On the face of it, it sounds like a robust settlement process, because the blockchain-powered transfer of funds between the two companies is free, and confirmed within 10 minutes. But because we were mimicking the traditional remittance process so closely, we were also mimicking its costs.
The First and Last Miles Are Where All the Costs Are
There are obviously big expenses involved in operating the kiosk on the sending side: customer acquisition costs, physical costs, and big hairy issues like licensing and compliance.
Every country handles this differently, but with a little bit of math you could make some practical guesses regarding these respective costs. The average remittance shop at the average strip mall does no more than two hundred transactions a day, and we know from our own experience and World Bank estimates that the average transaction size is somewhere between $50 and $100. The fees that the kiosk must charge its customers therefore need to match up against whatever costs they’d normally expect to incur over their first year or so. In districts with Filipino migrant populations in Canada, Hong Kong, Singapore, these economics have resulted in a flat fee of US$3–5 and an FX spread between 1–1.5% per remittance transaction.
Once the money has made its way to the receiving country, things get even hairier. The Philippines is an archipelago with 90 million people spread out across 2,000 islands, 75% of which are unbanked. But traditional remittances are a huge industry here, and the main reason is because the pawnshops have collectively built a domestic backbone through which $60B flow through annually.
The pawnshops are the last mile, and no one, not even the banks, are close to touching their aggregate dominance. As such, they can charge whatever they want for their services, and the oligopolistic nature of the industry ensures that there is very little difference in price between one provider and another.
As a prime example, to send the equivalent of $10 (about a day’s minimum wage in the Philippines) from one city to another, all the pawnshop networks charge 6-7%. When a Bitcoin company, acting on the behest of its sending customer, wants to transfer $10 from one city to another, it needs to pay that same fee.
If the beneficiary is lucky enough to have a bank account, the costs become flat fees, but banks have traditionally never been good at moving small amounts. (At the $10 range, their flat fees are 10–20% of the principal amount.)
All told, the costs of a Bitcoin-powered remittance and that of a transaction powered by Moneygram or any other traditional provider are markedly similar, because the local currency needs to flow through the same physical channels.
So is there any benefit at all to using Bitcoin?
Bitcoin Makes Starting a Remittance Business Cheaper
As evidenced by the proliferation of “rebittance” startups, there is one thing that Bitcoin does enable: instant settlement across borders.
The implications of this are initially hard to discern. From the customer’s point of view, all digital money appears to be instantly settled. But from an infrastructure standpoint, whenever a customer uses Paypal to “instantly” pay a merchant across the world, a complex series of clearing-house and correspondent-bank settlements have to take place in order to compensate the local bank that eventually makes hard currency available at withdrawal.
So when Western Union promises to make your money available to your beneficiary within an hour’s time, how do they actually do it? Simple. By having a huge cash reserve already sitting in the receiving country, ready to be paid out to beneficiaries at any time.
This “prefunding” strategy is the main thing that young Bitcoin remittance startups are now able to circumvent. Instead of advancing $100,000 to each country that you want to do business in, you settle each transaction in real-time with Bitcoin. From a cost perspective, it’s the difference between building a huge server farm versus pay-as-you-go hosting on Amazon AWS, and is a true game-changer.
But if the cost of capital is lowered, shouldn’t the overall cost to the customer be lowered as well? Not really. What the blockchain has allowed us to do is circumvent the SWIFT network, but as soon as we hit the ground and go back to local currency, all the same strictures are then re-applied.
You Can’t Make Remittances Cheaper without Changing the Last Mile
Remittances can only become cheaper by reworking the last mile itself, and there are a handful of startups focusing on this problem. Remittance darling Abra has been attempting to establish partnerships with logistics providers and airtime top-up vendors, essentially anyone with a large footprint and decent amounts of liquidity. The new strategy is a marked difference from the original “human ATM” pitch that got so much flack from industry pundits earlier this year. Meanwhile, local startup Ayannah is aggregating the pawnshops themselves under a single web-service umbrella. In both cases, it’s likely that the existing oligopoly will keep fees at the same level it always was.
Outside of Bitcoin, what other solutions exist? Mobile money seems to be most practical candidate. It replaces the dependency on a physical last mile with that of an over-the-air, good-as-cash technology. Unfortunately, it’s tough to find a really successful implementation of mobile money outside of Kenya’s MPesa, and perhaps Zimbabwe’s EcoCash. Even in the Philippines — where mobile money was first invented in 2001 — the cold war between our two largest telco’s and their respective mobile money services make a unified solution unlikely.
The perfect solution appears to be a hybrid between the current Bitcoin remittance startups and the ubiquitous mobile-money network that currently exists only in our dreams. If we can use Bitcoin to replace SWIFT during the international part of the journey, and employ the local mobile money network as our domestic transport, then we’ve really got something. Until then, associating Bitcoin so directly with cheaper remittances is perhaps missing the point: the cryptocurrency has illuminated an even larger problem that it simply cannot solve on its own.