How abnormally large token transfers into exchanges can signal price movement
We recently released our second signal for token transfers moving into exchanges.
Specifically, users can choose to receive alerts for our second signal by adding an ERC20 token into their watchlist, and our bot will alert the user when a transaction larger than 99.9% of all historical transactions move into an exchange.
The focus of our research article today is to analyze the effectiveness of our new signal by doing case studies on our recently fired signal variants.
Case one - KNC experiences heavy sell pressure as ~$600K USD worth of KNC move to exchanges
The price of KNC had dropped by approximately 10% just less than 24 hours of three signals firing. In particular, a huge sell off occurred around Nov 13 1PM HKT causing the price of the token to plummet 3% in less than 5 minutes.
The vertical looking slope near the end of the price chart is the 3% price drop in less than 5 minutes. This is most likely the result of one or very few large sell orders liquidating as it had occurred so quickly, especially on a coin with relatively decent trading volume. This is also noted by the fact there wasn’t any obvious increase in the trading volume during that time.
As we had received three signals alerting us of large KNC movements into exchanges, we did expect the price of the token to be much more volatile. The three signals together had an aggregated balance of $600K USD worth of KNC, which made up more than a quarter of KNC’s 24 hour traded volume across all exchanges.
Any movement of such a degree would undoubtedly cause the price of the token to swing wildly within short periods of time.
Below are the three signals we had received. As a reminder, the signal variant only triggers when it’s a transfer larger than 99.9% of historical transactions.
For convenience, I’ve also posted the Etherscan links for each of the three transactions.
Transfer #1 - 558,408 KNC (~$198K USD) tokens into Binance
Transfer #2 - 676,637 KNC (~$240K USD) tokens into Binance
Transfer #3 - 446,000 KNC (~$158K USD) tokens into Binance
Case two - Is a whale of Decentraland (MANA) looking to sell on a high note?
A whale has recently transferred 13.5 million MANA tokens (~$1.3M USD) into Binance after MANA has gone up 25% in the last week.
For MANA investors who bought before Nov 9, 2018, it has been a very merry week as prices had gone from 1200 sats to over 1700 sats at its peak in just 5 days. This is despite a falling BTC which had gone from the mid $6,500s range down to the mid $6,300s range.
However, on a less granular level, it now looks like prices are starting to get pushed back down, with prices having broken the 1600 sat support levels.
Interestingly, this drop seems to coincide with the signal we received that alerted us on 13.5 million MANA tokens moving into Binance. This single transfer is equivalent to more than 10% of MANA’s 24 hour trading volume across all exchanges.
This seems like a case where a whale is looking to lock their profits in while the prices remain high. At 25% capital gains, the whale would be locking in well over $270K USD in profits if he manages to sell his tokens at the current price range.
As we can see, case examples like these show why it's important for investors and traders to monitor large token transfers into exchanges.
Importance of tracking token transfers into exchanges
As mentioned in our previous research article, movements of assets between personal wallets and exchange wallets can be seen as an intent to make a trade or hold. When we move assets into an exchange, we have an intention to make a trade in the short term, and when we move assets into a personal wallet, we have an intention to hold in the short term.
We also discovered that abnormal ETH movement is correlated with much higher price volatility, with ETH prices going down in the first 24 hours and going back up the following 24 hours.
For tokens moving in or out of exchanges, we are looking at essentially the same thing, but on a larger scale. Compared to ETH, most tokens have a much lower trading volume and market cap than its parent blockchain, which makes tokens naturally more sensitive to price movements. $100,000 in tokens are much more likely to cause a dent in prices than $100,000 in ETH going to exchanges. This was clearly seen in our two case studies above.
Because of this sensitivity, it is especially important for us to have a high completeness level in the number of exchange wallets tagged. That is, we want to ensure we track as exchange wallets as possible as any large token transfer going into or out of exchanges can have a material impact on the token’s price.
Although Etherscan has a list of labelled exchange wallets, this doesn't nearly cover all the exchange wallets that currently exist out there. Through a combination of machine learning and a team of blockchain experts, we constantly work to uncover unlabeled exchange wallets by finding characteristics in a wallet unique to exchange wallets.
You can find out more about how we identify exchange wallets here.