In the traditional world of securities exchanges, it would be akin to defying gravity. But at Kraken, one of the most popular venues for trading cryptocurrencies, a coin called Tether does it again and again.
Huge trades move prices about the same as small ones, ignoring the normal rules of economics, according to a review of Kraken’s public order book—a pattern that experts on market manipulation view as a red flag. The mystery is bracketed by another quirk: Oddly specific order sizes—many going out to five decimal points, with some repeating frequently. Another red flag.
Andrew Rennhack, a former professional poker player who’s taken an interest in cryptocurrencies, got the discussion started when he began downloading data from Kraken’s order book and puzzling publicly about what he found. Bloomberg News did its own analysis, pulling more than 56,000 trades placed on the exchange between May 1 and June 22 and sharing the information with New York University Professor Rosa Abrantes-Metz and former Federal Reserve bank examiner Mark Williams. Both agreed that they’d never seen a market behave like Kraken, where large Tether orders fail to sway prices much.
Their questions join many others swirling around Tether, one of the world’s most-traded cryptocurrencies. On June 13, University of Texas Professor John Griffin, known for flagging suspicious activity on Wall Street, released a research paper focusing on the use of Tether to buy Bitcoin on another exchange, Bitfinex. He concluded transactions elevating Bitcoin’s price were “consistent with a manipulation hypothesis.” Bitfinex disputed his findings. Bitcoin tumbled following the report.
Read more about John Griffin’s Bitcoin report
At Kraken, the third-most-common trade during the period Bloomberg examined was for 13,076.389 Tethers. (Behind 75 and 1,000.) Many other orders go out to five decimal places, numbers like 34.08652. Abrantes-Metz and Williams suspect that such numbers could be signals to cheaters’ automated trading programs. One possible explanation: The software would look for orders with a unique size, and trade against that. Taking both sides of a transaction is known as wash trading, something banned in regulated markets like stocks because it can give a false impression of market supply and demand. Kraken isn’t similarly regulated by the government.
Tether-USD*
Number of trades:
5K
75
4
Trades at a volume of 75 were by far the most common, and made up 7.9% of total trading activity on Tether between May 1–June 22.
1000
3
13076.389
2
1
0
2.00E-08
5
10
29.98441
29.98681
29.9985
30.0054
30.0075
30.01651
30.02642
30.02763
30.03033
30.05199
30.06946
30.07127
30.10174
30.11262
30.12623
30.21939
31.43768
31.52276
32.81565
33.90948
33.97089
34.03828
34.04254
34.0426
34.05478
34.06503
34.14089
34.16856
34.2725
34.30535
34.31576
34.36073
37.5
50
75
100
125
200
250
500
925
1000
2500
9000
13076.389
Bitcoin-USD
Number of trades:
25K
The most common volumes traded on Bitcoin tend to be simpler, and the top three combined made up 8.1% of activity in this market.
1.0
20
0.1
15
0.01
10
5
0
1.00E-08
4.40E-07
0.00099956
0.001
0.002
0.0023
0.0025
0.003
0.004
0.005
0.006
0.007
0.008
0.009
0.01
0.011
0.012
0.013
0.014
0.015
0.016
0.02
0.025
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.12
0.15
0.19
0.2
0.25
0.3
0.35
0.4
0.48604187
0.5
0.6
0.8
0.97900622
1
1.5
1.79
2
3
5
Tether-USD*
Number of trades:
5K
75
4
Trades at a volume of 75 were by far the most common, and made up 7.9% of total trading activity on Tether between May 1–June 22.
1000
3
13076.389
2
1
0
2.00E-08
5
10
29.98441
29.98681
29.9985
30.0054
30.0075
30.01651
30.02642
30.02763
30.03033
30.05199
30.06946
30.07127
30.10174
30.11262
30.12623
30.21939
31.43768
31.52276
32.81565
33.90948
33.97089
34.03828
34.04254
34.0426
34.05478
34.06503
34.14089
34.16856
34.2725
34.30535
34.31576
34.36073
37.5
50
75
100
125
200
250
500
925
1000
2500
9000
13076.389
Bitcoin-USD
Number of trades:
25K
The most common volumes traded on Bitcoin tend to be simpler, and the top three combined made up 8.1% of activity in this market.
1.0
20
0.1
15
0.01
10
5
0
1.00E-08
4.40E-07
0.00099956
0.001
0.002
0.0023
0.0025
0.003
0.004
0.005
0.006
0.007
0.008
0.009
0.01
0.011
0.012
0.013
0.014
0.015
0.016
0.02
0.025
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.12
0.15
0.19
0.2
0.25
0.3
0.35
0.4
0.48604187
0.5
0.6
0.8
0.97900622
1
1.5
1.79
2
3
5
Tether-USD*
Number
of trades:
5K
75
4
1000
Trades at a volume of 75 were by far the most common, and made up 7.9% of total trading activity on Tether between May 1–June 22.
