Remember the day when a coin was just a bothersome piece of metal that you were happy to toss in the tip jar to benefit the underpaid baristas at your local coffee shop? Well no more. Now a coin might be worth more than many people make in a week. Over the next few weeks Economist David Evans will be taking a deeper look at bitcoin–why it is so special, whether it can provide stable value for investors and if it has a future far beyond its embryonic phases as a digital currency.
Missed the first three editions? Don’t worry, you can read the first installment of the Digital Currency Deep Dive: What Makes Bitcoin So Special here, the second installment It’s The Protocol Stupid here, and the third installment Digital Currency Deep Dive: Hey Who’s In Charge Around Here? by clicking here. You can also get a look at Dr. Evans’ technical paper on his SSRN page here.
Bitcoin is going to clean the clocks of the traditional payment providers because it eliminates the middleman and it’s really cheap. That’s a frequent claim among advocates for bitcoin and the other public ledger digital currencies. As Bitcoin Daily, echoing many others, put it:
“Bitcoin’s efficiency and low cost compared to the traditional payment methods that we are using today will ultimately prove too tempting to merchants, businessmen, and individuals. Bitcoin processors only charge 1 percent for bitcoin transactions, compared to the 2-3percent ‘standard’ fees paid by merchants for credit card processing. Bitcoin allows for the simple and secure transfer of value online, without the need for intermediaries. These include credit card network, card-processing companies, or money transfer agencies that rely on skimming transaction fees for their revenue.”
Is this true, and how would we know? That’s the subject of this edition of the Digital Currency Deep Dive.
To get an understanding of how Bitcoin could disrupt traditional payments let’s start by describing how traditional payments currently work.
The basic building block of payments involves sending and receiving money. That’s easy with cash: I give you cash, you get cash. It is much more complicated with electronic payments.
Electronic payment systems typically have a computer network for sending and receiving funds. Some are closed private networks. A remittance provider such as Western Union, for example, usually sends money to, and receives money, from its agents using its private money transfer network. Others, such as those networks used by banks domestically and internationally, are multilateral systems with clearing and settling. They may send messages concerning debits and credits and periodically settle accounts based on net positions. Four-party payment networks such as MasterCard work this way too. The network operates a clearing and settlement system for card issuers and merchant acquirers. In all these cases, the network ultimately enables people, businesses, and other entities to send money to, and receive money from, each other.
Many payments businesses wrap a variety of other services around the elementary function of sending and receiving money. In some cases the entity that operates the core network provides other services; in other cases third-party businesses provide other services relying on the system for moving funds.
Consider some examples.
Remittance companies usually operate networks of agents around the world, often in poorer neighborhoods and remote locations, that enable people to send and receive funds in cash, which is what many people need in many lesser-developed parts of the world. Credit card companies like American Express provide consumers (senders) with a bundle of services including credit lines, fraud protection, and a dispute process for charges from merchants and provide merchants (receivers) with useful data for marketing and a pre-qualified group of consumers. Banks include the ability to receive money into an account and the ability to send money out of an account through ACH-type systems as part of the basic checking account that people and businesses have.
Many other payments businesses provide services that focus only on either the sending or receiving part of the transaction. In the payment card industry, card issuers specialize in servicing senders of funds (typically consumers) and merchant acquirers specialize in servicing receivers of funds (typically merchants). In the banking industry some banks have businesses that focus on originating the transfer of funds such as bill pay services. And still other businesses provide services to these other specialized businesses.
I said above that the basic cash transaction is very straightforward. But even the cash business has a relatively complicated ecosystem built around sending and receiving funds. The ATM system, including ATM/debit cards, enables consumers to get cash from their banks accounts. Cash-in-transit businesses such as Loomis move cash between banks and between merchants and banks. Then there are businesses that focus on providing the paper stock for the currency, printing it, providing machines for counting, and cash registers for accepting it.
The bottom line is that most senders and receivers of money are not just obtaining a simple payment service. They are usually getting an extensive bundle of services that includes, and is ultimately predicated, on sending and receiving funds. Moreover, payments-related businesses may specialize in different elements of this bundle. As a result one must be careful in making sure that comparisons are apples-to-apples.
