Is Bitcoin dirty?
This month Bank of America Global Research issued a 49-page report on Bitcoin titled “Bitcoin’s Dirty Little Secrets” that comes across as a hit piece.
It has five main sections:
- What drives Bitcoin prices? 12 pages
- How does Bitcoin compare to traditional portfolio assets? 9 pages
- How does Bitcoin score on ESG? 11 pages
- Are CBDCs Kryptonite for crypto? 6 pages
- Is DeFi potentially more disruptive than Bitcoin? 7 pages
Right on page 1, the executive summary, the plan of attack is laid out clearly (italics mine):
“Bitcoin supply is artificially scarce, demand drives prices”
“No good reason to own BTC unless you see prices going up”
“Our Bitcoin ESG read: low on E, mixed on S and G”
“CBDCs are kryptonite for crypto, but DeFi is intriguing”
These follow the outline, with the last statement collapsing the last two sections.
So let’s look at these in turn.
“Bitcoin supply is artificially scarce”
I object, as does the Bitcoin community, to the term “artificial.” The scarcity is by design. It is a deliberate feature, not a bug. For money to fulfill the store of value requirement, it must be scarce. Indeed Bitcoin’s “minting” or “mining” or emission schedule was designed with two main principles: unlike fiat currency, the supply must be finite.
Last year the M2 money supply of the US dollar increased by 26%. In a single year. The money supply has been increasing every year, but 2020 set a new record since the Federal Reserve was created over a century ago. This was largely in response to COVID-19, typically in recent years, the supply has grown in the range of 4 to 10% since the Great Recession. However, the outlook in the next few years is for significant fiscal stimulus and large deficits, which will push M2 growth higher than during the prior decade.
The M2 money supply grew by 26% in 2020 (source: Federal Reserve)
The development of Bitcoin was motivated in part by the continuing inflation of money supplies by central banks around the world, and the growing instability of the fiat system, evidenced by the Great Recession and by squeezes in the Repo market and Eurodollar market (overseas dollars, not Euros).
As the authors point out, the Bitcoin supply rate drops in half every four years. Technically this is artificial, although what it really is is an algorithmic enforcement of scarcity. It reflects a clear design principle by Satoshi Nakamoto of allowing for rather rapid expansion of early supply as Bitcoin was just getting its footing, then continual reduction over many decades to a lower and lower inflationary environment. This long-term plan allows the Bitcoin-related crypto economy to stabilize around a disinflationary ethos.
Bank of America lives in and benefits from the fiat inflationary world. Due to their banking license, they are able to create money instantly with every new loan they give, in a fractional reserve system. (Strictly speaking, the fractional reserve limit was eliminated as of a year ago, but Basel III equity capital standards provide constraints on the loan book.) Banks benefit by being first in line as new money is created before it gradually loses value year by year. In recent years low-interest rates have cut into their profit margins, but long-term rates are rising again.
There are a lot of nice charts in the BofA Global Research paper. In particular, they look at institutional demand, something that became much stronger in 2020. They are trying to show correlations of Bitcoin price to things like Google trends, institutional announcements, flows into GBTC trust, but with charts that span only a few weeks, seemingly targeted toward traders. This is not surprising as a commodity strategy paper.
The authors, however, fail to discuss to any extent the overwhelming long-term drivers of value in increasing stock-to-flow, increasing hash rate/difficulty and thus security, and the rapidly growing network effects due to the growing adoption and ease-of-use of Bitcoin from one year to the next.
“No good reason to own Bitcoin unless you see prices going up”
The second section starts out by talking about Bitcoin’s “high concentration ownership ratio.” They state that 2.4% of the addresses have 95% of the Bitcoin. While true, this fails to recognize the actual beneficial ownership. A large fraction of Bitcoin is held in exchange addresses or in the GBTC trust or in corporate accounts that have many shareholder beneficiaries. These balances, like Bank of America’s deposits, represent a large population of users. It’s like ignoring that many retail owners of Amazon stock have the shares held in street name at their broker.
A very large money center bank attacking Bitcoin’s GINI coefficient, incorrectly calculated, is rather rich. On the other hand, Bitcoin’s market cap is now three times that of the Bank of America.
The authors then go on to look at Bitcoin’s high volatility, which exceeds that of gold and many emerging market currencies. This is true, but Bitcoin’s high volatility is convex, favorable to the owner. Emerging market currencies are generally concave, dropping in price over the long term. Bitcoin has a strong upward bias in its value, most emerging market currencies have a strong downward bias relative to the dollar.
