The Biggest Crypto Fantasy Of All
- “Spend cash, invest in Bitcoin. Cash is trash.” – A ‘bon mot’ from one of the Winklevosses
- “As of the end of Q2, we have converted approximately 75% of our Bitcoin purchases into fiat currency,” – Tesla Quarterly Earnings Statement (July 2022)
- “Bad coin and good coin cannot circulate together.” – A version of Gresham’s Law
This week brought news of “The Merge” – a purportedly momentous singularity in the history of humankind, or at least a big new loop in the tangled story of crypto-currency. We are told that, from now on, Ethereum (the second most popular crypto offering) will no longer be based on “Proof of Work” but rather on “Proof of Stake” (whatever that turns out to mean). The big selling point of the Merge is that the cost of “mining” the stuff will be reduced. In fact, mining won’t even have to happen anymore (too bad for Nvidia, by the way, which has floated its market capitalization on sales of the mining hardware – its share price is 92.88% correlated with the price of Bitcoin since November 2021).
In any case, “Merge parties” are said to have broken out world-wide, live-streaming from every time zone. Meanwhile, the hedge fund world has crowded into the options market to trade on the volatility that they expect.
- “‘A lot of smart money [is] buying’… Outstanding option contracts have risen from 1.2mn at the start of the year to more than 4.6mn by Wednesday. About 80 per cent of these contracts are call options… It’s a sign of ‘massive bullish sentiment.’”
Of course, the existence of “massive bullish sentiment” would normally signal a decline in the market (investor sentiment is usually a contrarian indicator). The next few weeks will reveal how the Merge will play out for speculators. But we already know that in practical terms it won’t make any difference for the long-run fundamental prospects of crypto.
MORE FOR YOU
The Flaws in the Model
There are so many things wrong with crypto-currency it is hard to know where to begin. Whack away at this piñata of fallacies and false hopes, and the goop comes gushing out. The ugly stuff has to do with what could be called the “moral flaws” in the crypto model: the facilitation of illicit transactions, the money-laundering, the ransomware extortions, the drug dealing, porn, and god-knows-what. Then there is the astonishing wastefulness of the mining process by which crypto tokens are “created.” Prior to The Merge, Ethereum’s carbon footprint was said to be “roughly equivalent to that of Finland.” Bitcoin’s climate-damage price tag is evidently somewhat larger, on a par with the total energy budget of Sweden. Add in dozens of hacking scandals, and background everything with the pain suffered by naive investors dabbling in the totally unregulated crypto market. The overall Ponzi-ish nature of the phenomenon has been clear from the beginning. (A typical Bitcoin come-on was the one proffered by the Bitcoin Savings and Trust – in 2011-2012 – which promised investors up to 7 percent weekly interest. It ended predictably and badly.) Today even crypto proponents admit that the “value” of their holdings is based on the Greater Fool showing up when they need him. Last Fall, the GF disappeared – and the price of Bitcoin fell by 70% (though not quite as bad as the Chinese stock market — another field of dreams – demonstrating that centralized and decentralized economies can match each other in value destruction).
Future historians will post the “What Were They Thinking” narratives of the crypto-currency bubble alongside the Dutch Tulip Mania and Bernie Madoff in the gallery of scandal and folly.
But aside from all that – there is a deeper problem. Crypto doesn’t work. It does not and cannot serve a valid economic purpose. Not as a store of value, obviously (see previous chart). Not as an inflation hedge. Nor can it compete with the dollar, nominal or inflation-adjusted. Fiat money beats DeFi hands down.
But the most important failure of crypto-currency, however, is its most practical failure: as a potential medium of exchange.
Can you buy things with it?
The short answer: You really can’t. For two reasons.
First is the question of “uptake.” Where can one actually pay for a real-world product (not just another crypto-object) with a bitcoin or similar? There is a lot of noise about that, and there are said to be workarounds for vendors like Amazon that won’t accept bitcoin directly. But the fact is that after a dozen years of promises crypto-currency qua currency is still not really fungible. It is steeply discounted in those few transactions where it may ostensibly be employed.
A currency that can’t be traded on, can’t be spent, fits the definition of “bad coin” under Gresham’s Law. In the presence of “good coin” (e.g., the U.S. dollar) crypto can’t circulate. Even in El Salvador, which has declared Bitcoin its national currency, and even in “Bitcoin City” – which is El Salvador’s showcase for the crypto-economy – the Barron’s reporter who went to check things out found that “the money was nearly useless; only 3 merchants out of 10 that we met in Bitcoin City would accept payment in Bitcoin.” Crypto remains a speculative vehicle, with extremely limited convertibility to the real economy.
