Category: Uncategorized

• Crypto is an Unproductive Bubble

  This post sparked a great conversation on Hacker News! See the comments here.
  Another update: Scott Alexander just (12/8/22) wrote the best counterpoint to this that I’ve read – see it here

  “The four most expensive words in investing are: ‘This time it’s different.’”

  John Templeton

  I’m writing this essay not so much to convince anyone of my point of view as to outline my current thinking around cryptocurrencies and the mania surrounding them. Because this is a controversial subject over which smart people disagree, I’m making these claims in public so that I might be proven either right or wrong later.

  If you disagree with any of my claims here, or have more to add, I’d appreciate you leaving a comment with your thoughts, both for my benefit and for that of other readers.

  Claim 1: Crypto is a Bubble (Confidence: High)

  The hallmark of a bubble is people buying an asset primarily in the hopes of future appreciation driven by other buyers, rather than because of any notion of its long-term value. Cryptocurrencies have satisfied this criterion starting a number of years ago, and from what I’ve seen continue to do so. The trouble with bubbles is that once the supply of hopeful buyers runs out, prices stop rising and start falling as fear replaces greed. I predict that this will happen to all cryptocurrencies, including Bitcoin and Ethereum, within a decade. This goes doubly for double-bubble assets like NFTs.

  Crypto boosters claim that cryptocurrency has long-term value as a digital currency, but I disagree for the reasons outlined in Claim 3.

  Claim 2: Blockchain technology has no non-monetary applications (Confidence: High)

  While this mania seems to have calmed down as of late, I’ve seen a number of claims that blockchain technology – i.e. the ability for a network of people to maintain a distributed ledger without trusting a central authority or authorities – unlocks non-monetary uses, e.g. supply chain transparency. However, any use of blockchain technology in which the ledger is not self-contained, but is instead tied to the physical world, has the fatal flaw that there can be no trust-free link between the ledger and the physical world. If you don’t trust anybody to enter a delivery of wheat into your ledger, then your supply-chain ledger will be empty, but if you do trust certain people, then you are relying on the honesty of specific people or on other safeguards outside the ledger itself, undermining the purpose of a distributed ledger.

  Claim 3: Future monetary use of blockchain technology will be minor (Confidence: Medium)

  This is where it gets more tricky. Blockchain technology does unlock an interesting use case, which is the ability of participants in a network to maintain a ledger, or record of holdings, without trusting a central authority. In theory, this approach could replace fiat currency, which is controlled and tracked by trusted authorities like banks. However, I see two major roadblocks to the widespread adoption of cryptocurrencies:

  1. Inefficiency – Requiring every participant in a network to interact with a distributed ledger imposes substantial transaction costs which make cryptocurrency transactions more expensive than transactions in fiat currency, where the network is supported by trusted central authorities like banks and the government. To my knowledge, nobody has yet found a way to correct this flaw and it’s likely a feature of the way distributed ledgers work. If that’s the case, then cryptocurrency adoption will be limited to marginal use cases where fiat currency is impractical, e.g. online black-market transactions, since people prefer paying lower rather than higher transaction fees.
  2. Control of monetary policy – When the world was on the gold standard and central banks had limited ability to control the money supply, extreme and frequent booms and busts were a major feature of the economy and a constant detriment to people’s lives and livelihoods. Fiat currencies administered by competent central bankers have the attractive feature that their governance can be used as a tool to cool excesses and calm panics when the credit cycle goes through its inevitable revolutions. Cryptocurrencies, to my understanding, dispense with this notion of central control on purpose, and their adoption as the primary currency in an economy would subject that economy to pre-fiat-era booms and busts or worse.

  Unless both of these factors are somehow addressed, I predict that cryptocurrencies will play a marginal role in the future, if they exist at all. Even if they continue to grow in value as more speculators pile in, their usage as currency will be limited unless the inefficiency problem is solved. If that does somehow happen as well, we may see widespread use of cryptocurrencies in at least some countries, until a reckoning with the subsequent booms-and-bust dynamics, which may take decades, forces a return to fiat currency or its equivalent. (This could take the form of something which is a cryptocurrency in name only, with a “ledger” maintained by the government and banks.)

  I also believe that it’s unlikely we’ll get even that far, as governments will not be keen to lose control over monetary policy. Authoritarian governments are likely to restrict or ban cryptocurrencies if they get much bigger than they are now, and even democratic governments will have to weigh voters’ enthusiasm for cryptocurrencies with the importance of being able to control monetary policy.

  Some bubbles are caused by over-enthusiasm over genuinely innovative and productive assets, like the Railway Mania of the 1840s or the Internet Bubble of the 1990s. When those bubbles pop, they leave behind large amounts of investment in assets like rail and software that can be used for productive purposes later. (For more on this, see Carlota Perez’s Technological Innovations and Financial Capital, or a summary.) Other bubbles are triggered by enthusiasm over assets which have much less enduring value, like Tulip Mania, the South Sea Bubble, and Beanie Babies. It’s my prediction that the cryptocurrency bubble will turn out to have been another one of those unproductive bubbles, fueled by low interest rates and speculative enthusiasm from a wide base of retail investors rather than by a genuine advance in financial technology.

  March 18, 2022

  ---

• Regulating Resiliency in Supply Chains

  COVID has exposed just how vulnerable to shocks the systems we depend on for physical goods are. Just about every good – from lumber to computer chips – has been subject to shortages, and each shortage has reverberated downstream through supply chains to cause shortages in many other goods, as with the millions of cars that aren’t being made because of chip shortages. All these shortages are having a meaningful effect on consumers’ quality of life and on firms’ ability to produce goods and conduct R&D – both a short-term and long-term hit to our economy and well-being.

