Full disclosure: The author works at Visa, but the views herein are his own and do not represent his employer. The author also owns some Bitcoin. Follow him on Twitter at @ggollerk.
Bitcoin is in the news again this week, after the price dropped over 10% at one point. Predictably, this led to a spike in media coverage, with pundits offering their explanations for the recent price volatility. Culprits identified included NYDFS’ proposed “Bitlicense” regulations, CFPB’s warning on virtual currencies and margin trading at some Bitcoin exchanges. This focus on price — i.e., Bitcoin as currency — is all too typical of mainstream coverage, but it often comes at the expense of understanding Bitcoin as a protocol and as a network. This post will focus on these latter two aspects of Bitcoin, and we’ll also take the pulse of Bitcoin’s performance in a few typical use cases. Before we proceed, it behooves everyone to know that the right way to think about Bitcoin is the following: protocol first, network second and currency third.
Bitcoin as Protocol
The Bitcoin project is a breathtaking accomplishment: an open-source, distributed, peer-to-peer, cryptographically secured protocol that effectively solves the Byzantine Generals' problem and delivers a mechanism to transmit value anywhere in the world, without a central authority. This is the stuff of dreams (payments geek and cypherpunk dreams, anyway) and probably deserves a Nobel Prize. Bitcoin still faces an uncertain future, but regardless of the outcome, some of the key ideas represented by Bitcoin are probably here to stay: cryptographically secured transactions, the power of decentralized networks, “smart” transaction fees (i.e., market-determined), a stable, non-arbitrary growth in the money supply. Bitcoin has forced people to ask fundamental questions about the nature and role of money in an economy, the purpose of central banks and state-issued fiat currencies and what we really mean by “good money.” This is an encouraging development in these unprecedented monetary times, and for that, we have Bitcoin to thank.
Bitcoin as Network
What about Bitcoin’s key features as a payments network? Bitcoin offers push-based, “sender-pays” transactions with cash-like properties that ensure the funds are good as soon as the network confirms the transaction as valid (~10 minutes). This stands in direct contrast in almost every way to most retail payments today (in particular, card-based payments) which are pull-based. (At least in the developed world; the developing world is overwhelmingly cash-based, which is a form of push payment.) These “receiver-pays” transactions have an authorization/clearing/settlement framework and can take 2+ business days for funds to be delivered to the recipient. It’s useful to unpack the jargon here to really understand the differences.
First, push-based payments mean that the sender does not reveal any payments credentials in the process of transacting. The receiver publishes a public address (e.g., via a QR code) to which the sender then “pushes” funds. (Email is a good analogy, with email IDs being the equivalent of public addresses and the sender not having to reveal his email password to complete an exchange.) A traditional card-based payment today is pull-based, where your card credentials are read (or in some cases, stored on file) during every transaction, which is clearly not preferable — at least in terms of network design.
What about practical use, though? A formidable security infrastructure has grown up around card-based payments over the decades (EMV and tokenization are recent examples). Reading the data-breach headlines in recent months (e.g., Target, PF Chang’s, etc.) you would think that card-based fraud is rampant and exploding. In fact, according to the Nilson Report, global card fraud was about 5bps (i.e., 5c per $100) in 2012, and has remained remarkably stable for ~20 years. Fraudsters and payments security professionals battle every second of every day of every year, but thus far, nobody is decisively gaining ground. (Recall the useful analogy of fraud being like a balloon — you press down in one place only to see it pop up in another.)
Second, Bitcoin is a sender-pays model. This model inverts the economics of traditional retail payments, where the receiver (typically, a merchant) pays a discount rate (also called an interchange fee) to facilitate the transaction. Interchange fees vary across the world, but according to the Federal Reserve, they average ~150-175bps globally. Recall that card fraud was about 5bps, which means that interchange pays for much more than just card fraud (more on this later). Bitcoin also changes the economic model in another way — Bitcoin transactions feature a fixed-fee per transaction, rather than the ad valorem fee associated with traditional card payments (leaving aside exceptions created by regulation like the Durbin amendment).
