The cost of making a bitcoin transfer is often compared to the cost of an international wire transfer or remittance. It’s said that it’s cheaper to move value using cryptocurrencies than traditional payment networks. But there’s a right way to do this comparison and a wrong way.
I was reminded of this when I saw a tweet from prominent crypto commentator Joseph Young slide on to my Twitter timeline a few days ago:
PayPal says its bank account withdrawal fee is $1, which is true. BUT, the conversion rate is where they get you.
I received $1,000 this week from a friend. PayPal shows $955 as withdrawable amount.
$45 fee for a $1,000 payment, while Bitcoin processes $194 million with $0.1 pic.twitter.com/FdXmi6eb8A
— Joseph Young (@iamjosephyoung) December 1, 2018
Now I don’t want to single out Joseph here. He seems like a nice guy. He’s written some good articles on the Korean bitcoin markets. But his tweet is the perfect example of the wrong way to benchmark bitcoin against wires. That he received over one thousand retweets—way more than any tweet I’ve ever sent out—means a lot of people have the wrong impression about the relative merits of wire transfers and bitcoin.
Joseph made the same claim a month ago when Nouriel Roubini began his Twitter onslaught against bitcoin. In both tweets, Joseph is failing to set up an even playing field. The bitcoin network does one thing for customers. It moves value. Transferwise and PayPal do two things for their customers. First, they move value. Next, they convert this value into the local medium of exchange. Bitcoin payments are only cheaper than PayPal or Transferwise payments because the bitcoin network doesn’t provide this second service. If Joseph were to add the cost of converting bitcoin to fiat, he’d see that the comparison isn’t quite so favorable.
That this foreign exchange conversion must be absorbed is a fact of life. The recipient of Joseph’s $1 million transfer can’t meet his or her living expenses and debts with bitcoin, they need local currency.
Here’s the right way to benchmark a fiat-to-fiat cross border transfer against a bitcoin transfer. According to Joseph, it costs 10 cents to move $1 million, or 300 bitcoins, along the bitcoin network. Let’s compare this to how much it costs for the legacy cross border system to move $1 million in U.S. dollars without doing a foreign exchange transaction.
If bitcoin were treated like local currency, then expensive conversion costs would no longer hobble its usage as a cross border payments option.
The personal situation for each individual will vary, but here is how it breaks down for me in Canada. I have a U.S. dollar account at my Canadian bank. I often receive U.S. dollar wires from American businesses. On the U.S. side, the sender of the $1 million will be charged $35 for wiring me $1. That’s based on fees quoted by Citibank. Meanwhile, my bank charges me a flat $15 to receive a U.S. dollar wires.
So the all-in cost of a $1 million U.S. dollar-to-U.S. dollar wire works out to $35+$15=$50.00. That’s certainly more than the 10 cent mining fee for using the bitcoin network, but far less than the $7,500 that Joseph cites in his first tweet.
No matter which way you send me the $1 million, bitcoin or fiat, I still need to convert it into local currency. This is where the adventure begins. There are two costs involved in converting $1 million worth of bitcoin into Canadian dollars. First, I’ll have to pay a 0.5 percent trading fee to QuadrigaCX, one of the two largest Canadian bitcoin exchanges. This works out to $5,000.
The next fee is sometimes referred to “paying the spread.” The spread is the gap between the price at which buyers are bidding for bitcoin and sellers are willing to offer. Ideally, I’d like to sell my entire $1 million stash of bitcoins at the half-way point between QuadigaCX’s best bid and ask price. In reality, I’ll have to move down and sell at the price of the highest bidder if I want to get my hands on dollars.
Unfortunately, the Canadian dollar-to-bitcoin market is relatively illiquid. Volume on QuadrigaCX rarely exceeds 1000 bitcoin or so per day. The first bidder will only be able to absorb a few thousand dollars’ worth of my 300 bitcoin sell order, so I’ll have to move down to the next bidder, and the next. By the time I’ve liquidated all 300 bitcoins, I’ll have driven the price of bitcoin on QuadrigaCX quite a bit below its market price. If I end up selling my coins $10 below fair value—a conservative estimate—then I’ll have paid an extra $3,000 in fees.
To sum up, I will have to pay a total of $3000+$5000=$8000 to convert $1 million worth of bitcoins into Canadian dollars. But the costs don’t end there.
