Decoupling, Digital Currency, and the Consumer of Last Resort

Plus! A SPAC Modest Proposal; Scraping for Profit; Post-Cookie; Best Buy Bundle; Debt is Here; The Short-Selling Cycle

Decoupling, Digital Currency, and the Consumer of Last Resort

It would be accurate but incomplete to think of the global economy in terms of the circulation of goods and services around the world. There's regional variation in what people consume, mostly driven by income differences, and there's regional variation in what they produce, driven by a wider variety of factors that range from historical accidents to entrepreneurs to government policy to natural resource endowments. Another view, that's surprisingly accurate but still limited, is to view the entire global economy in terms of the circulation of savings: there's one dominant producer of reserve assets, the United States, and everyone in the world either directly or indirectly demands them. And there's a third model, which has the same traits. It works like this: China is the best country in the world at adding value to raw materials and intermediate goods; raw materials and components flow into the country, and finished goods flow out. In this model, the function of the global economy is to find end consumers for more of these goods, and the result of that process is demand for certain high-value input goods that can't be made there just yet—M1 chips from Taiwan, memory chips from Korea, storage chips from Japan; oil, iron, copper, and food from around the world.

In a world without widely-trusted fiat currency, the first model would have all the explanatory power we needed. There might be countries that chose to accumulate savings, and others that chose to grow their debts, but over time, these imbalances would even out through financial crises, which would tend to hit the local economy hard while leaving the rest of the world unscathed. (The way to claw your way out of debt when you can't issue currency is to reduce imports and/or increase exports, i.e. for a given level of labor and capital, you need more of the returns to go to people other than the local laborers or capitalists.)

The modern world does not work that way; there are countries that can issue widely-circulated currencies, and thanks to the network effects of money, the most widely-circulated ones are in the highest demand. This leads to a dynamic where the US is the consumer of last resort ($): a financial crisis is often ultimately a shortage of dollars, and America can print those just fine.1 Since currencies adjust to exports, the usual rule is that countries that want to pursue an export-driven policy end up accumulating dollars (and, to a lesser extent, other foreign currencies) to keep their currency from appreciating. Sometimes this takes the form of a currency peg—Saudi Arabia can keep the riyal steady because it sells lots of oil and stashes away lots of dollars—and sometimes it takes a softer touch, as in this in-depth look at how Taiwan's currency policies work.2

Industrializing through exports is a policy choice, and often a good one, but every country that makes that choice puts the US in an annoying position: those countries can't be net exporters and dollar-accumulators unless someone is willing to take the other side of the trade, by running a trade deficit that's financed by allowing trade counterparties to accumulate assets. This is not a terrible deal for America by any means; for a given level of production, we get more consumption, at the cost of some ugly financial statements. It's certainly a choice that China made; China's forex reserves total some three trillion dollars. What does that buy them?

  1. The flexibility to set their exchange rate wherever they want it to be, at least within reason, while still allowing enough wiggle room to get price signals from the market.
  2. The ability to provide liquidity to Chinese companies should they find themselves short of foreign currency.

Export-dependent countries are vulnerable to crises because they rely on demand in other parts of the world, and a huge cache of reserves (plus capital controls to keep money from leaking out) is a way to ride out the crises.

The US obviously benefits from this relationship in an immediate material sense: the products China exports are much cheaper than the alternatives, and the complex global supply chain enabled by this—specialized components produced in places that have very narrow competitive advantages, natural resources produced wherever they can be found, and all of the above assembled affordably and flexibly in China—has made the global economy much more functional. Replacing that is, in practical terms, impossible: there just aren't other parts of the world with the requisite mix of labor market slack and state capacity, and even if a shift did happen, the world would lose some of the unique advantages of having so many different suppliers all crammed together in specific specialized cities, from Shenzhen's electronics to Dafen Village's oil painting replica assembly lines. The next-best alternative is for US imports to be more expensive, and for the hardware supply chain to be far less adaptable. It might be an acceptable cost, but it will be a steep one, and the companies exposed to that cost—US hardware companies that sell to Chinese manufacturers, retailers that buy Chinese products, American brands like Starbucks, Marriott, and Nike that have expanded into China and rely on it as part of their growth story—will no doubt lobby fiercely against this outcome.3

If the US plans to decouple from China, it's a multi-decade process involving a little bit of reshoring and a whole lot of different-shoring; convincing manufacturers to move out of China and into India, Vietnam, Mexico, and other places that combine a) cheap enough labor, and b) closer diplomatic ties to the US than to China. There are parts of Africa that seem like good options, but China is years ahead there, and has also invested a lot in getting closer to governments in that part of the world—partly to access natural resources, partly to have more influence in international bodies that weight each country equally.