Government regulation is the other factor to consider. It bites different players differently. Governments around the world have imposed many regulations involving sending and receiving money to prevent money laundering and funding terrorism. Those regulations are particularly onerous for international money transfers. Governments also have many regulations for the various services such as lending that are bundled with sending and receiving money.
But one thing is true for sure: as Bitcoin Daily says, there are many intermediaries in the payments business and they all charge something to make money.
Domestic or international computer networks are typically the workhorses of sending and receiving money. The computers are usually linked together through private communication networks. These networks face the usual path dependence problem of computer systems. They have been built over time and therefore include elements that are far from the frontiers of information technology. They are kludgy.
Nevertheless, between scale economies and efforts to operate them efficiently these specialized computer networks are able to send and receive money at far less than 2 percent of the transaction amount. There isn’t publicly available data on this but based on my conversations with people in the business the average cost of moving money over these networks is trivially small on a per transaction basis. That’s obvious from considering how these systems are run. They are sophisticated computer systems involving little manual intervention that are moving massive sums domestically and globally.
But as evidence of the low cost, here are some examples. The cost to financial institutions of transactions by one of the ACH operators in the U.S. is less than 2 cents. In fact, the low cost of these networks is one of the reasons that banks in developed countries typically provide direct debit and credit transfers at a very low cost, if not for free. Merchants that belong to MCX, the new merchant-run ACH-based debit card system in the U.S., are paying only 4 cents per transaction to FIS, which is operating the system, as a result of being able to rely on low-cost ACH.
Visa and MasterCard, based on data they reported to the Federal Reserve Board, charged acquirers and issuers an average of .3 percent of purchase volume (10.4 cents per transaction) in 2009. Remember, in the case of Visa and MasterCard, that revenue covers the cost of the entire brand positioning and marketing these networks do on behalf of their customers – the issuers – as well as operating the clearing and settlement network. The PIN debit networks in the US were even cheaper. They charged a total of .1 percent or purchase volume (4.1 cents per transaction).
The traditional payments industry therefore has inexpensive methods for moving money—at least compared to the 2-3 percent figure that is sometimes thrown around. The frictions that people complain about generally come from the various intermediaries that plug into the clearing and settlement systems and, significantly, from the regulations these intermediaries have to comply with. They also come from the cost of dealing with fraud and disagreements among merchants and consumers.
At the heart of the Bitcoin ecosystem is the decentralized digital public ledger (aka the blockchain). That is the network that provides authentication, clearing, and settlement. It consists of decentralized servers operated by individuals around the world for hosting the blockchain and a labor force of “miners” who provide the processing services that results in authentication (is it a valid “non-double” transaction) and clearing and settlement (bitcoins are effectively moved from the sender to the receiver). The miners have formed “pools” which in some cases control a significant amount of the processing capacity. In fact Ghash has just acquired 51 percent of network power.
As with traditional payments, many other businesses hang off of this network and provide value-added services. During its early stages of development, people sent and received bitcoins directly over this network. But as Bitcoin has tried to move into the mainstream, wallet providers have arisen to make it easier for nontechnical consumers to use bitcoin.
Wallet providers enable consumers to put bitcoins in a wallet—much like putting dollars on a prepaid card. They then enable consumers to pay with those bitcoins at participating merchants who accept the wallet for payment. Like traditional payment intermediaries, wallet providers bundle many services in addition to facilitating the sending and receiving of funds. Coinbase, for example, insures merchants against exchange risk by paying US merchants in dollars rather than bitcoins and they provide consumers a presumably safe place to store their bitcoins and replenish their stocks of bitcoins.
Many other businesses have formed as complements to the public ledger. These include most exchanges, vaults, manufacturers of mining equipment, and many other specialized value-added service providers.
Like traditional payments, bitcoin is evolving into an ecosystem with many specialized players.
Unfortunately, The Bitcoin Daily comparison—and one that is commonly heard in discussions of the advantages of bitcoin over traditional payment methods—is an apples-to-kumquats comparison. To really compare transactions across payment methods you have to make sure that the comparison includes the costs incurred by both the sender and the receiver and that they are for the same things. Bitcoin Daily compares the cost to the receiver for bitcoin with the cost to the sender and the receiver combined for credit cards and then further ignores the differences in the services provided.