  Why are physical supply chains so fragile? Part of the reason is just that they’re so complex. A firm making a physical good likely sources components and tools from dozens, if not hundreds or thousands, of upstream suppliers. Often, many of those upstream suppliers are in a different country than the firm in question. It only takes an issue with one of those upstream suppliers – or with the ability to transport one of their products – to throw a firm’s operations off-kilter. In some cases, a shortage of a component or tool means scrambling to find an alternative supplier, but in some cases – as with chips – there may be no alternative, and the firm’s operations become limited by the quantity that a single supplier is able to deliver. Either way, production slows or stops, having cascading effects on the downstream customers that depend on the firm in question’s products.

  So, any link in a supply chain can cause disproportionate harm to the entire supply chain relative to its own size. But in many cases, the links – firms producing physical goods – are actually incentivized to be efficient to the point of fragility. Many physical goods, whether toilet paper or lumber, are essentially fungible commodities in near-perfectly competitive markets, meaning firms producing those goods must eke out every possible efficiency to defend their already-thin margins. This means running production equipment 24/7 at full capacity and maintaining stocks of inputs and inventory that are as small as possible to minimize capital outlay and storage costs. 

  This is, of course, a fragile way to do things, but it’s often the only economically-viable one, and firms are forced to risk being unable to meet a once-a-decade crisis in order to preserve day-to-day profitability. The trouble is that when a firm is forced to slow down or shut down by such a crisis, only a small fraction of the economic costs of the shutdown are borne by the firm itself – many more are borne by the supply chain and economy within which the firm operates because of goods left unproduced and unused. (a negative externality, in economic terms)

  There are also cases where a physical good, rather than being produced by many near-identical firms, is produced by a single firm, or only a few. This is the case with both computer chips and chip production equipment, with top-end supply dominated by TSMC in Taiwan and ASML Holding in the Netherlands respectively. Just as with the paper-thin operational slack in commodity goods, the concentration of production in goods like chips makes sense: for goods that are extremely expensive to produce in terms of fixed costs of intellectual property and physical infrastructure, a natural monopoly or oligopoly is the expected result. 

  The trouble is that when only one or a few firms are at a critical supply chain chokepoint, a single event can cause a worldwide disruption, as with the 2011 floods in Thailand that upended the hard drive industry. Moreover, firms operating expensive capital equipment are also incentivized to run at 100% capacity just like a toilet paper producer, meaning that there’s little room to absorb spikes in demand – as with TSMC struggling to meet chip demand during the COVID recovery.

  These kinds of vulnerabilities might seem inevitable, but there is room to reduce them with regulation. 

  Another industry, finance, is concerned with moving money rather than physical goods, but has similar vulnerabilities – individual financial firms are also incentivized to operate in risky ways that make sense for each of them individually but not for the system as a whole, since any firm’s failure can be magnified through the whole financial system and real economy.

  And indeed, the world of finance used to suffer from catastrophic booms and busts every decade or so, impacting the real economy of goods and services in major and painful ways. However, regulation in the last century has reduced, though not eliminated, the frequency of cascading failures in the financial system. A similar approach might be useful for physical supply chains.

  Of course, there are limitations to the analogy. In a financial crisis, the government can inject money it creates out of thin air to improve liquidity. Unfortunately, the Fed is not able to materialize computer chips or toilet paper out of thin air in the same way that it can manifest new dollars. So, regulation for physical supply chains should be aimed at making them more resilient before a crisis strikes.

  There are a few ways in which this can be done.

  One is the “too big to fail” principle, or ensuring that no single firm is so big, or so much the exclusive supplier of some critical good, that its failure would cause the whole system to crumble. In addition to sheer scale, in physical supply chains, there is also the importance of international borders – even if there are many suppliers of a good, but all are located in a single foreign country, then it is that country’s government’s willingness to export that good that becomes a single failure point.

  Thankfully, the US is already starting to wake up to this concern. Projects like the new Intel plant and new TSMC plant, both to be located in the US, are reducing vulnerability in chip supply chains. Similar projects, supported by the state both politically and with investment capital, could reduce vulnerabilities in other key supply chain chokepoints.

  Another issue in physical supply chains is somewhat analogous to bank reserve requirements. Banks are incentivized to keep relatively little money on hand – what is not lent out is not earning interest. But low reserves increase the fragility of the whole financial system, since banks with low reserves are more vulnerable than those with ample ones, and a single bank failure can cascade through the whole system.

  Similarly, firms producing physical goods in highly-competitive spaces are incentivized to maintain very low inventories of both inputs and finished goods, and very low if any excess production capacity. But just like with banks with small reserves, goods-producing firms with small inventories and small excess production capacities are vulnerable to crises and cannot meet bursts in demand. Just like with banks, this makes the whole system more fragile.

  One can imagine the government stepping in and instituting similar “reserve requirements” for goods-producing firms as for banks. For example, firms over a certain size might be required to hold a larger inventory buffer of inputs and outputs on hand than would be economically rational, and to be subsidized by the government for the additional costs incurred, just like banks are paid interest by the government on their reserves. 

  In addition, the government might incentivize or require firms over a certain size to maintain excess production capacity. This could take the shape of a Federal program where firms were periodically required on short notice to produce an extra 20% of their annual production of a good within some number of months, which would then be bought at above-market rates by the government. That’s one example, but I’m sure those more familiar with industrial regulation could imagine even better ones.

  Regulating companies producing physical goods is no easy matter, and any regulation that improves supply chain resiliency would impose costs on taxpayers. But if done properly, it could also reduce our vulnerability to future supply chain shocks and serve as a form of worthwhile insurance to our economy. As we navigate the COVID supply chain crisis, it’s worth considering how to reduce the severity of the next one.

  October 8, 2021

  ---