China has a very different problem in decoupling from the US: the problem of finding a replacement for the utility of owning a vast amount of FX reserves and having a trade relationship with a country that has an effectively unlimited appetite for goods.

And this is where things get interesting. The WSJ has written a piece on China's digital currency ($) with some very telling details. For example:

The U.S., as the issuer of dollars that the world’s more than 21,000 banks need to do business, has long demanded insight into major cross-border currency movements. This gives Washington the ability to freeze individuals and institutions out of the global financial system by barring banks from doing transactions with them, a practice criticized as “dollar weaponization.”


The digital yuan could give those the U.S. seeks to penalize a way to exchange money without U.S. knowledge. Exchanges wouldn’t need to use SWIFT, the messaging network that is used in money transfers between commercial banks and that can be monitored by the U.S. government.

A good reason to hoard a lot of currency is the risk of losing access to more of it while still needing to use it; a popular alternative reduces that risk.

Even more interestingly:

The money itself is programmable. Beijing has tested expiration dates to encourage users to spend it quickly, for times when the economy needs a jump start.

Programmable money, tied to real-world identities, and universally tracked by a central bank, starts to look suspiciously like a substitute for the consumer of last resort. Every year that China gets richer, domestic consumption plays a bigger role (exports were 26% of China's GDP in 2010, and 18% last year). If domestic consumption can be tightly controlled, then it's a way to not just increase the volume of consumption but to control the variance of demand for the goods China produces. It's not yet enough to match the size and variability of global demand for China's exports, but every year it gets closer.

The US and China are both talking seriously about decoupling, but the digital yuan indicates that China's government is more effectively planning for it.

Further reading: Trade Wars Are Class Wars is my go-to book on the currency consequences of export-driven growth. And while I started this piece before the panel happened, this discussion of the US, China, and currency issues is worth reading. Reserve currencies are incredibly sticky, but technology shifts have a way of making some economic variables less important. The real risk is not so much that the dollar loses its reserve currency status, it's that owning reserve currencies—and having a good relationship with their issuers—becomes less important.


I had a very enjoyable interview with Mish Saul, where we discussed everything from existential risk to trade policy to IPOs to kids. And in a few months I'll be speaking as part of On Deck's Investing Fellowship.

A SPAC Modest Proposal

Regulators are questioning the fact that SPACs include aggressive projections when they merge, while traditional IPOs largely don't, and rely on backwards-looking data instead ($, FT). It's healthy for markets to have diverse regulations, but SPACs have a dangerous combination of a) looser rules on what companies can say about the future, and b) surprisingly high fees subsidized by retail investors. Either one is fine, but both together means adverse selection.

On the other hand, SPACs have shown that there's a lot of demand for early-stage companies, and that's a good thing. As IPOs happen later and later, a greater share of total returns accrues to VCs and early employees instead of public investors. Microsoft went public with a market cap of under a billion dollars in 1986; by the time Facebook went public, it was already worth $100bn.

One option for SPACs would be to give traditional IPOs more of a safe harbor for disclosing projections about the future. Right now, the thinking is that a SPAC is a better deal for a company that either a) is willing to pay a lot to be public, and is willing to accept underperformance once it's public, or b) a company that has a good story that doesn't show up in the financials just yet. If more companies in category B could go public the usual way, SPACs would go back to being a bad signal, which would make it harder for the worst deals to get off the ground.

Scraping for Profit

LinkedIn joins Facebook in being victimized by a leak-by aggregation: details on half a billion users are for sale on a hacking forum, but the details are all things LinkedIn surfaces on its own. LinkedIn, like Facebook, tries to rate-limit people who access pages en masse, in part to prevent this exact kind of issue: the data was public, but deliberately inconvenient to access, which made it essentially private. Since sites now have fewer legal protections against scrapers, LinkedIn, Facebook, and others will have to rely on technical countermeasures to prevent more scraping.

Interestingly, that problem—taking something from technically accessible to trivially accessible—was part of the blowback when Facebook launched the News Feed. News Feed didn't have any information that wasn't already on profiles, but it was still a disconcerting experience for users.


Apple and Google are making it more difficult for advertisers to track individuals across sites, which is naturally leading the ad industry to adjust. Apple is offering some data, not tied to users, on ad performance for apps. And Twitter is partnering with Nielsen to better measure Twitter video ad performance. If direct response ads get harder to measure, more inventory may be shifted over to brand advertising, where measurement was hazy to begin with. So Apple and Google may end cutting into traditional TV advertising budgets, too.

In a related story, Procter & Gamble worked with the China Advertising Association on their user-tracking workaround ($, WSJ). This piece is notable in part for its profile of P&G's sophisticated advertising strategies, which combine first- and third-party data:

Frustrated with what it saw as tech companies’ lack of transparency, P&G began building its own consumer database several years ago, seeking to generate detailed intelligence on consumer behavior without relying on data gathered by Facebook, Google and other platforms. The information is a combination of anonymous consumer IDs culled from devices and personal information that customers share willingly. The company said in 2019 that it had amassed 1.5 billion consumer identifications world-wide.

