone down. But once they do that, they've introduced a new reason for crypto to correlate with everything else! When they rebalance, they'll sell whatever went up (so crypto correlates on the way up), and when they lose money and find that they're more levered than they like, they sell (so it tags along during declines). This is something we saw a bit of in March of 2020: before that, it was very hard to find any real-world variable that crypto correlated with—it wasn't a risk-on trade or a risk-off trade, but an idiosyncratic factor. But during the Covid crash, crypto dropped along with everything else, just faster; whether people were getting margin calls on equities and selling crypto to meet them, or getting margin calls on crypto as a result of somebody else selling for this reason, it crashed hard.

So in the long run, an asset that's most attractive because its price performance makes it a pure diversifier will end up correlating with whatever existing holders are diversifying from. What does make an asset a good diversifier is if its fundamentals are uncorrelated with whatever drives other asset prices. And in that case, diversifying into it isn't just nice for market participants—it's also a way to provide capital for activities that diversify the overall economy.

Continuation Funds, Club Deals, and the New Equity Market

Software PE firm Vista is working on a deal that helpfully combines many of ways private equity has evolved over time ($, WSJ):

1. The deal in question is a continuation fund: they'd raised a fund in 2014, which acquired several software companies that they've placed under the same holding company umbrella. Now that the 2014 fund is reaching the end of its life, they're raising a pool of capital to buy that asset so they don't have to sell it just yet.
2. The company itself is not just a single business, but a collection of them, and they plan to expand it through acquisitions. PE used to be more of a composting and scavenging enterprise—take a big business, often one that grew through acquisitions, and sell off some parts of it to more strategic owners before taking the more-focused remainder public. But PE firms increasingly like to back companies that flip this model, rolling up numerous companies in a category and taking them all public (this has worked for car washes ($, Diff), bowling alleys ($, Diff), and others).
3. The holding company merged with also-PE-acquired Citrix, and paid with equity, so its cap table includes not just Vista but Elliott Associates.
4. If this deal goes through, it means that some of the companies involved will spend even longer as private companies rather than public ones.

What all of these have in common is that they're making the substance of PE a lot closer to that of public equities: a public company typically has a broad investor base, may grow through cash- and stock-based acquisitions, and has an indefinite life. A PE portfolio company used to have one shareholder, was more likely to shrink than grow, and would spend a finite amount of time in that state before getting sold. Estimating the long-term risk-adjusted returns of PE relative to public equities in this environment is best left as an exercise for the reader.

Implementation Details

The FT has found that $84bn worth of projects covered under the Inflation Reduction Act and the CHIPS and Science Act have been delayed or canceled ($, FT). One way to look at this is that it's fundamentally a transaction cost program. You can explain the absence of particular real-world investments through some combination of 1) a shortage of capital targeting them, and 2) a surplus of rules constraining capital from acting. The capital problem is straightforward to solve—just write a check! But the rules problem is hard, because it requires so much local knowledge. If a given renewables project isn't viable, it might be because of land-use limitations for the project itself, but it might also be because of a lack of transmission lines, insufficient batteries, or not enough power consumers whose power needs can fluctuate based on wind speed or cloud coverage. And tracing why those constraints exist will run into additional thickets of nth-order consequences of rules. The natural time pressure of legislation means that a quicker decision, like offering subsidies and tax credits, will generally win out over a slower one, like figuring out exactly what we mean by "permitting reform." And this problem compounds, since at least some of the projects that do manage to get started will lead to new rules that ensure that they're the last project of that specific kind to have a quick approval process.

Anti-Goldilocks Economy for Junk Loans

A good way to think about asset classes is to look at which macro fundamentals they (approximately) respond to: if you're positive on growth, you buy stocks; if you're worried about inflation, you choose commodities; if you're concerned about a long deflationary period, you might have a balanced portfolio of quality consumer staples and long-duration treasury bonds. And assets themselves can be rearranged to change their sensitivities to these factors: one thing that originally made junk bonds so popular as an asset class is that they gave bond investors a way to prosper in a high-growth, rising-rates environment—yes, the present value of the bond's cash flows is lower if, all else equal, rates are up—but that's an economic scenario where growth is probably improving the probability of receiving those promised cash flows, in a way that offsets some of the decline in their present value. (In practice, that's only fully-offset when the bonds are very junky indeed, but relative performance helps.) Less creditworthy corporate borrowers have correctly responded to this dynamic by opting in to floating-rate products, but now the funds that track these are experiencing outflows ($, FT): a low-growth, low-rates world is especially bad for companies that were already cutting things close and will now be rewarding investors less for the risk that they take on.

Diff Jobs

Companies in the Diff network are actively looking for talent. See a sampling of current open roles below:

• Growing edtech company with product/market fit is looking for an evangelist who is passionate about gifted and talented education and can tell data-driven stories online. (Remote)
• A new data startup that's seeing early traction bringing transparency to a $1.5tr asset class is looking for a frontend-focused generalist engineer. 1-3 years of experience, or compelling internships and side projects. (West Palm Beach, FL).
• A company building ML-powered tools to accelerate developer productivity is looking for ML researchers with a knack for converting research into Github repos. (Washington DC area)
• A well-funded company building AI-powered compliance tools is looking for legal analysts who can help build and refine AI models. 2+ years of experience in a highly regulated industry, with compliance exposure. (NYC)
• A company building the new pension of the 21st century and building universal basic capital is looking for an experienced frontend engineer with iOS, Android, or React Native experience. (NYC)

Even if you don't see an exact match for your skills and interests right now, we're happy to talk early so we can let you know if a good opportunity comes up.

If you’re at a company that's looking for talent, we should talk! Diff Jobs works with companies across fintech, hard tech, consumer software, enterprise software, and other areas—any company where finding unusually effective people is a top priority.