Bitcoin has also been strongly outperforming gold. It has been rising faster than exponentially in terms of gold ounces and is now worth around a kilogram of gold per Bitcoin. Their paper points out that the notional dollar value of Bitcoin trading is higher than that of the GLD ETF, one of the most popular paper gold vehicles.
I also give credit to the authors for noting that the market cap of gold (around $12 trillion) or the size of the Fed’s balance sheet ($8 trillion) are interesting reference points for Bitcoin ($1 trillion market cap). These are interesting potential targets as Bitcoin’s 21st-century monetary technology accretes value from the existing financial system that was built in the 20th century.
The Federal Reserve Balance Sheet has jumped from $4 trillion to nearly $8 trillion with the COVID-19 response.
The authors examine a number of correlations with Bitcoin and other assets. But over long periods, their data shows that Bitcoin correlations with other asset classes are very low, less than 0.2, and with some, slightly anti-correlated. Nevertheless, they state that Bitcoin is not for “diversification or inflation protection, but a rather sheer pice appreciation.” The reality has been that Bitcoin’s shorter-term correlations with other assets tend not to persist.
What does persist? Bitcoin has the highest return over the past decade of any major asset class. Long term, it is crushing the dollar, other currencies, treasuries, stocks, commodities, real estate, and gold. Its long-term rapid price appreciation is actually the surest protection against long-term inflation.
The authors note that money inflows have a large impact on Bitcoin prices.
This is a good thing for those who can handle the volatility, lots of bang for the buck. Again, it’s a highly convex asset whose performance per $million of money flow into the asset is over 20 times higher than gold or long-term treasuries, according to their own analysis. New institutional money is coming into Bitcoin and is highly impactful on its price.
“Bitcoin ESG read: Low on E, mixed on S and G”
The main front of their attack on Bitcoin seems to be ESG: environmental, social, and governance issues. Especially environmental. “Enormous environmental costs” (on page 24). “Bitcoin’s annual energy consumption now rivals that of some small developed countries like the Netherlands and the Czech Republic.”
This shibboleth has been repeated by many in various forms. I saw another bank report that claimed Bitcoin’s energy consumption equaled that of Japan, which is completely ludicrous.
The nuanced reality has also been discussed by many Bitcoin experts. One of these is Nic Carter, whose findings are summarized here: https://www.coindesk.com/the-last-word-on-bitcoins-energy-consumption .
I personally have written several articles on this issue over the last few years, including here, here, and here. The first of those was written three years ago in response to a prediction that Bitcoin was going to consume a huge share of the world’s electricity. It is still well under 1% of the total. The dire predictions of a UC Berkeley lecturer in computer science have yet to be realized.
But how much energy does gold mining use, how much electricity does the world’s financial system use? In the aforementioned article, I noticed that a 5% efficiency increase in gold mining would cover all of Bitcoin’s electricity needs at that time.
Now Bitcoin’s value has risen a factor of nine since that article was written. That seems to have been a very good investment in electricity. Once a Bitcoin is mined, it is quasi-eternal, like gold.
Bitcoin mining is largely operating at the most efficient frontier of energy production, which is increasingly renewables. It has been estimated that over one-half and perhaps as much as 70% or more of Bitcoin’s electricity is from renewable sources. Hydropower, in particular, is one of the lowest-cost sources, utilized in places like Sichuan province, Quebec, and western Washington state.
The BofA authors note that the major share of Bitcoin mining happens in China. I do twice yearly reports on Bitcoin mining, here is the latest: https://medium.com/the-capital/crypto-super-top-50-fifth-edition-bbce1d15a6bf. Mining pools are gradually becoming more globally distributed. The authors make the gross error of assuming Bitcoin mining in China comes mainly from coal when large amounts of mining happen at cheap hydropower locations in Sichuan and Yunnan. They quote a Cambridge Center for Alternative Finance study noting that lots of Bitcoin mining happen in Xinjiang but that only sampled 37% of the IP addresses. And Xinjiang also has significant hydropower and natural gas sources.
The Chinese government in some provinces has restricted crypto mining https://supchina.com/2021/03/09/bitcoin-mining-is-still-huge-in-china-despite-new-ban-in-inner-mongolia . However, data is hard to come by since there are many mining operations, even though they combine into large pools. An American sitting in his Mom’s basement can also put a mining rig into a number of these pools.
The newest trick by Bitcoin miners is to protect the atmosphere and utilize very low-cost electricity by capturing (otherwise) flared natural gas at the source. Multiple companies have installed generators in gas fields in Alberta and west Texas to run Bitcoin ASIC miners in situ. This is close to free electricity and is about as green as one can get. Bitcoin is also an excellent load-leveling tool since miners can easily stop operating when electricity is expensive at certain times of the day. This strategy is being adopted in Kazhakstan also, next to China.