Consider Coinbase. This bitcoin wallet charges the merchant 1 percent of the transaction amount. But that doesn’t include the round trip for the transaction. Coinbase also charges 1 percent of the dollars loaded into the wallet plus 15 cents. The round trip cost of a transaction that begins with the consumer buying bitcoins for the wallet is therefore 2 percent, ignoring the 15-cent fee.
In the Bitcoin Daily example, the merchant card processor charges 2-3 percent. But that fee, which presumably refers to a credit card given its size, is dominated by revenue that the merchant processor passes on to the card issuer. The card issuer, in turn, offers various services to the consumer including dealing with merchant disputes, insurance for fraud, float between the time of the charge and the time of the bill, and possibly rewards for using the card.
If we consider the round trip cost of the transaction (the total charge to the merchant and the consumer) Coinbase is currently charging fees that are roughly in the same range as credit and debit cards. A large merchant would typically pay less than 2 percent as the blended rate for credit and debit cards and the consumer would typically get some value back in rewards for credit cards. The round trip transaction fee is therefore less than 2 percent for credit and debit cards. In addition the merchant and consumer are receiving other valuable services that Coinbase isn’t offering.
Of course, Coinbase is still very small and it is possible that its fees will decline as it realizes scale economies.
Nevertheless, it would seem that Coinbase would have a long way to go to compete with ACH-based transactions. In the U.S., it is possible by linking one’s bank account to do person-to-person transactions for free using PayPal or SquareCash, and in the U.K. with Paym. Even for international remittances, PayPal linked to a bank account results in less than a 1 percent total fee (sender plus receiver) between many developed countries. Of course, perhaps these fees will go up as these services gain more traction, but the fact that domestic transactions are free now confirms the lost cost of sending and receiving funds using ACH-type systems.
The decentralized digital public ledger performs similar clearing and settlement functions as traditional payment networks. Wallets and other value added service providers then rely on the public ledger to support the services they provide senders and receivers of bitcoin.
While the public ledger may support product innovations that traditional payments can’t it is not at all obvious that the blockchain is a more efficient technology for processing transactions at least as it is currently organized. The public ledger is operated by a global network of miners, who are paid transaction fees for doing the processing work as well as randomly generated rewards that increases the supply of bitcoins. There is nothing inherent in that system that makes it obviously more efficient than a centrally controlled computer network.
In fact, as of now the bitcoin public ledger is a far more costly system for processing transactions than other networks for sending and receiving funds. Miners prioritize transactions based on the fees offered for a transaction. That results in senders having to pay fees in order to make sure their transactions get processed quickly. Coinbase reportedly pays .002 btc per transaction (roughly 11 cents as of today). One report from early February 2014 indicated that processing fees were less than 40 cents, which implies they were sometimes almost as high as 40 cents.
People often claim that with Bitcoin “you can send money between any two points on earth for free”. While that is true in some cases, sometimes a transaction fee is required. The fee, when it is required, is usually worth less than 40 U.S. cents.
If we take the Coinbase 11 cent per transaction—much less than the almost 40 cent max reported above–then the Bitcoin public ledger currently charges around the same price for clearing and settlement as Visa does (10.4 cents as of 2009). But as I noted Visa’s fee covers the cost of running a global branded payments network that provides significant brand and marketing support for issuers and acquirers and makes a significant operating margin on the transaction fee. If Visa broke out the fee for simply moving money, independent of other things it does, it would almost certainly be less than 10.4 cents even with a nice profit margin.
And that’s the most favorable comparison for the public ledger transaction fee. The 11 cent per transaction fee for the miners is also almost three times higher than what PIN debit networks charge in total. More importantly, the 11 cent bitcoin per transaction fee is an order of magnitude greater than the prices for clearing and settlement systems for banks. For example, we know that the clearing and settlement system operated by the Federal Reserve Board costs less than 1 cent per transaction (based on data here).