For small advertisers, a change in policy at a big tech company is basically equivalent to a new law, or a court ruling—they can be mad, but they don't have any practical way to change things. But for bigger advertisers, it's more like the start of a negotiation.

Best Buy Bundle

Best Buy is testing a $200/year membership program, with free shipping, lower prices, and unlimited tech support. One way to define a commoditized business is that it's one where you win sales rather than customers—after every transaction, the customer is gone and has to be brought back for the next. Much of the retail business consists of finding ways to move along this continuum, whether it's through convenient locations (win the customer because they walk or drive by frequently), loyalty programs, or targeted marketing. Membership programs are an increasingly popular option, and can essentially create a surplus for both Best Buy and the consumer by shrinking the amount of marketing and rent Best Buy needs to pay to close each incremental sale.

Debt is Here

About a year ago, Alex Danco wrote the canonical piece on why subscription companies were going to start funding their growth with debt. The thesis is older than that, but was typically kept quiet—Vista, for example, made some very good returns from the realization that a SaaS company's revenue is more like the coupon on a very senior bond, and that looking through the P&L to the implicit bond bundle could reveal some striking mispricings. But the trend has accelerated. One obstacle is that many of the companies that borrow against subscriptions are small, and the institutional money that can fuel this tends to get distributed in big chunks. This post goes into some detail on that problem.

But now, smaller deals are happening ($, WSJ):

The loans backing the complex bonds—known as asset-backed securitizations, or ABSs—can be small, like the $25 million Golub provided to software delivery specialist CloudBees Inc.

This is a very big deal: since the capital gets reinvested in growth, small deals beget larger ones; CloudBees is going to turn around and invest that $25m in ramping up revenue fast enough that they can do a $50m or $100m deal after. Connecting institutional capital to the SaaS business—a business which, in economic terms, keeps on minting fixed-income products but mysteriously funds them by raising equity instead of debt—will rationalize balance sheets and accelerate growth. (And, at some point, overshoot; every new asset class goes through a Campbellian hero's journey, which is not complete without something terrible happening.)

The Short-Selling Cycle

Short-focused funds have lost significant assets in the last year. There are two causes:

  1. It's been a tough year for short sellers, with lots of demand for exactly the kinds of companies short sellers classically bet against—overhyped businesses with shaky stories.
  2. In part because of this, it's become easier to manage a short book as part of a larger multistrategy fund. Because of the distribution of returns from short selling, an investment in a short-focused fund tends to need more active attention than one in a balanced fund (if you're an allocator, you have to back such a fund knowing that there will be times when markets are doing well and that fund is the specific reason you're underperforming rather than outperforming).

But the opportunity set for these funds is bigger than it was a few years ago, simply because there's more variance in new companies going public, more uncertainty about the future steady state, and more businesses that survived the last year because of external liquidity rather than internal durability. Good times to short stocks are often bad times to be running a short-focused fund, because the assets under management cycle is roughly inverse to the opportunity cycle.

  1. This doesn't mean that recessions are completely optional, or that the US should always run deficits, cut rates, and engage in quantitative easing in response to the slightest deceleration in global GDP. It just means that in many slowdowns, the most directly solvable problem is the dollar shortage, and fixing that produces additional real GDP that can be applied to solving deeper systemic problems.

  2. This piece may be my single favorite bit of macro research ever, since it combines a look at industrial policy, monetary policy, trade policy, asset class stickiness, and statistically-savvy forensic accounting.

  3. The US companies present in China reveal that there are two kinds of global races to the bottom. There's the classic country-to-country one, where labor-intensive and capital-light activities flee expensive places and cluster in cheap places—and keep hunting for cheaper ones. So some US textile manufacturing and appliance assembly left for Japan (not in the sense that manufacturers shut down factories in one place and opened them in another, but in the sense that buyers started choosing more cost-competitive Japanese suppliers). Then those jobs left Japan for Taiwan, and Taiwan for China.

    But there's another kind of race to the bottom that involves financing and revenue as well as costs, and that one is more granular: if opening a factory means getting tax perks, cheap financing, and access to a large and growing market, then the full bundle can be worth it even if wages are high. (This is part of why some jobs migrate back to the US.) China, with its high savings rate, growing consumer class, and KPI-driven governance, has had a growing advantage in this race to the bottom that offsets the erosion of their labor arbitrage. Every time wages in China go up, opening a factory gets less affordable, but when some of those wages are recycled into savings that fund investment, and when the rest turn into consumption, China's cities and provinces can have a net more compelling offer to make multinationals.