If crypto mining electricity utilization gets too high in certain localities, those authorities will restrict their electrical consumption (this has already happened in certain locations in the US also) and miners will have to move elsewhere or switch to a lower hash rate (less valuable) cryptocurrencies. Even if the Bitcoin mining electricity budget was globally capped, it would not change the rate at which Bitcoins are produced, currently 6.25 Bitcoins per 10-minute block, regardless. It might increase transaction fees for recording onto the blockchain somewhat.
The real problem with the Bank of America authors is they don’t compare Bitcoin’s electrical consumption to that of the banking system and financial industry. The financial industry is a huge consumer of computer power and electricity, and of office space in central business districts. In the US, the financial industry accounts for 21% of the US GDP if you include real estate finance.
It is reasonable to assume the financial industry uses over 10% of the electricity consumption of the US, since commercial use is about half of all electricity. It also consumes energy in other forms to run its offices and branches. I guarantee you that moving $1 million in cash by armored car requires much more energy per mile than moving the same amount across the Bitcoin blockchain. Try doing that for $36 over a distance of thousands of miles (see next paragraph). Moving money by bank wire is much slower (days) and more expensive as well.
Bitcoin is making the financial industry more efficient by automating processes such as international money transfers, driving out middle layers, and acting as a final settlement asset for cash, bonds, and other instruments. The value of transactions carried on the Bitcoin network ranges from about 1000 to over 3000 times the value of the Bitcoin mined per block. Checking blockchain.com, the latest block as I write this is #676938 (2021–03–30 12:09 UT), the block reward is 6.25 Bitcoin, but the total transaction value carried in the 10-minute block is 20,386 Bitcoin with a value of $1.172 billion, spread across 2350 transactions. The average transaction was $498,623 for an average fee of just $18. Bitcoin can move several million dollars across the globe for a single transaction in one hour (six confirmations), any time of day, for less than the cost of a bank wire. This may be BofA’s real concern.
Their next topic raised is S, the social issues. They agree that Bitcoin has the potential for the democratization of money and benefits for small businesses and for payments by individuals. It also provides pseudo-anonymity. They then express concerns: not everyone has access to the Internet (but over half of the world’s adult population has a smartphone), volatility of the price, and the need for regulation.
This next one is interesting, they say blockchain illicit activities totaled 1.9 billion USD in 2020. But that was across all cryptocurrencies, and most of it was frauds, probably overwhelmingly attributable to altcoins or problems with exchanges. The Bitcoin blockchain itself is clean, has never been stolen from, and cooperative by design. Like any other monetary asset, there will be bad actors trying to separate people from their money. Cryptocurrency regulations and enforcement have strengthened steadily over the last few years.
The last is G, for governance. It’s a bit strange for a money center bank, in a monopoly central bank-driven system, to be deeply concerned about whether Bitcoin is sufficiently decentralized. There are four money center banks in the US, which hold roughly half the deposits in the country.
The analysts’ concerns are the number of large mining pools running transactions and the size of the core developer community. While four mining pools earn half the new Bitcoin rewards, they represent many firms and individuals who place mining rigs into the pools through Internet connections unlike large money center banks, they hold onto only a small fraction of the Bitcoin “money supply.”
Bitcoin has a governance system for a voluntary community that looks like this: the constitution is the Nakamoto consensus and the executive branch composed of miners commits transactions to the blockchain. The developers and the community at large have the legislative function since changes require buy-in not just from developers but miners and exchanges by their signaling of support for protocol proposals. The judiciary is the thousands upon thousands of full nodes that maintain copies of the entire blockchain to authenticate and audit transactions. These full nodes can cost as little as a couple of hundred dollars to set up and can function as senders and receivers of Bitcoin transactions, including micro-transactions on the Lightning second layer.
The exchange industry is young, and no doubt some are prone to market manipulation. As the value of Bitcoin grows and the industry matures, there is increased liquidity provision to smooth out irregularities. Decentralized exchanges help abstract away manipulation and can be used to quickly move liquidity between exchanges.
“Are CBDCs Kryptonite for crypto?”
Their fourth question suggests Bitcoin could be under attack by CBDCs, but they use a sleight of hand in the formulation, changing it to ‘crypto’? A strange question with an attempt to be cute, actually. Which crypto? Stablecoins? Not Bitcoin.
CBDCs will end up being an on-ramp for Bitcoin savings and trading, as stablecoins have been.
