Over the past week, we have seen a couple of cases where the only way to explain the price movement in Bitcoin is if traders are dumb, evil or a combination of both. Before delving into the specifics, it is important to understand that the only way such things will cease occurring is if traders get smarter and start using tools (such as provided by CoinRoutes) that incorporate data from all relevant markets.
The two examples that will be described in this note were the price drop of over $400 this past Sunday morning and a price divergence which happened yesterday morning (Thursday), where markets went in opposite directions simultaneously.
Example # 1: Bitcoin plunges $400 on a Sunday morning…
The technicals — BTC had a sharp, $200 coordinated rally on Saturday after grinding higher most of the week and then continued to trade sideways, after the move through early Sunday morning. At that time, however, bitcoin suffered a straight line, sloppy fall of $400, followed by stabilization.
What does this mean? First, it’s instructive to compare the moves. The up move was about half the size of the down move and also took place on multiple exchanges in a coordinated fashion. That up move was roughly a $200 spike, which looks from an outside perspective as being the result of a buyer checking with multiple OTC desks for liquidity, with the result that the market was impacted and the move took place. I postulate this because the move was well coordinated between exchanges, indicating that relatively sophisticated firms were doing the buying.
The downward move on Sunday, on the other hand was most likely either an example of stupidity or manipulation. Stupidity, if it was a legitimate seller, as they indiscriminately hit bids on exchanges driving down the price while creating arbitrage opportunities due to the sloppy nature of the selling. It is also important to note that the downward move took place during the least liquid time of the trading week (Sunday Morning US time as Asia is sleeping and most US and Euro participants are not trading).
The following “zoom in” charts, produced by the CoinRoutes software, describe these two moves well. First, the upward move, where there is almost no separation between the best bid and best offer:
Next, we zoom in on the downward move, where one can see a large gap between the lowest offer and the highest bid that reached over $150 at one point:
These charts show that, when the drop of $400 occurred, the seller(s) did not take as much care with their orders as the buyers did, and, most likely sold too aggressively on the markets or with the broker they were using.
Unfortunately, however, there is another potential reason that this occurred, which is quite troubling (hence the use of the word “evil” in the title). It could have been an example of market manipulation.
To explain the potential strategy a manipulator might have used, consider the following. Lets say that a firm, planning this on Friday, started to accumulate BTC while simultaneously hedging that position by establishing a short futures position on a market such as Bitmex. After they had accumulated enough BTC in the spot market, they might well have continued to sell futures for a couple of hours, as the liquidity, particularly over the weekend, is larger in the futures market. Sunday morning, they might have then dumped all the BTC they had accumulated over the week as sloppily as they could, with the goal of driving down the futures price, so they could gain on their short position. I want to make clear that I have no direct knowledge that such a strategy was undertaken, but it certainly could have been used. Such trading strategies, often referred to as “liquidity arbitrage” are considered manipulative by all major regulators today. (Although it should be noted that such trading strategies were undertaken by the largest players in markets such as Japanese index futures and stocks in the 1990s quite regularly.)
Example #2: Bitcoin traders go in different directions
As I was doing a demonstration of the CoinRoutes platform yesterday morning (Thursday at around 1040 am) we noticed that CoinRoutes RealPrice, was showing a higher bid for Bitcoin than offer. While crossed markets in the gross price of Bitcoin are common, it is quite rare to see RealPrice crossed. That is because RealPrice takes into account fees as well as the size of orders on the book and, in Bitcoin, the fees are typically much larger than the spread. When we investigated what was happening, we saw that the offer price on Coinbase Pro was roughly $90 below the bid price on itBit, Gemini, and other markets. After about 20 minutes, the situation returned to normal, but the graphs from the CoinRoutes historical tool showed the divergence quite well. First, lets look at the 24 hour chart to put the move into context:
Notice the strange looking volatility with both up and down moves happening almost simultaneously, representing what looks to be around one hundred dollars of difference between the bid and offer. To see what actually happened, we can zoom in to the specific period where the divergence occurred:
As you can see from the graph, traders on each individual platform acted without knowledge of what was happening on other markets, bidding up the price of Bitcoin on some markets while simultaneously pushing it down on another. This, of course, created unnecessary volatility, but it is important to point out that those investors who sold at the lower prices or bought at the higher prices lost a fair amount of money. It is important to understand that institutional investors are troubled by incidents such as this, as such outcomes contravene reasonable principles of best execution. The good news, however, is that traders who use trading algorithms, such as CoinRoutes provides, will have an enormous advantage, as they will be on the correct side of such divergent trades.
It is also worth noting that it is unlikely that this example had a malicious intent, as it is hard to see how such trading could be considered manipulative.
Conclusion:
While these two events are different in scope and likely cause, they both point out that it is critical for crypto investors to trade with systems that are aware of the prices on multiple markets. Without such knowledge, events like these will continue to happen, damaging investor confidence. If, however, a critical mass of crypto investors use such tools, the ability of such algorithms to systematically exploit price divergences while trading will dampen their impact. Such developments, after all, are a natural part of the evolution of financial markets and would be particularly welcome for crypto markets.