Americans receive wages and business income in dollars, not bitcoin. So even before the U.S.-based payor can send me $1 million in bitcoins, she has to buy them on a bitcoin exchange. Coinbase Pro, one of the largest U.S. exchanges, charges 0.3 percent per transaction, which means she’ll be charged $3,000 in fees in order to buy $1 million worth of bitcoins.
By now, readers will be able to see that the 10 cent mining fee that Joseph cites in his tweets is irrelevant. Most of the resources used up in making a $1 million cross border payment are due to the cost of converting local currency into bitcoin and bitcoin back into local currency. In the case of the US-Canada route, this comes out to at least $7,500 + $3000 + $0.10 = $10,500.10. That’s a lot!
Let’s see how this compares to using a non-bitcoin route for moving US$1 million to Canada. Our first option is Transferwise, a remittance provider that offers to route US$1 million to Canada for $4934. That’s quite a bit cheaper. Beware, Transferwise’s quote may not accord with the rate that one actually receives.
While Transferwise does a good job with a few thousand dollars, a better option for converting large amounts of U.S. dollars into Canadian dollars is to open an account at an electronic brokerage like Interactive Brokers, or IB. An electronic brokerage like IB provides regular retail customers with direct access to the interbank foreign exchange market. The interbank market is where large financial institutions trade currencies between each other. According to the Bank for International Settlements (BIS), $218 billion is traded each day in the U.S. dollar-to-Canadian dollar market alone.
Once my Canadian bank has confirmed that I’ve received US$1 million from Citibank (remember, this cost $50), I’ll send the funds to my IB account. IB will charge me a fixed fee of $20 to convert US$1 million into Canadian dollars. I will also have to pay the spread, but with so many large players active in interbank markets, the gap will be incredibly tight. The all-in-cost of the full US dollar-to-Canadian dollar payment will come to less than $100. That’s minuscule compared to the bitcoin route, which added up to $10,500.10. So there is no reason for me to choose bitcoin as my network for making cross border payments.
Joseph's homage to 10 cent fees on bitcoin cross border payments ignores all the foreign exchange hassles that go with it.
I’ve provided the Canadian experience, but other destinations will be different. If I were living in Japan, for instance, the calculation would tip more in bitcoin’s favor. BitFlyer, Japan’s largest bitcoin exchange, charges just 0.02 percent on extremely large bitcoin transactions. A sale of $1 million worth of bitcoins for yen means paying a small $200 fee, far less than the $5000 that QuadrigaCX charges. The full cost of the US-Japan bitcoin route comes out to just $3200 ($3000 to purchase the coins on Coinbase plus $200 in Bitflyer fees and a 10 cent bitcoin fee).
That’s certainly better than Transferwise, which quotes $8397 to send $1 million to Japan. But I doubt the bitcoin route could beat a US dollar-to-US dollar bank transfer combined with a trip to a broker like IB that provides individuals with direct access to the interbank foreign exchange market.
Exotic corridors may prove more fertile for bitcoin cross border payments. Because very little trade or investment occurs over most of the world’s 16,000 or so currency corridors, there isn’t enough liquidity to provide direct conversions of one fiat currency into another. To transfer money from Chile to Zambia, for instance, Chilean pesos will have to be sold for U.S. dollars, and then the dollars sold for Zambian kwacha.
Since fiat transfers along exotic corridors incur twice the fees, an equivalent bitcoin payment compares more favorably. However, nations with illiquid currencies are likely to be small and undeveloped, and so may also lack liquid bitcoin markets. Zambia, for instance, doesn’t have a bitcoin exchange, so selling $1 million worth of bitcoins for kwacha may prove to be costly.
Joseph’s homage to 10 cent fees on bitcoin cross border payments ignores all the foreign exchange hassles that go with it. These hassles may one day disappear, but only if bitcoin is able to attain local currency status. That is, people must become comfortable receiving a bitcoin payment and holding on to the proceeds for a few days because they know that their grocer or landlord also accepts bitcoin. If bitcoin were treated like local currency, then expensive conversion costs would no longer hobble its usage as a cross border payments option. Since bitcoin would no longer be starting the race ten steps behind fiat, its 10 cent fees would be able to shine.
But this is a big challenge. Can a volatile and awkward instrument like bitcoin ever become docile enough to be used as a local currency? I’m skeptical, but we’ll see.