Third, Bitcoin is like digital cash. The act of completing a transaction embeds the traditional steps of authorization, settlement and clearing (think of it as the digital equivalent of handing over dollar notes). With Bitcoin, transactions are confirmed by the network of Bitcoin nodes; it is the network that confirms that ownership of the funds has been transferred from sender to recipient. Traditional payments network operators like ACH, Visa/MasterCard, etc. cannot “settle in Bitcoin” like they can in other fiat currencies, unless they join the network of nodes running the Bitcoin protocol.
Where does this leave us? It’s probably fair to say that if we were designing a global payments network from the ground up today, it would look a lot like Bitcoin. Bitcoin has some interesting network/protocol design advantages over prior payments systems. But these design advantages tend to be marginal to anyone who is not a payments geek or Bitcoin believer. The ordinary consumer just doesn’t care enough about how the plumbing of payments works behind the scenes. Now is a good time to recall a simple rule of thumb that tends to hold in most two-sided markets: to dislodge an incumbent, a new service needs to be 10x better than the existing options (indeed, Uber comes to mind as the archetypal example of that 10x improvement). The conclusion is inevitable — Bitcoin’s mere creation and continued existence are not enough to achieve mainstream usage. The Bitcoin ecosystem of entrepreneurs, Bitcoin enthusiasts, early adopters, merchants and technology providers will need to build products that demonstrate Bitcoin’s utility in real-world use cases.
Crypto-currency confronts the Real World
This section will evaluate Bitcoin’s progress in a few typical use cases: remittances, eCommerce and retail payments.
Remittances: On the surface, remittances would seem to be a use case tailor-made to demonstrate Bitcoin’s potential. Remittances today are often expensive and slow, just like they were decades ago. Why can’t we just use Bitcoin to “solve remittances” and allow anyone to transfer money anywhere in the world? Slam dunk, right? It turns out that once you wade into the remittances world, regulation (KYC, AML, etc.), fraud, the need for physical end-points that provide cash-in/cash-out services, consumer education and the challenges of managing inflow/outflow imbalances across various remittance corridors will quickly halt you in your tracks. You can’t “move fast and break things.” This industry has already seen newer entrants like TransferWise and Xoom innovating and building next-day (even same-day in some cases) remittance services, all within the existing payments infrastructure. This is hard, unglamorous work, but these innovators are materially improving the lives of migrant workers and immigrants, i.e., those who need it most. There are many Bitcoin-based companies that are hard at work as well, and one of them may very well be building the “10x improved” remittance solution right now — we just haven’t seen it yet.
eCommerce: This is a space where Bitcoin has made the most visible progress, particularly in the U.S., with Bitcoin payments processors like Coinbase and BitPay signing up big merchants like Overstock, Dell, Rakuten’s US subsidiary, etc. What do these merchants have in common? They are hypersensitive to “cart abandonment.” The holy grail of eCommerce appears to be frictionless commerce, i.e. building an Amazon-style 1-click checkout [An aside: as a consumer, I personally find the friction in eCommerce quite helpful in thinking through whether to really buy something. I leave items in my Amazon cart for a while before pulling the trigger, and highly value the option to reconsider whether to buy after a night’s sleep.] Are Bitcoin transactions frictionless? Not really, not true blockchain transactions anyway, which involve scanning a QR code or copy-pasting a Bitcoin address or some other way to get hold of the recipient’s address.
Some Bitcoin service providers offer smoother checkout processes by using off-blockchain transactions, but at that point, you’re using “Coinbase-coin” (or “Circle-coin”), not actual Bitcoin. A closed-loop service like that isn’t much different from PayPal or Google Wallet, and it changes the nature of Bitcoin by reintroducing an intermediary and requiring the re-establishment of a trusted 3rd party relationship. Clearly, merchants find it useful to offer any payments mechanism that has a chance of ensuring a purchase conversion (see the screencap below from Rakuten’s website, featuring a multitude of digital wallet checkout options). Merchants receive a PR boost from aligning their brand with Bitcoin and they may also win the loyalty of the Bitcoin community. What about the mainstream consumer though? Sure, it may be about as frictionless as other digital wallet services (although a smoother Bitcoin checkout requires a trusted 3rd party as noted above). Does this meet the bar of a 10x improvement over existing options? No.
