Usability and volatility prevent Bitcoin’s use as a medium of exchange today
Technology and time will allow it to be used as such.
This is the second of a five-part series. Read here the first part.
The first part of this series covered the economics of money and its evolutionary stages. It also demonstrated how the willingness to hold cash balances is what gives money value. However, this is not meant to be a case against Bitcoin’s use in commerce nor a defense for it to be solely a SoV never to be used in exchange. Of course, and absolutely not.
But, I do think that trying to use Bitcoins (or any other cryptocurrency) today in everyday commerce makes little sense due to volatility and usability (e.g., confirmation times, fees, TX throughput, available wallets, and custody). The latter will be solved through technology (better software, apps, payment protocols, and second layer solutions all addressing the “how”). The former, through the passage of time (the “when”), which will bring increased knowledge and, thus, confidence leading to broader adoption.
Usability of Bitcoin as MoE
First, let’s understand usability with a focus on capacity.
Bitcoin’s base layer, at the protocol level, cannot scale exponentially. Linear and incremental scaling is possible, and it will continue to be upgraded to accommodate more throughput along with other optimizations. For onboarding billions of users, however, exponential scaling is necessary, and blockchains are unfit for this need. Any cryptocurrency claiming otherwise is giving up some essential features to accomplish this, although its creators may not disclose or even realize it. On-chain exponential scalability with paramount security and decentralization is infeasible, there are always engineering trade-offs.
Many blockchain projects strive, at all costs, to sidestep the inevitability of protocol development in layers, similarly to the Internet’s protocol stack. While this article is not about scalability, since the concept is so intertwined with the SoV vs. MoE debate, it cannot be avoided.
Protocols on top of Bitcoin are a logical outcome of its technical on-chain limitations. To date, the Lightning Network (LN) is the most promising secondary layer solution for increased transaction throughput and reduced confirmation times. This approach will not be the only nor the last of possible decentralized and trust-minimized layering solutions for scaling. What most people fail to comprehend is that the LN protocol does not create a new coin as LN is not a currency network. Instead, it is a payment network with no currency issued on it and no coin transferred through it. LN routes payments by transferring rights to on-chain balances, which are cryptographically linked to specific “unspent transaction outputs” (UTXOs).
LN builds on precursor ideas like DigiCash’s Ecash system, which allowed for private, untraceable payments with its native UoA (cyberbucks) as well as with national fiat currencies via traditional banks. A payment network can then provide for greater capacity and privacy.
Developing Bitcoin through layers makes it resemble the current financial system, but only from a functional point-of-view, and not one of security, as there will be fewer (if any) trusted third parties that are unavoidable in the existing monetary order.
Every national fiat currency is unscalable by design. Even gold is unscalable by itself. Banks are essentially payment networks working on top of fiat currencies enabling the exponential use of base money. PayPal is another payment network providing a greater use of fiat currencies in the digital world.
Although economists argue commercial banks create money out of thin air, the reality is not the case. Banks issue liabilities, or deposits that can be redeemed on demand for fiat currency (base money). Bank liabilities are promises to pay (comparable to off-chain transactions) and are only settled when paper currency is redeemed (on-chain transactions).
On the opposite side of the equation lie the assets a bank holds with varying degrees of liquidity, maturity, and risk to fulfill its obligations. Cash, securities, loans, and other assets are perfectly acceptable instruments in the market with which a bank must manage its liquidity to meet its liabilities sufficiently and prudently. Thus, demand deposits are not money per se, despite being used by individuals to transact in the economy. Bank deposits may function as money, but they are not money. This is why demand deposits are not legal tender, it would be ludicrous to grant such a privilege[1] to commercial bank debts.
So, our current financial system features a myriad of payment networks (hierarchical, parallel, subordinate, and, often, overlapping) which is fraught (for better and worse) with trusted relationships. Precisely with this characteristic — the need for trust — lies the innovation of the Lightning Network, since payment channels are not promises to pay, but are bilateral contracts linked to specific committed and time-locked UTXOs powered and secured by clever use of cryptography.
When envisioning a stack protocol development for Bitcoin, similarities with the current financial system from a functional perspective include:
1) Base layer protocol or high-powered money, base money, and outside money: gold, fiat paper currencies, and Bitcoin.
2) Secondary and outer layer protocols or payment networks: banks, financial institutions, PayPal, ApplePay, WeChat Pay, and the Lightning Network.
The distinction between a financial system based on fiat currency (or gold) and one based on Bitcoin is the security models of trusted interconnected parties vs. trust-minimized networks or centralized trust vs. decentralized trustless, which are valid for both the base layer protocol (fiat currency vs. Bitcoin) and the outer layer protocols (banks vs. LN). Gold’s “base protocol” is akin to Bitcoin’s as it is no one’s liability.