Even that is not the worst of it. One could imagine, or at least argue, that over time the adoption of crypto by merchants in the real economy might somehow improve, perhaps with certain safeguards. That is the premise of the Stablecoin concept, where crypto values are allegedly guaranteed by a peg to real money, i.e., U.S. dollars. Rather like the relationship of the dollar to gold in the old days.
But there is a flaw in the crypto framework which neither the Merge, nor the Stablecoin kluge, do anything to fix. It is the extremely low transaction processing capacity of the underlying blockchains that support Bitcoin and Ethereum. For Bitcoin, the limit appears to be about 7 transactions per second – for the entire planet. For Ethereum, the limit is said to be around 15-20 transactions per second.
Apple sells about 10 iPhones every second. That would use up Bitcoin’s capacity right there. On Amazon, the average number of transactions for is 18.5 per second throughout the day – much higher during busy hours. Or perhaps it is 100 orders per second. In the Christmas season, the volume reaches 300 orders per second. Walmart processes something on the order of 100-200 transactions per second. No crypto-currency could even cope with Walmart’s or Amazon’s current level of business – let alone the transactions flowing in from tens of millions of other vendors around the globe.
The extremely low limits on transaction processing capacity for Crypto translate into higher costs, and long delays. Barron’s’ reporter found that even a simple funds transfer with Bitcoin took six hours to execute. (Consider how long you might wait for a credit card transaction to confirm at the supermarket before you lost patience.) As to cost, Ethereum transaction fees averaged about $2.00 recently — but earlier this year the fees were in the multiple tens of dollars per transaction with odd spikes reaching as high as $200. (It is not clear that the Merge will change this.)
Current reality-based payment systems like Mastercard, Visa, and American Express process about 1250 transactions per second on average. Peak volumes can be much higher, and these real-world systems are designed to handle it. Visa is said to be able to process up to 65,000 transactions per second. Alipay (Alibaba’s version) has been said to have achieved actual payment volumes as high as 250,000-500,000 transactions per second during some of the key shopping days in China.
Crypto promoters have their answers, in the form of future plans. In the next big Ethereum modification, something called “sharding” will be introduced – which will split the blockchain into many pieces and increase the tps rate to 1000, supposedly. The following modification may involve further reconstruction of the eco-system, with “roll-ups” — and will bless the waiting world with volumes up to 100,000 tps. Then there is Solana, which supposedly has a “proof of history” mechanism for validating its blockchain, and says it can achieve much higher throughput:
- “Solana has been gaining traction… Its blockchain processes 50,000 transactions per second, and its average cost per transaction is $0.00025, according to its website. Ethereum can only handle roughly 13 transactions per second and transaction fees are substantially more expensive than on Solana. But the last year and a half has laid bare [Solana’s] trade-off as the blockchain network has suffered multiple outages. Most recently, on May 1, Solana locked up for several hours before it was similarly brought back online following a restart of its validator network.”
The breakdowns have proliferated:
- “On 14 September 2021, the Solana blockchain went offline after a surge of transactions caused the network to fork and different validators had different views of the state of the network. The Solana blockchain again went offline on May 1st, 2022 and May 31st, 2022. The [first] outage lasted 7 hours, and the [second] one lasted 4.5 hours. On the 3 August 2022, the Solana ecosystem was targeted by hackers, affecting 9,231 Solana wallets [stealing] $4.1 million in total from victims. On July 1, 2022, a class action lawsuit was filed against Solana.” (Wikipedia)
It would be interesting to know the scope of the disruptive “surge” that seems to have broken the Solana system.
Blockchain is Not Designed for This
No crypto-currency has anything like the capacity to serve as a medium of exchange in the modern retail economy. Blockchain is an interesting and useful technology, suitable for small networks, centrally administered, that need quick and reliable execution on high-value transactions. The Onyx system developed by JP Morgan for settlement of wholesale inter-bank repo transactions is an example that seems to be successful. But blockchain cannot currently address the capacity requirements of massive retail payments systems.
Neither Bitcoin nor Ethereum as they stand can serve as a medium of exchange. The Merge, whatever it amounts to, does not change that.