Japan may not be quite the cryptocurrency powerhouse that the world thinks it is.
A deep dive by CoinDesk has found methodological flaws in widely-cited bitcoin exchange data that appear to overstate the importance of the Japanese yen as a trading pair. Our analysis of trading data collected from July 26-30 suggests that the U.S. dollar, not the yen, is the dominant currency traded for bitcoin by a wide margin.
Currently, analytics sites CryptoCompare and Coinhills offer a breakdown of bitcoin trading by currency pair, and until recently the data from both sites indicated that over 50 percent of bitcoin trading is denominated in Japanese yen.
The problem is that the vast majority of yen-denominated transactions are not “spot” trades of actual bitcoin for yen. Instead, they are derivative products: contracts that derive their value from the performance of an underlying asset.
In other words, the parties to these transactions are betting on the price of bitcoin, but no bitcoin is actually changing hands. While there’s nothing inherently wrong with these contracts, selectively mixing derivative and spot volumes can paint a misleading picture.
Specifically, Coinhills and CryptoCompare, whose data has been cited by major outlets such as Bloomberg and the Wall Street Journal, did not distinguish between the spot and derivative volumes of Bitflyer, Japan’s largest exchange. In other words, both types of trading were counted toward the total for yen-bitcoin activity.
However, their calculations exclude equivalent dollar-denominated derivative markets such as those on Bitmex.
As a result, the yen and dollar totals were not an apples-to-apples comparison, since the former includes derivative trades and the latter does not. When correcting for the misclassification, the data compiled by CoinDesk paints a starkly different picture.
To be sure, Japan remains a global hotspot of cryptocurrency interest, thanks in part to a law that took effect early last year recognizing bitcoin as legal tender and regulating the country’s exchanges. And the CoinDesk analysis covers only a five-day period, so it isn’t quite conclusive evidence that the U.S. dollar underpins most cryptocurrency trading.
Still, after being contacted by CoinDesk, CryptoCompare changed its methodology, and its data now shows the dollar out-trading the yen.
Further, the inconsistencies in how different types of transactions are counted across exchanges, and the wildly different results when these are addressed, underscore the nascent state of the cryptocurrency industry’s data practices.
As mentioned, the issues with currently available data stem from the classification of Bitflyer’s various exchange markets.
Bitflyer handles both spot trades and derivative trades, but it is the sheer scale of these derivatives trades that can distort market data.
During the time period of CoinDesk’s snapshot, Bitflyer’s Lightning FX derivatives service processed the equivalent of nearly $2 billion in yen-denominated trades every day. These derivative trades accounted for 90% of CryptoCompare’s observed yen trading and 85% of Coinhills’ yen trading.
By itself, the inclusion of derivatives volume would not necessarily be a problem if CryptoCompare and Coinhills also counted dollar-denominated derivatives markets when tallying the total volumes. Trouble is, they don’t.
For perspective, Bitmex, the largest dollar-denominated derivatives market, recently set a record of over $8 billion of contracts traded in a single 24-hour period (July 23-24), dwarfing Bitflyer’s volumes. These derivative trades are uncounted in CryptoCompare’s and Coinhill’s respective measures of total USD-BTC trading.
To create a more balanced comparison of global trading activity, CoinDesk took a snapshot of both the spot market, which excludes all derivative trading, and the total market, which includes all spot trading and the dollar yen volume at the four major derivative exchanges: Bitflyer’s Lightning FX, Bitmex, CME and Cboe.
Both these measures indicate that the U.S. dollar dominates worldwide, with the yen a distant second.
For example, the dollar accounted for only 17 percent of CryptoCompare’s tally and 21 percent of Coinhills’ figure for those five days in July. In CoinDesk’s apples-to-apples snapshot, by contrast, the dollar makes up 56 percent of spot market trading and 68 percent of total trading when including major global derivative markets.
The findings are notable as the spot market is particularly relevant to officials concerned about potential illicit uses of cryptocurrency.
Any unsavory actor attempting to make use of ill-gotten gains today must rely on spot exchanges to convert cryptocurrencies into fiat currency.
The dollar’s role in this exchange ecosystem extends the reach of the U.S. government. Just as the dollar’s preeminence in the international financial system gave U.S. regulators the leverage they needed to shape global anti-money laundering (AML) practices in the wake of 9/11, the dollar’s importance in global fiat-to-cryptocurrency trade could give them outsized influence as governments around the world mull new regulatory frameworks for cryptocurrencies.
For example, Japanese officials have thus far led the push for global cryptocurrency AML standards in international forums like the G20 and the Financial Action Task Force, but continued dollar dominance of cryptocurrency trading could inspire a more active U.S. presence.
Yaya Fanusie, director of analysis at the Center on Sanctions and Illicit Finance at the Foundation for Defense of Democracies and co-author of a report on bitcoin laundering, told CoinDesk:
“If your assumptions are correct and the dollar in fact dominates on crypto exchanges around the globe, this data could prompt U.S. regulators to take a more active role.”
To be fair, the uncertainty, complexity and constant churn of the exchange market leaves analytics sites like CryptoCompare, Coinhills and others struggling to keep up. In such an environment, errors are bound to occur, even with data from large, highly regulated exchanges like Bitflyer.
When contacted about the inclusion of Bitflyer’s derivative data, both CryptoCompare and Coinhills acknowledged that data from Bitflyer’s Lightning FX derivatives market was the root cause of their observed yen dominance.
“We do currently count Bitflyer’s Lightning FX volume, and thank you very much for pointing this out. We plan on excluding Bitflyer’s futures volumes from calculations at the end of this month,” Constantine Tsavliris, an analyst at CryptoCompare, told CoinDesk.
Subsequently, CryptoCompare, which recently announced a data partnership with Thomson Reuters, indeed removed Biflyer’s derivatives market data from its calculations.
A Coinhills representative stated that the company is exploring adding other major derivatives markets such as Bitmex.
While the industry continues to mature by the day, it remains the Wild West for all observers hoping to gather a clear image of the cryptocurrency exchange market.