Hence, to claim that the LN replicates the traditional banking system with crypto is to miss the most glaringly evident difference. The former is an open and trust-minimized network of bilateral contracts where no party unilaterally has custody of the underlying assets. The latter is a closed and trusted network where centralized parties and customers contractually agree on the transfer of ownership and control of the underlying assets in exchange for a promise to pay on demand (in a nutshell, the relationship between bank and depositor). So, LN payment channels are contracts between two parties with equal rights, whereas the nature of a contract between a bank and depositor is a debtor-creditor relationship.
Now that we have established the similarities and distinctions between each financial system, we turn to answer the question of how Bitcoin can be used as a MoE.
Ponder the following: today, when you make a payment through your bank or PayPal, would it be correct to say you are using the dollar? Is the dollar of a wire-transfer used as a MoE? Paper dollars are not physically handed over to the other party, right? We could conclude that the dollar is not a MoE in this transaction. The actual transaction is a transfer of rights over “promises to pay paper currency on demand.”
Technically, the above is true. But, this observation is irrelevant for all practical matters. Yet, this line of reasoning is equivalent to BCHers’ view that Bitcoin transacted off-chain means it is not being used as a MoE.
The truth is money is indeed a MoE, and so is the dollar, but that does not mean every and any transaction in an economy should occur with paper currency directly (or with gold or on-chain). This scenario does not scale, and payment networks are needed with payments linked to (or backed by) fiat dollars. So it is immaterial if we use physical currency or a payment network to make payments.
If money’s function of MoE only manifests itself when two parties physically hand over paper currency (or gold), we could make the case that money is rarely used as a MoE, especially in advanced economies where cash payments are increasingly occasional. Base money is currently so segregated from everyday economic life that it lacks practical significance, from this base money perspective, attempting to spot if the MoE function is present. Most payments are effected not with currency but are denominated in it and use it as the asset through which promises to pay may be settled.
Paper currency forms the foundation (i.e., it is the monetary standard) of our financial system on top of which payment networks operate (i.e., off-chain) and with which payments are settled (i.e., on-chain). The dollar, then, is used as an SoV — though precarious over the long run — as a MoE and UoA.
The same working model applies to a Bitcoinized financial system. Bitcoin can be used as SoV, MoE, and UoA (or, at least, it is theoretically possible). But, this does not mean every and any transaction must occur on-chain. Bitcoin can (and will) be used as a MoE. However, over the long run, it is possible a significant amount of transactions will take place over payment networks which transfer rights to on-chain balances. These rights may or may not be promises to pay (i.e., we may have LN and traditional banks, just as we already have centralized exchanges) that can be settled on-chain through the transfer of UTXOs.
Currently, Bitcoin is already spectacularly suitable for secure and seamless global transactions. For medium to large transactions, it is an excellent “medium for wealth transfer,” to borrow from Szabo’s terminology.
So, although frequent and small payments might eventually take place primarily over payment networks, they do not have to be realized in actual Bitcoin but are denominated in it and use it as the asset through which bilateral contracts (e.g., LN or trust-minimized) or promises to pay (trusted third parties) may be settled.
In this sense, the base protocol can be considered the final settlement layer — the layer on which larger transactions tend to occur — whereas the secondary layers will cater to recurring and smaller payments.
So, with this approach, we answer the question of how Bitcoin will be used as a MoE.
Liquidity, SoV, and transaction costs
Let me now turn to volatility, the other fundamental issue.
There was a growing trend from 2012 to 2015 to onboard merchants to accept Bitcoin as a form of payment. Bitpay and Coinbase began as intermediaries, or payment gateways, enabling customers to pay with Bitcoin while merchants could opt to receive in local fiat money or Bitcoin. Of course, most merchants preferred local currency.
Don’t get me wrong, as I do commend both companies for these initiatives. Nonetheless, it was a self-defeating purpose since no sane merchant would risk its daily business operations by holding a cash balance that could fluctuate considerably within minutes. Even though price volatility has declined in recent years, it remains a relevant impediment for traditional business to accept and hold Bitcoins today.
If merchants are not holding Bitcoins due to lack of trust in the currency and unwillingness to live with its unpredictable short-term price swings, then they are not demanding Bitcoins and contributing to the stability of its value. Instead, they are quickly unloading it back onto the market once received, which effectively implies a reduction in the demand for Bitcoins[2].
