The New York branch of the Federal Reserve published a paper today called Is Bitcoin Really Frictionless. It digs into the prices of bitcoin on three major USD spot exchanges: Bitfinex, BTC-e, and Bitstamp. It's an overall interesting paper with some good analysis and graphs and appears to be a genuine attempt to investigate bitcoin price issues.
The main point of the paper is to analyze how and why bitcoin price on these three exchanges differs so much, and to state that bitcoin is not really frictionless in this sense. Some of the culprits they point to are volatility, using this graph as a starting point they see the relationship it has between price differentials:
The main point of the paper is to analyze how and why bitcoin price on these three exchanges differs so much, and to state that bitcoin is not really frictionless in this sense. Some of the culprits they point to are volatility, using this graph as a starting point they see the relationship it has between price differentials:
The main reason why price differences would exist between two exchanges on a fungible commodity like bitcoin is that there are barriers to efficiently arbitraging, or profiting from these price differences. Normally if Exchange A has a price at $400 and Exchange B has a price of $410, you can simply buy at A and sell at B and earn $10. However in practice there are issues in waiting for wire transfers, bitcoin confirmations and risking volatility in the meantime, as well as fees:
The authors did a good job of analyzing the raw empirical data, compiling interesting relationships since 2013 on the three exchanges in question:
They even addressed the issue of exchange risk to explain why BTC-e trades persistently at a discount:
Exchange failure or fraud is another source of risk. Exchange failure is not merely a theoretical possibility in bitcoin markets—it occurs regularly. A study in 2013 reported that eighteen of the forty bitcoin exchanges analyzed—almost half—ultimately failed. Most notable among all bitcoin exchange failures is that ofMt. Gox, an exchange that once commanded the largest share of the market and lost roughly $460 million worth of its users’ bitcoin to hackers in 2014. Counterparty risk could help explain the consistent discount realized on BTC-E. Unlike Bitfinex and Bitstamp, BTC-E does not publish the location of its operations, and little is known about its owners. Such opacity may deter users from using the exchange for greater perceived probability of bankruptcy, which would endanger users’ accounts, or fraud.
While the methodology seems intellectually honest and the conclusions are not meant to disparage bitcoin, the title is a bit of a troll. When bitcoin proponents say that it is frictionless it is referring to the low friction of bitcoin payments, not fiat on and offramps which have their own frictions due to the banking system and state regulations. They actually reveal that they know they are strawmanning when they write:
Bitcoin-to-bitcoin transactions between digital wallets can be performed at a negligible cost relative to transaction amounts.
In total it's an interesting and well done paper, considering they are researchers working at the Fed.