3
13076.389
2
1
0
2.00E-08
5
10
29.98441
29.98681
29.9985
30.0054
30.0075
30.01651
30.02642
30.02763
30.03033
30.05199
30.06946
30.07127
30.10174
30.11262
30.12623
30.21939
31.43768
31.52276
32.81565
33.90948
33.97089
34.03828
34.04254
34.0426
34.05478
34.06503
34.14089
34.16856
34.2725
34.30535
34.31576
34.36073
37.5
50
75
100
125
200
250
500
925
1000
2500
9000
13076.389
The most common volumes traded on Bitcoin tend to be simpler, and the top three combined made up 8.1% of activity in this market.
Bitcoin-USD
Number
of trades:
25K
1.0
20
0.1
15
0.01
10
5
0
1.00E-08
4.40E-07
0.00099956
0.001
0.002
0.0023
0.0025
0.003
0.004
0.005
0.006
0.007
0.008
0.009
0.01
0.011
0.012
0.013
0.014
0.015
0.016
0.02
0.025
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.12
0.15
0.19
0.2
0.25
0.3
0.35
0.4
0.48604187
0.5
0.6
0.8
0.97900622
1
1.5
1.79
2
3
5
Tether-USD*
Trades at a volume of 75 were by far the most common, and made up 7.9% of total trading activity on Tether between May 1–June 22.
Number of trades:
5K
75
4
1000
3
13076.389
2
5
1
0
Bitcoin-USD
The most common volumes traded on Bitcoin tend to be simpler, and the top three combined made up 8.1% of activity in this market.
Number of trades:
25K
1.0
20
0.1
0.01
15
0.002
0.05
10
0.5
5
0
No evidence that Kraken itself is involved in any manipulation has emerged. “Nothing looks out of place to us in our publicly available data feed,” Kraken Chief Executive Officer Jesse Powell said in a statement. Regarding the trades Bloomberg examined, he added: “We have not verified the legitimacy of the data set you asked us to review.”
Powell has previously dismissed the need to police for cheating in cryptocurrencies. In an April blog post he wrote that being protected from manipulation “doesn’t matter to most crypto traders” and likened efforts to curtail it to “mind control.” His broader point was that most crypto traders don’t care about an industry examination begun by former New York Attorney General Eric Schneiderman. Powell said in the same post that he wasn’t opposed to working with the government.
In any case, the U.S. government is paying a lot of attention to the overall $248 billion crypto market. The Justice Department is conducting a criminal probe into whether traders are manipulating the price of Bitcoin and other digital currencies, with techniques including wash trades, Bloomberg revealed last month.
And the U.S. Commodity Futures Trading Commission has been examining Tether. In December, the agency subpoenaed Tether and Bitfinex, seeking proof that Tether is backed by a reserve of U.S. dollars, Bloomberg reported in January. Neither have been accused of wrongdoing. In June, Tether released a report in which Freeh Sporkin & Sullivan LLP, a law firm co-founded by former FBI director Louis Freeh, said Tether had $2.55 billion deposited in banks as of June 1. The analysis wasn’t an official audit, however.
Read more about the mysteries shrouding Tether
Tether fetches about $1 because each coin is supposed to be backed by $1 in a bank—letting it play a pivotal economic role in the cryptocurrency market. It’s pitched for its stability—a haven for investors when prices go wild for other coins. Because many venues are unable to secure bank accounts to give their customers access to dollars or other traditional currencies, investors are often paid with Tether when they cash out.
Kraken is among very few markets worldwide that let investors trade U.S. dollars for Tether and vice versa. So Kraken should play a large role in establishing Tether’s price.
But the normal economics of supply and demand don’t always appear to apply.
Take May 7, for example. Within a span of less than two minutes, there were 31 consecutive trades to buy the currency, representing a total of 159,487 Tethers. But Tether’s price didn’t budge from 1, despite the unrelenting purchases.
Price
Volume (Sell)
Within a span of 2 minutes on May 7, there was a tight sequence of buy orders with a total volume of 159,487.1013 that did not move the price
PRICE: $1.00
Price
Volume (Sell)
Within a span of 2 minutes on May 7, there was a tight sequence of buy orders with a total volume of 159,487.1013 that did not move the price
PRICE: $1.00
Price
Volume (Sell)
Volume (Buy)
Within a span of 2 minutes on May 7, there was a tight sequence of buy orders with a total volume of 159,487.1013 that did not move the price
PRICE: $1.00
Price
Volume (Sell)
Within a span of 2 minutes on May 7, there was a tight sequence of buy orders with a total volume of 159,487.1013 that did not move the price
PRICE: $1.00
On June 3, someone bought 37.5 Tethers, and the price rose 0.0002. That was immediately followed by another purchase of 37.5 Tethers, which didn’t move the price from 0.9999. Then came one of those orders for 13,076.389 Tethers, this time a buy. Even though it was 349 times bigger than the previous trade, it nudged Tether’s price up by 0.0001, less than the much smaller order.
