Cryptocurrency market fall
On September 24, we observed a sharp drop in the price of Bitcoin and followed by the drop in the entire cryptocurrency market. The answer to the question whether the market sentiment has changed from bull to bear will give us some figures.
First of all, we looked at movement of count of trades and volume over the past 24 hours of some top exchanges on the pairs BTC / USDT, BTC / USD. Our greatest interest was on Huobi Exchange, with the clearly unstressed movements of these two indicators (Fig. 1).
The disparity between increase/ decrease in the number of trades and volume, and sometimes divergence, may tell us about the presence of the manipulation in trading on exchange. For example, over the same time period, the graphically considered indicators on Binance exchange on BTC / USDT look like this (Fig. 2):
We can see an increasing number of trades and volume (red arrows Fig. 1, 2) during the emergence of the largest red candle in Bitcoin price on September 24, but looking at Huobi, we can see that the increasing of volume was bigger. This gives us the reason to assume that there were large market orders to sell with the full execution of a number of limit buy orders in order book of BTC/USDT. And then unnatural behavior (divergence) of these indicators when the price of Bitcoin has reached the local bottom (green arrows Fig. 1).
These unnatural changes have affected limit orders. In the visible part in order books (up to a hundred positions for sale and buy). We saw a not entirely logical change on some top exchanges in amount of Bitcoins paired with stablecoins and fiat currencies, but what caught our attention was a dropped number of Bitcoins paired with altcoins which happened after drops of tickers and before delisting a small list of pairs (Fig. 3).
The red line is Bitcoin price movements from June 2019 to the present, the blue line is the sum of all bitcoins in the visible part of order books of BTC with altcoins on Binance exchange. This drop was from 23.2 thousand Bitcoins, up to 6 thousand. Nobody answers on where are Bitcoins, the exchange managers will not tell us, because it was the main provider of liquidity on the buyer side in pairs with altcoins.
We may assume that several players on exchanges knew in advance about that impending price would fall and initiated it together. We mat also suppose that this should not be without knowledge of the main player (Tether) on the cryptocurrency market. Previously, there have been many studies on the correlation of USDT emissions and Bitcoin prices. Below is one of them.
In TokenAnalyst’s tweet, we don’t see a strong inverse correlation between USDT ERC-20 emission and USDT Omni burning, but this is another question … Our next goal was to see if there were anomalies in large wallets that we associate with Tether (1st level — Fig. 4): 0x5754284f345afc66a98fbb0a0afe71e0f007b949,0xc97a4ed29f03fd549c4ae79086673523122d2bc5,0x6485fc77be2186fc60feea38d0ef4331d8404b60. Next are transactional wallets 2nd and 3rd levels (Fig. 4) These are subsequent wallets that were somehow connected to cryptoexchanges on Ethereum network. However, we did not see any abnormal transactions. Try it yourself, a convenient option for the reader maybe would be sites bloxy.info or etherscan.io.
Intermediate wallets between exchanges (Fig. 4) are wallets allocated by us, connecting exchanges to the largest volume of USDT ERC-20 transactions.
In this case we were interested not in arbitrageurs’ wallets among them, but in transfer-wallets between exchanges, such as:
0x91A3317308A2b60c849788Fe7a72589CD48E7159 — total amount of USDT transfers through it to Huobi from Binance, Okex and a small number of third-party wallets about 378 million tokens without reverse transfers;
0x3ad63fe9ce75874cd4f88422c44bfd17cb5f4754 — the next wallet to replenish Huobi, mainly by Binance, a little with Okex and third-party wallets, the amount of transfers is 138 million;
0x6cc5f688a315f3dc28a7781717a9a798a59fda7b — total amount of USDT transfers through it to Okex from Binance, Huobi and a small number of third-party wallets about 330 million tokens without reverse transfers;
0x3f5CE5FBFe3E9af3971dD833D26bA9b5C936f0bE — total amount of USDT transfers through it to Binance from Okex, Huobi and a small number of third-party wallets about 267 million tokens without reverse transfers.
Other USDT intermidiate wallets addresses for Huobi: 0x97ee7b5740ba5831f76af9efa576bbd0309b07c0, 0xb98232c3e43aa16a9d28350ac279eed519aae213, 0x9679601a09247e2eec3500f4bb6560049e457e46, 0x5f86fe0e62d1f0226f535ebe5d98a5d3edc9a615; for Binance: 0x94f0ceaebf8d95032a06732230265503d711488d, 0x18a065a08a806bc8f49edc7581c0af05b4c71dce, 0x25959edb755d176d067803189af214f23bda94b2; for Okex: 0x23ce5fabcb45e3dd3bdc458a10f21e4fad54f663, 0xbf92681aea2c163776feebcbc57c5dcb26d56a82; for gate.io: 0x744727a6fc563f54fd0f6f0442c0bd1e212011f9.
This is not a complete list of wallets for rebalancing USDT-ERC20 between Asian cryptocurrency exchanges. Analyzing the inputs and outputs of USDT on these wallets, we saw that volume of transactions of stablecoin on exchanges decreased after the fall. Moreover, using data on these wallets, we can see an increasing activity before fall comparing with mean values.
The increased activity of these wallets indicates a high probability of joint actions of Huobi, Binance and Okex exchanges. However, we cannot assert with absolute certainty that these exchanges are in collusion without owning accurate information on Bitcoin transactions between these exchanges.
However, if you look at information in open sources, it becomes clear that Binance, Huobi and Okex exchanges are closely connected to BitFinex exchange and its parent company iFinex Inc, which is the issuer of stablecoin USDT Tether.
It can be assumed that it was these participants in cryptocurrency market that could manipulate the price of Bitcoin. The main motive in this case may be the ongoing trial between the US prosecutor’s office and IFinex. It seems logical to try to lower price of Bitcoin in order to earn on closing margin positions on these exchanges, as well as IFinex, to make a profitable purchase of a large number of cryptocurrencies (especially Bitcoin) using the USDT token. It is also worth mentioning the margin positions on exchanges such as BitMex and CME and their interest in such a price movement, but establishing their involvement in this fall in this study failed.
In this regard, the fall of September 24 is considered as misunderstanding, and we expect early return of Bitcoin to its local maximum.
