Cryptocurrency trading bots: how do they work?
Most trade deals in financial markets are done by automated trading systems or, more simply, by bots. The share of such transactions in the total volume of trading operations in the financial market is more than 80%.
Being in a legal vacuum, the crypto market is no exception — auto-trading and manipulations, prohibited in traditional finance, are also flourishing here. Moreover, crypto exchanges themselves often encourage high-frequency trading, bringing liquidity and, of course, considerable fee income to the platforms.
Many experts believe that the poorly regulated digital currency market is at the mercy of trading bots, which affect trading volumes and pricing. The problem of market manipulation is the subject of close attention from regulators, including the US Securities and Exchange Commission (SEC), which by rejecting bids to launch a bitcoin-ETF prevents institutional investors from entering the crypto market.
What is automated trading and why do I need trading bots?
Automated trading systems (ATS) or trading bots are programs that serve for full or partial automation of trading. They use special trading strategies in which bots open and close positions within a split second.
ATSs are often mistakenly identified with the term “algorithmic trading”. The fact is that the latter is not intended to make a profit. Algorithmic trading is a method of execution of a large order, when it is divided into several sub-order with special characteristics of price and volume. Each of them is sent to the market for execution at a certain time. Algorithmic trading is designed to reduce the cost of execution of a large order, minimize the impact of the latter on the market and reduce the risk of transaction failure.
Bots, on the other hand, interact with the stock exchanges through the API, receiving and interpreting market information, placing the appropriate buy or sell orders.
Bots’ actions are subject to the specified rules and algorithms. For example, a bot can take into account such market parameters as price, volume, available in the orderbook, time, etc. It can also take into account data from technical indicators, such as exponential moving averages (EMA) or Bollinger bands, as well as various variables set by the user. In addition, bots can interact with each other and different trading platforms (if arbitrage strategies are involved).
Trading bots allow a trader not to devote much of his time to market analysis and price movements on different currency pairs. Bots have advantages over live people, mainly due to their ability to make transactions in a fraction of a second, in the 24/7 mode. Unlike people, programs are devoid of emotions and can work with a large number of exchanges and currency pairs.
Bots are used by professionals, including fairly large financial companies, and “amateurs”, simple owners of cryptocurrencies, seeking to gain passive income and increase their capital. Progress is not static, algorithms become more complex, and different bots compete not only with people, but also with each other.
Types of bots and basic trading strategies
Trading bots differ among themselves in terms of complexity, principles of the device and, of course, the price. There are three main categories of such programs:
- Simple bots with predetermined logic;
- Software based on artificial intelligence and machine learning (“smart” bots);
- Advisors.
Simple bots operate on the basis of ready-made scenarios, which they are guided by in a given situation. The algorithms they contain can usually be edited, as it is unlikely that once configured bots will provide a stable income in the long term.
“Smart” bots are usually capable of self-learning. They can be based on neural networks and machine learning algorithms that increase the speed and depth of analysis. Such bots are usually more expensive and difficult to use.
Advisor robots may belong to both the first and second categories. It is clear from the name itself that such solutions make recommendations, not make deals. Such programs are also often used in the context of trust management, when trades are managed by a remote broker via API.
In automated trading various strategies are possible, among them:
- Arbitrage — earnings from the difference in digital coin prices on different exchanges (or between the underlying asset and its derivative);
- Market-making — taking advantage of the difference between the purchase and sale prices of coins and their derivatives.
When cryptocurrency trading was still in its infancy and market efficiency was even weaker than it is now, many traders earned on arbitration. In other words, they bought assets on one exchange and sold them on another at a higher price, earning profits from the difference excluding fees.
The fact is that a few years ago the market volume was much smaller than it is now. There were not so many people who traded, there were fewer exchanges and, accordingly, there was not so tough competition between the trading platforms. All this caused imbalances on different platforms, from which one could benefit. As the infrastructure developed and the market grew, price differences were smoothed out, and gradually this became less relevant.
Market-making strategy implies speculative profit. In accordance with price fluctuations, the trading bot places limit orders and gains profit on the difference between the buy and sell prices. For providing liquidity that improves the quality and attractiveness of the trading platform, traders can receive bonuses in the form of reduced fees.
Are trading bots effective?
As mentioned above, opportunities for arbitration in the crypto market are not so attractive anymore. As for market-making strategies, many bots, for example, are based on EMA and other lagging indicators. The values of such instruments are based on past history, which is an undoubted disadvantage. Thus, most automated trading systems perform analysis only in retrospect.