Price
Volume (Sell)
Volume (Buy)
13076.389
The next buy order of 13,076.389—349 times larger than the previous one—nudges the price up by 0.0001, less than the much smaller order
A buy order of 37.5 increases the price by 0.0002
4895.630579
$1.000
37.5
37.5
$.9999
PRICE
$.9997
VOLUME
925
Price
Volume (Sell)
Volume (Buy)
13076.389
The next buy order of 13,076.389—349 times larger than the previous one—nudges the price up by 0.0001, less than the much smaller order
A buy order of 37.5 increases the price by 0.0002
4895.630579
$1.000
37.5
37.5
$.9999
PRICE
$.9997
VOLUME
925
Price
Volume (Sell)
Volume (Buy)
13076.389
A buy order of 37.5 increases the price by 0.0002
The next buy order of 13,076.389—349 times larger than the previous one— nudges the price up by 0.0001, less than the much smaller order
4895.630579
$1.000
37.5
37.5
$.9999
PRICE
$.9997
VOLUME
925
Price
Volume (Sell)
Volume (Buy)
13076.389
A buy order of 37.5 increases the price by 0.0002
$1.000
37.5
37.5
$.9999
PRICE
$.9997
VOLUME
925
The next buy order of 13,076.389—349 times larger than the previous one— nudges the price up by 0.0001, less than the much smaller order
On May 9, eight sell trades for 13,076.389 Tethers each occurred in succession over 16 seconds, yet Tether’s price was unchanged at 0.999. The next trade—for just 75 Tethers—pushed the price up 0.0001.
Price
Volume (Sell)
Volume (Buy)
$.9992
...but a buy order of just 75 moved the price up .0001
Multiple sell orders of exactly 13,076.389 are executed, and the price didn’t budge...
200
200
75
$.9991
$.9991
$.9990
$.9990
$.9990
13076.389
13001.389
13076.389
Price
Volume (Sell)
Volume (Buy)
...but a buy order of just 75 moved the price up .0001
$.9992
Multiple sell orders of exactly 13,076.389 are executed, and the price didn’t budge...
200
75
$.9991
$.9991
$.9990
$.9990
13001.389
13076.389
Price
Volume (Sell)
Volume (Buy)
Multiple sell orders of exactly 13,076.389 are executed, and the price didn’t budge...
...but a buy order of just 75 moved the price up .0001
$.9992
200
75
$.9991
$.9991
$.9990
$.9990
13076.389
13001.389
13076.389
Price
Volume (Sell)
Volume (Buy)
Multiple sell orders of exactly 13,076.389 are executed, and the price didn’t budge...
...but a buy order of just 75 moved the price up .0001
$.9992
200
$.9991
$.9991
$.9990
13001.389
13076.389
“Large trades are not impacting prices,” said Abrantes-Metz, the NYU professor and an expert on manipulation who has testified about rigging of the globally used Libor interest-rate benchmark, among many other cases. “I’ve looked through lots and lots of data, and I don’t think this is real,” she said, referring to the trading on Kraken.
Such patterns appear throughout the data set Bloomberg downloaded from Kraken. While buying was greater than selling—about $50 million versus $32 million from May 1 to June 22—Tether’s price only ranged from 1.0095 to 0.989. Amounts such as 34.05478 and 30.06946 and 34.14089 are relatively common. Why would trades be so precise?
“It could be a signal, because it is really, really strange,” Abrantes-Metz said.
Another expert also shared misgivings. “Many of the trade amounts are frequently occurring to the fifth decimal point, a unique identifier which increases the probability it is being generated by the same person or entity,” said Williams, the former Fed examiner who studies risk management as an executive-in-residence at Boston University. The supply-demand situation at Kraken is unusual, he said. “Counter to basic trade economics, large and frequent trade volumes appeared to have less influence on price than small trade volumes.”
Bursts of Tether buying on Kraken don’t seem to drive its price higher, Griffin said in an interview. During periods of heightened demand, one would expect its price to reach perhaps $1.10, leading to issuance of more Tether that moves the price back to $1, he said. “But we’re not seeing that.”
He said the oddly specific order sizes in the data set Bloomberg downloaded are “suggestive of wash trading”—a maneuver, banned in regulated markets, in which cheaters trade with themselves to create a false impression of market demand. (Crypto markets, in contrast, are relatively unregulated). “Not sure what the motivation is,” he added.
Rennhack, the data sleuth, was originally long Bitcoin, betting it would rise. But after he started reading online warnings about anomalies, specifically from the Twitter account @Bitfinexed, he sold and went short.
“The more I looked into it, the more sketchy it seemed,” he said.
Dave Weisberger also has questions. Before creating cryptocurrency data and order-routing company CoinRoutes, he was among the foremost experts in the inner workings of the $30 trillion U.S. stock market following a career that included stints at Citigroup Inc. and Two Sigma Securities. He can’t wrap his mind around why orders for 13,076.389 Tethers keep popping up on Kraken. “No human would enter that order,” he said. “It doesn’t make sense.”
The Tether trades he’s reviewed on Kraken are unusual enough that, if they were happening on a stock exchange, regulators would be investigating the trades, Weisberger said. “Is there manipulation? I don’t know, but I know they’d be digging into it.” Such mysteries are probably keeping big investors out of cryptocurrencies, he added. “Institutional investors, before they commit their capital, want to see the market is fair.”
June 30, 2018: Updates sixth paragraph of story published June 29 with CEO comment