Here is a taxonomy. Governments produce fiat, which is a poor store of value and that operates primarily as a medium of exchange. CBDCs will be just another form of fiat tied to the national or transnational currency (e.g. Euro) or global reserve currency (dollar) held in centralized double-entry ledgers. Actually, this is the opposite of private, decentralized monetary technology implemented by Bitcoin, which has advanced to triple entry ledgers widely dispersed in thousands of copies.
In other words, CBDCs are each likely to be a government stablecoin that is stable in the short term and whose value inflates away in the longer term.
Media of Exchange and Stores of Value. It’s a spectrum, money and money-like assets have both properties, but tend to lean toward either MoE or SoV.
What will CBDCs look like? Will they operate on a single level, incorporating retail usage directly or in a hierarchical fashion like the present monetary system? If the CBDCs allow for direct retail (business and individual) wallets with deposits held at the central bank, then this is a clear alternative to commercial bank deposits. CBDCs are much more in competition with those and with non-governmental stable coins.
And central banks may choose to go that route in order to “bank the unbanked”, in order to more efficiently distribute relief payments (welfare, pandemic relief), to speed up and make more efficient all payments in the economy, and to broaden the reach of monetary supervision and monetary policy implementation. Direct retail wallets would allow interest payments to the population or even negative interest rate charges during low growth periods for an economy, as we have seen on larger deposit accounts in Europe.
The authors wait until the very last paragraph of this section to note the following potential “drawback”: “Banks could see cheap deposit funding curtailed.” Cheap means they pay low interest to the depositors. The report also notes that there could be new risks of runs on the commercial banks.
Could CBDCs be kryptonite for commercial banks like the Bank of America?
“Is DeFi potentially more disruptive than Bitcoin?”
DeFi is, of course, decentralized finance, which disaggregates financial processes and functions and moves them toward more decentralized models, ledgers, and workflows.
There are many ways in which DeFi can develop but the authors focus on Ethereum based platforms. Because of its “Turing complete” smart contract capabilities, Ethereum has been the most popular platform for the development of ICOs, dapps, asset tokens, and decentralized exchanges.
The most high-profile activities in the past year have been cryptocurrency-based lending and NFTs, non-fungible tokens used to tokenize art and collectibles, and potentially many other asset categories as well.
Other services possible under DeFi are derivatives (options, etc.) and asset management, including custody and legacy planning.
The authors correctly note that the scale of DeFi is not yet huge, perhaps $35 billion in assets tied up at present. There are also concerns about the scalability on top of Ethereum, which its community is actively seeking to scale by adding sharding, rollups, moving to Proof of Stake.
They also note that true credit is not being created since smart contracts don’t have banking licenses. However, there is a very active crypto collateralized lending market.
They actually do promote the idea of a distributed ledger central bank rail for settlements within the banking system.
What the BofA authors don’t mention is that it is increasingly possible to implement DeFi applications on Bitcoin, given second layer solutions like Lightning and side-chain solutions such as Liquid. Discreet log contracts that take advantage of external oracles can allow lending functions, for example, over Bitcoin.
Again the question posed is a bit strange. It is like asking if automobiles will disrupt roads. Certainly, that was relevant when roads were of lower quality and only traversed by horses. But the two technologies go hand in hand, they are different layers of a transportation system, or in our case, a financial system. Automobiles drove progress in road technology.
Ethereum is one type of road or base layer, Bitcoin is another.
Bitcoin will continue to be the most important store of value in the cryptocurrency world, with the largest market cap. The chance that its market cap will succumb to that of Ethereum seems rather remote, especially with Ethereum moving to Proof of Stake, which increases centralization, weakens security, and removes an effective floor on the price due to the cost of mining. Ethereum and some of the other altcoins with similar functionality will continue to grow in importance for payments and as media of exchange.
And stablecoins and CBDCs will also attract much use, again as media of exchange, certainly not as stores of value.
What DeFi is most disruptive for is the existing financial system, the existing network of commercial banks and exchanges, including the Bank of America. Perhaps they are seeking to buy time by knocking Bitcoin, since Goldman Sachs, PayPal, Square, and others seem to be ahead of them in implementing Bitcoin availability for customers.
Is Bitcoin dirty?
Is Bitcoin dirty? No, it’s a pristine asset, a final settlement asset, quite different from debt-based fiat. It’s intrinsically much cleaner than the fractional reserve fiat system that is fraudulent by design. Banks create and lend money they don’t have, and the government treasury / central banking / commercial banking system erodes the value of said fiat inexorably every year, every month. And without a doubt, Bitcoin is more efficient energy-wise and faster in moving larger sums of money around than current banking networks.