In general, many people criticize technical analysis referring to the key role of external, non-market factors. However, critics often overestimate its capabilities, forgetting that analysts’ assumptions are only probabilistic. Technical analysis is only an aid in decision-making, not a tool for rapid enrichment. Under conditions of not so liquid, volatile, subject to manipulations and acutely reactive to “loud” market news, it is difficult to predict anything ahead.
It should also be remembered that each market participant has its own tendency to risk. The latter represents the degree of readiness of a trader/investor to work with high-risk assets. This term is often still interpreted as the degree of uncertainty that an investor can afford with regard to a possible negative change in the value of his portfolio of assets.
All this means that certain trading strategies are suitable for some traders, but are completely unacceptable for others. Moreover, a bot that has shown good results will not necessarily be highly effective in the future. Ignoring this fact can lead to a so-called systematic survivor error.
Criticisms and risks of automated trading
Many experts believe that high-frequency trading puts excessive strain on the financial market infrastructure. Auto-trading systems can place several orders per second for each instrument, with only a small part of these orders leading to transactions. Thus, the exchange infrastructure is loaded, working idle most of the time.
Also automated trading systems involve risks of software, hardware or human errors. For example, bots played a significant role in the short-term decline of the US stock market in 2010, when high-frequency liquidity providers abruptly stopped operations. At that time, algorithmic and high-frequency trading were the subject of numerous proceedings initiated by the SEC and CFTC.
Another example is the so-called Knightmare on the New York Stock Exchange, which happened on August 1, 2012. At that time, the updated algorithmic engine of the Knight Capital Group company put out $3.5 billion in buy orders in 45 minutes and $3.15 million in sell orders due to mistakes in setting, and the market for some assets shifted by more than 10% due to incorrect software actions. Net loss of Knight Capital amounted to $460 mln. The next day the company declared bankruptcy.
In 2012, the European Parliament discussed the imposition of significant restrictions or even a total ban on high frequency trading. Despite the many proposals in both the EU and the US, few countries have introduced legal restrictions on high frequency trading.
One of the first countries was Italy, which on September 2, 2013 introduced a tax against high-frequency traders. Thus, transactions lasting less than half a second were subject to a 0.02% tax. At the same time, many researchers believe that automated trading improves market liquidity and helps reduce trading costs.
Bots and manipulations
Recently, the Wall Street Journal published an article that said that crypto market is at the mercy of trading bots, which significantly affect not pricing and trading volumes. Over time, however, the scale of the problem grows and becomes global.
In particular, unfair automated trading techniques such as spoofing and wash trades are flourishing on the cryptocurrency market. In the first case, a bot heaps up fake orders, creating the illusion of high market demand/supply and misleading traders. Wash trades imply a situation where the bot carries out buy/sell transactions with itself, creating the illusion of booming activity in the market, as well as artificially inflating trading volumes and asset prices.
Both types of operations are prohibited in traditional financial markets. For example, the New York Stock Exchange regularly monitors the correctness of trades and punishes violators.
Some ardent opponents of market regulation claim that there is nothing wrong with manipulation, and even openly support such actions. For example, trader Kjetil Eilersten thinks that it makes no sense for regulators to ban market manipulations. Instead, it is better to provide small traders with advanced tools to manipulate the market. According to Kjetil, this somewhat equalizes the possibilities of different participants.
“If everyone is manipulating, nobody is manipulating,” he says.
There are also bots whose purpose is the trade activity of large companies. Thus, the managing partner of Virgil Capital hedge fund Stephan Keene noted that his firm for some time caused a lot of trouble to some “bot-terrorist”, because of which Virgil suffered losses on deals with Ethereum.
As you know, crypto market is not so big and liquid yet. It means that even one such bot can affect the whole market. Besides, bots often play a key role in Pump&dump schemes, encouraging crowds of inexperienced market participants to buy coins at their all time high marks.
Manipulations on crypto market are increasingly seen by regulators of different countries. For example, last month the New York State Attorney General’s Office stated in its report on exposure of crypto exchanges to market manipulations. The agency also sent data on several exchanges to regulators due to possible law violations.
Automatic systems are not subject to an emotional factor in trading, they are guided only by calculations and clearly defined algorithms. Bots are able to detect trading signals against the background of market noise and process huge arrays of data, the processing of which cannot be handled by one person. If set up correctly, bots can generate passive income.
However, in the market, as in any other, even unemotional, cold calculation always involves risks that can be minimized, but cannot be completely eliminated. Also, bots are unlikely to be needed by Bitcoin maximalists and ardent supporters of the Buy & Hold strategy.
Trading bots are not the “Holy Grail”, even the most effective of them need periodic review of parameters. In other words, it is not a “free money” that you constantly get every day. However, there is a great opportunity to set up your own passive income source with our fair crypto lotto Ethex.bet!