(A separate, but related, point: the “power law” distribution of online retail means that most of the purchase volume flows through a handful of websites like Amazon and Apple, who already have consumers’ cards-on-file and thus already provide friction-free checkout. The top 5 eCommerce merchants are the ones who can actually move the needle and influence consumer behavior.)
Retail payments: More broadly, what are the other reasons why Bitcoin makes sense for retail payments, for both merchants and consumers? Lower fees associated with Bitcoin-based transactions are the most cited reason and there is clear value to the merchant from accepting lower-cost Bitcoin payments (although at the moment, the cost of converting from Bitcoin to fiat currency eats into the benefit of accepting Bitcoin). Some Bitcoin enthusiasts seem to believe that lower merchant cost is an end in and of itself. It is wise to remain less sanguine about the desire and intent of merchants, both big and small, to pass along these savings to consumers.
Recent history tells us that when merchant acceptance costs are lowered, merchants do not pass along the savings; instead lower costs enrich their bottom-line. Australia is a good example, where interchange was regulated over a decade ago but consumers have seen very little benefit from lower acceptance costs. Global fraud rates, as we noted earlier, are stable and low. Much of the interchange fee charged to merchants is actually delivered to consumers in the form of rewards. As fashionable as it is these days to rail against the big banks, the reality is that a credit card offers consumer protection in the form of zero liability (either the bank or the merchant bears the risk of fraud, depending on the type of purchase), the benefit of float / cash flow management in the case of credit products, and a robust rewards program. Essentially, your bank gets you a discount on every single purchase, paying you back in the form of cash, points, miles or another currency of your preference. As a baseline, let’s say that the typical card product offers you 1% cashback (although it’s possible to do much better, just ask Samuel L. Jackson). Let’s relax our 10x improvement rule and use 5x instead—this also happens to be the value proposition of Target’s highly successful REDcard product. That means Bitcoin merchants need to offer a 5% discount for Bitcoin-based payments to have a chance at mainstream acceptance. While a few merchants have offered discounts for Bitcoin payments (it’s the only time I’ve been convinced to buy anything with Bitcoin), much more work is needed before mainstream consumers have a reason to pay attention.
(Note: Bitcoin early adopters receive an additional benefit from spending their Bitcoin since the price has increased multi-fold since their purchase; this is a benefit unavailable to mainstream consumers, and even this incentive is diluted by the recent IRS guidance which treats Bitcoin as an asset, requiring tax reporting for every single Bitcoin purchase.)
Grade: Merchant value proposition A-, Consumer value proposition C+
The Road to Mainstream Adoption
It is still early days in the Bitcoin experiment and many people are hard at work building great products and services that are aimed at bringing Bitcoin to much wider usage. The purpose of this post was to take a clear-eyed, dispassionate view of Bitcoin’s progress-to-date. Much has been accomplished in a short period of time, but in absolute terms, Bitcoin has only reached the foothills—there are mountains to climb ahead before Bitcoin finds its way to the mainstream.
If this post has sounded pessimistic, it is only because I have tried to view Bitcoin from a mainstream user’s perspective rather than through rose-tinted glasses. I believe Bitcoin has the potential to change the world—beyond payments, exciting new possibilities are on the horizon—distributed p2p asset markets can help address the absolute power of Wall Street, trustless Bitcoin-based contracts can introduce huge efficiencies in contract negotiation and enforcement, machine-to-machine payments running on Bitcoin could completely reshape commerce, and Bitcoin could emerge as an alternative asset class (i.e. the new gold) or even as an international reserve currency (i.e., bancor for the 21st century). These transformative “Bitcoin 2.0” applications deserve a closer look, but that is a topic for a separate post.