With the rapid development of computer technologies at the end of the 20th century the process of trading in financial markets changed and became completely electronic. There also appeared a separate segment of trading known as algorithmic trading.
Algorithmic trading is an automated system for placing and managing trading orders on various financial instruments through computer programs based on mathematical algorithms. In algo trading trades take place without human participation. An algotrader or a quant trader only describes the algorithm of behavior of the robot (mechanical trading systems (MTS)) in different situations in the programming language. Based on the analysis of the previous prices of financial instruments, they predict the probability of falling of the future price in a given range. Robot enters into a transaction or quits it in case of certain changes in the price chart of the trading asset. A popular method of algorithmic trading is considered to be High Frequency Trading (HFT), that is to say the conduction of electronic trading at very high speeds. High frequency robots with the aim of generating high profits open and close short-term positions with high volumes.
Algorithmic trading strategies
There are many strategies of algo trading, which are installed in a trading robot by programmers. Here are its main strategies:
VWAP (Volume Weighted Average Price) – Distributes the volume of requests uniformly within a certain period of time at the price of better supply or demand, but it does not exceed the volume weighted average price over a specified period.
TWAP (Time Weighted Average Price) – Executes requests and evenly divides them into equal time intervals. The strategy does not consider predicted changes of trading volumes, which may negatively affect the market.
Percentage of Volume – Supports the fixed percentage of participation in the market chosen by a user. It makes small and frequent transactions by well reacting to the jumps of volume.
Iceberg – Set selling or buying request, which does not display the entire size of the market requests. Potential buyers see only a part of the request and only after its execution the next part is being published. And this continues until its full implementation.
Trend – following strategy – the main objectives of the strategy are: the early detection of the emerging trend through various indicators of technical analysis, the release of signals for trading in the direction of a trend and the release of signals about the closure of the position when signs of a trend termination appear.
Arbitrage – Foreign Exchange market robot, fixing divergence of the prices on the same or equivalent instruments in various market places, buys cheap in one place and immediately sells in another place with the expectation that prices of the instruments will coincide and the positions will be closed with profits. Arbitrage is considered to be a risk-free strategy, because the robot buys assets for a short period of time, thus avoiding sudden price fluctuations during the time. Accordingly, the income from arbitrage transactions is also insignificant and the total profitability is formed by the frequency of transactions.
Scalping – a strategy for short-term intraday speculative transactions. High-frequency robots are the most commonly used robots for scalping, which open and close positions during seconds, in case of making a small profit in a few pips. Basically, the strategy is used in the derivatives market, where the commission from the turnover is significantly lower.
Pair trading or statistical arbitrage – the strategy aims to identify the correlation between various instruments of the market and make profits from the imbalance between them. In other words, in small time intervals one asset can be undervalued or overvalued against the other one. Robot uses that very moment by fixing the deviation of the current ratio from the value of its moving average.
Algorithmic trading with all its advantages regarding the speed of trading, absence of emotions, providing high market liquidity, decrease of volatility in the market, etc. has also several disadvantages:
- High-frequency algorithmic traders often make the market operation complicated, by making an excessive number of requests.
- Unreasonable increase of volatility of the market. For instance, on May 6, 2010 for a few minutes, the Dow Jones Index dropped 8.6% (the loss of the market was more than $1 trillion). Then, during 90 seconds, the index regained 543 points (4.67%). The reason was that the high-frequency robots in case of uncertainty liquidated all their positions. The sharp outflow of liquidity on the background of the started drop of the index led to its excessive strengthening without any economic basis.
- Failure of the algorithmic systems. There are some cases, when the major players of the market were on the verge of bankruptcy because of the failure of the program.