This scenario is why forcing merchant adoption is futile today and should not be a priority for the community. Pushing for its use primarily in exchange now is putting the cart before the horse. First, we need Bitcoin to establish itself as a good SoV. Then, MoE will soon follow as a natural consequence. Therefore, instead of merchants simply allowing for customers to spend their coins, we need them to want to be paid in Bitcoins, to hold Bitcoins, to hold cash balances in Bitcoin.
So, how do we break away from the circular trap of merchants not holding Bitcoin because of volatility while this volatility is due to demand being fickle? Through education and understanding, solid protocol development, and, above all, time. Sound economics can furnish us with important insights on how this may be accomplished.
With all else being equal, consumers will always prefer a more general (liquid) MoE. A regular merchant today would likely only prefer Bitcoins over dollars (or the local currency) if the former provided a significant decrease in transaction costs or if an understanding of the potential of the technology leads to a willingness to hold onto Bitcoins in speculation of future value.
To borrow from Peter Šurda’s analysis, there are three factors that influence the choice of a MoE: liquidity, SoV, and transactions costs in the narrower sense. All three factors can be considered transactions costs in the broader sense.
First, we will address transactions costs in the narrower sense, which comprises technical logistics, fees, storage and transportation costs, authentication, censorship resistance, counterparty risks, regulatory burdens. Today, in many instances, Bitcoin presents a significant improvement over the existing monetary base, which results in reduced transactions costs in the narrower sense for its users. Especially in cross-border payments, for example, Bitcoin is already a superior MoE when compared to paper money or gold, both of which involve the use of payment networks (off-chain), whereas Bitcoin transacts with base money (on-chain) for international transfers.
In many other environments, Bitcoin does not entail relative advantages (if at all) for everyday payments with base money (on-chain) or through payment networks (off-chain). There are countless jurisdictions where domestic transactions occur without friction and virtually instantaneous through cash, the banking system or with the use of other methods, such as credit and debit cards or mobile payments.
As cryptocurrencies and so-called blockchain technology bring more competition to the world of finance, by considering base money and payment networks together, it is just a matter of time for the existing financial infrastructure to catch up and improve considerably. While enhancements may not be enough to offset all reduced transactions costs in the narrower sense, they should be sufficient to reduce the gap.
In other aspects, however, Bitcoin will most likely maintain its dominant position. No other financial system can be as open and inclusive since anyone with Internet access can start transacting within seconds, and as resistant to censorship, which is valid for both base money and payment networks (e.g., on-chain and off-chain).
Reduced transaction costs in the narrower sense alone can influence the choice of currency. However, with intense competition, it has a limited and ephemeral effect. Even though Bitcoin’s transactions costs in the narrower sense are lower than base monies of paper or gold (on-chain), when individuals compute overall transaction costs, it makes little sense to accept the digital currency in everyday commerce.
Bringing us back to liquidity and the lack thereof, volatility, Bitcoin remains unstable and unpredictable in price, which amounts to a considerable barrier for widespread adoption as a regular MoE accepted and held by businesses and individuals. Presently, only people who anticipate Bitcoin can preserve or appreciate over time will accept payments and hold it as cash balances. These individuals who demand it as cash balances today do so because it has appealing SoV properties and not because it is a highly liquid good. Bitcoin is accepted and held not because it is a more “general MoE,” but because it may better preserve value and become a more general MoE in the future. Therefore, SoV takes precedence today over liquidity.
As Bitcoin’s value proposition is more appreciated, as its demand as cash balances marginally increases and as these behaviors are observed and imitated by more individuals[3], then its liquidity will increase while decreasing volatility. Network effects exert a positive feedback loop, enhancing Bitcoin’s liquidity, contributing to a more stable price, reducing overall transactions costs, and permitting it to be used more as a MoE.
So, the answer to the question of when Bitcoin will be used as a MoE is scantily now, but progressively later as it will occur gradually over a long period. We are simply not there yet. To accomplish this, Bitcoin needs time and, most importantly, must retain its utility as a good SoV.
Read the third part.
[1] I do realize these paragraphs may evolve into a discussion of its own (fractional reserve banking), but it is extremely insightful to understand how a Bitcoin standard can develop when we view accurately how central banks, base money, and banks relate to each other.
[2] By the way, for merchants that already understand the potential and want to be paid in Bitcoins today to contribute to its adoption, then a great solution is BTCPay Server, an open source payment processor that operates at no cost.
[3] Some will see this argument as proof for “Aha, Ponzi! You need more people to come onboard!” This accusation is absurd as Bitcoin works just fine and will remain functioning flawlessly with any number of users. The argument here is regarding liquidity and monetization of a good, which applies equally to gold and paper currencies. The only difference with paper currency lies in its source of demand which is highly influenced by legal tender laws.