Over time it is possible that the Bitcoin blockchain will become more efficient as the volume of transactions increases. It is also possible that the distributed labor force of miners will be replaced by huge processing companies that can get operate less labor-intensive and more efficient operations. Ghash apparently has gotten to 51 percent of network power through operating efficiencies. Then one could imagine that the costs could decline to at or below the levels that existing clearing and settlement entities incur. But since the cost of processing transactions is already extremely low it also isn’t clear how much that would matter in the end.
It is very difficult to make a case for bitcoin as a more efficient method for sending and receiving money once one makes apples-to-apples comparisons that account for differences in services and compliance with government regulations. It is also difficult to make a case for bitcoin as having an inherently superior technology for sending and receiving money.
That doesn’t mean that bitcoin isn’t a disruptive technology. It could very well be. But the proof will be whether this crypto-currency can enable new products or services that existing payment systems can’t. Some have argued, for example that bitcoin could enable the conditional transactions such as shutting down my leased Tesla if I haven’t paid my bill or support sophisticated derivatives. There could also well be ways in which some variant of bitcoin could eliminate inefficiencies in cross-border transactions. Creative minds will surely do more with the blockchain innovation.
There are also use cases for which bitcoin could well be more efficient than existing payment methods. Bitcoin, for example was originally conceived as a method for supporting small value transactions. That, in fact, has been one of the weakest parts of existing payment systems because it is hard to develop a pricing and business model that works for very small transactions. As is well known, bitcoin was a particularly useful currency for entrepreneurs operating dark websites engaging in illegal activities. Similarly bitcoin could be a very efficient payment system for gamers and others who live in a relatively closed ecosystem for which a standard electronic currency makes it easier to transact.
An oft-made claim about Bitcoin, reflected in the Bitcoin Daily quote above, is that it dispenses with the need for intermediaries (aka middlemen). That would indeed be miraculous since every payment method since barter has had intermediaries of one form of another.
I’m afraid there are no miracles here. The public ledger, and the miners who operate it, stand between the senders and receivers of bitcoin. And that intermediary charges senders and receivers for transactions as we’ve seen. Over time the decentralized miners, who have already formed pools that control a significant fraction of transactions, could be organized into even more formal intermediaries similar to Visa and MasterCard if Bitcoin becomes successful.
Many other intermediaries are emerging. Bitcoin wallets are intermediaries between merchants and consumers. Exchanges are intermediaries between buyers and sellers of bitcoin. In fact, if Bitcoin were to evolve into a fully functioning financial services system we would expect the same kinds of intermediaries, and perhaps others, to emerge in the Bitcoin ecosystem as in the traditional payments ecosystem. Clearly, though, the Bitcoin ecosystem is awash with middlemen.
Some supporters of Bitcoin seem to believe that this crypto-currency will dispense with the need for banks—and those greedy bankers—and other financial intermediaries who “skim” transaction fees, as Bitcoin Daily put it. Maybe if Bitcoin is successful it will, indeed, kill off banking as we know it today. But new middlemen will populate the new ecosystem, and those intermediaries will perform many functions similar to traditional financial services.
Moreover, those intermediaries will be seeking fees. And why shouldn’t they? They are in the business of making money and their VC investors are looking for a return on their investment. We already see this with the fees being charged by all the new players in the Bitcoin ecosystem including the miners (up to 40 cents), wallet providers (2 percent roundtrip for Coinbase), the exchanges and so forth.
In fact, we have seen the emergence of dominant intermediaries—Mt. Gox before its collapse—and Ghash now, all in the very early days of Bitcoin. This is an ecosystem that is ripe for large scale-based intermediaries.
Thus, some of the claims for the Bitcoin revolution should remind us of the sage lyrics by The Who:
I’ll tip my hat to the new constitution
Take a bow for the new revolution
Smile and grin at the change all around
Pick up my guitar and play
Just like yesterday
Then I’ll get on my knees and pray
We don’t get fooled again
Don’t get fooled again
No, no!
I’ll tip my hat to the new constitution
Take a bow for the new revolution
Smile and grin at the change all around
Pick up my guitar and play
Just like yesterday
Then I’ll get on my knees and pray
We don’t get fooled again
Don’t get fooled again
No, no!
Yeaaaaaaaaaaaaaaaaaaaaaaaaah!
Meet the new boss
Same as the old boss