A recent report into algorithmic trading by the Hong Kong Academy of Finance has cast an interesting light on this dark art. To put this into perspective, algorithmic trading accounted for:-
• 65% of total global equity market turnover in 2017, up from 25% in 2004
• 30% of foreign exchange trading in 2017, up from less than 2% in 2004
• 45% of European equity trading in 2017, from circa 3% in 2004
• 38% of Asian market equity trading in 2017, from around 1% in 2004
While the UK market is still well behind the US, where algorithmic trading accounted for 65% of turnover, it is markedly head of Asian markets.
What is algorithmic trading?
Algorithmic trading is also referred to as automated trade execution and is a low latency execution strategy which uses computer programs that follow a defined set of instructions. These instructions can be based on timing, price, quantity or any mathematical model. There are numerous benefits of this trade execution process, such as:-
• High-speed trading
• Ongoing analysis of prices and trends
• Trade frequency not possible for humans
Only since the introduction of electronic trading and the ongoing speed improvements has algorithmic trading been possible. As a result, you will often hear the term "low latency execution" concerning algorithmic trading. This is simply the speed at which your trading instructions move from the point of entry to execution, which can all be automated.
Algorithmic trading strategies
As trade execution is automatic within a low latency execution environment, it is possible to automate an array of trading strategies. These include:-
Technical indicators
The ability to monitor share prices and graphs on an ongoing basis while automating trade execution is beneficial for chart breakouts and technical indicators. Low latency executionmeans instructions are passed very quickly, carried out and reported almost instantly. The degree to which you wait until new trends are “more established” can easily be written into the algorithmic programme.
Arbitrage opportunities
There are numerous arbitrage opportunities, many of which require instant trade execution on both sides to benefit and remove the risk. This is where algorithmic trading shows through, carrying out relatively small but frequent arbitrage trades, thereby accumulating significant profits. In theory, if the two sides of the arbitrage are transacted simultaneously, it is possible to minimise or even remove risk.
Trading range algorithm
Often referred to as mean reversion trading, a trading range strategy means selling shares when they hit a high and buying them when they hit a low. The idea is that highs and lows are temporary, and eventually, the share price will snap back towards the mean average. The speed at which today's low latency execution trading platforms can operate means that some algorithmic trading strategies can, to a certain extent, dictate market movement – especially in relatively illiquid markets.
Mathematical model-based strategies
Due to low latency execution services, creating your own mathematical trading model is now possible. This may revolve around the relationship between a share price and a derivative, two share prices or the connection between a share price and a commodity. Whatever your requirement, it is possible to create a mathematical model based on your preferred trading strategy.
Summary
While automated trading/algorithmic trading is not necessarily favoured by everybody, it is a crucial element of the market. Due to the often frequent nature of this type of trading, there is evidence to show that it has a considerable impact on market liquidity and, by definition, "correct pricing". These systems have been made even more attractive by introducing low latency execution platforms that can pass your instructions and execute them in a split second.
While many people are concerned about “self-fulfilling prophecies” when using algorithmic trading, there is no definitive proof that this exists.