Publication date: May 14, 2025
New technologies requires a new approach to trading, where the speed of placing orders and the analysis of emerging trading options, while ensuring risk protection, becomes crucial for the success of trading activities. For this reason, we can currently observe the phenomenon of so-called algorithmic trading on the global market and on regional markets.
Managers are now flooded with data, and their processing, analysis, understanding and interpretation take a lot of time. Hence the growing importance of the role of automated decision-making systems (ADM) and replacing humans with robots. Artificial intelligence is mainly a technology, while ADM systems are social engineering processes, consisting in delegating part of the decisions to computer systems, often using artificial intelligence. Therefore, the term “ADM systems” is used to describe the entire decision-making process that replaces humans, being an algorithm stored in computer code, characterized by various levels of sophistication and complexity. It is precisely on such systems that algorithmic trading is based. Also called quantitative or automated trading, in simple words it describes the process of using computer programs to automate the process of trading (buying and selling) financial instruments (stocks, currencies, cryptocurrencies , derivatives). These computer programs are coded to trade based on the input data that has been defined for them. The input data can be based on a targeted strategy to take advantage of various market behaviors, such as a specific price change that can cause the algorithms to make specific transactions or other factors such as volume, time or advanced algorithms that trade based on trading indicators. Automation of decision-making is becoming increasingly common today. Algorithmic trading represents a trading technique where complex computer systems based on advanced mathematical formulas are used to generate profits. High-frequency trading (HFT) involves the use of a computer program that, using artificial intelligence or following a given instruction, places orders on trading systems. The speed of placing orders and concluding transactions is much faster than in the case of humans. In addition to creating opportunities for profits for companies using HFT, there is the potential to increase the liquidity of markets by increasing the systematicity of transactions from which human emotions are eliminated. By far the most popular fans of algorithmic trading are larger financial institutions, as well as investment banks, along with hedge funds , pension funds, brokers and market makers. HFT is used in many forms of trading and investment activities. For example, investors such as investment funds, pension funds or insurance companies use HFT to buy significant volumes of shares when they do not want to influence market prices. Companies making medium-term investments or selling assets (market makers, speculators, arbitrageurs) use HFT to create the necessary liquidity in the market. Companies taking advantage of market trends, hedge funds or looking for opportunities in the correlation of prices of different assets or arbitrage, as well as high-frequency traders use HFT to conclude transactions. Algorithmic trading is a solution for companies trading on the market in Europe, as it provides not only the necessary speed of concluding transactions, but also accuracy and immediate risk analysis, while eliminating the limitations associated with concluding transactions by humans: emotions, basing on intuition, the possibility of errors during decision-making and entering them into the trading systems of brokers and exchanges.
How Algorithmic Trading Works, Its Advantages and Disadvantages
Algorithmic trading is based on the extensive use of computer programs and artificial intelligence to systematically submit orders to trading systems. The basic benefits of algorithmic trading include: concluding transactions at the best price at a given moment; precision of placed orders at the right time; settlement of transactions at the right time and immediately, in order to avoid significant price changes; reduction of transaction costs; simultaneous analysis of the situation in many markets and identification of trading benefits; reducing operational risk in the order placement process, as well as eliminating emotions and instinct from trading. Trading strategies recorded in the algorithm can be relatively simple, based on price differences and volumes for transactions, as well as complex, using artificial intelligence.
However, they always contain 5 constant elements:
– analysis of electronic data (asset prices, exchange rates, information received in real time);
– transfer of data to the database;
– analysis of strategy results for current data and generation of trading decisions;
– implementation of the trading strategy based on the analysis results and placing orders on selected markets.
During such a process, computers collect market data, evaluate it according to the given criteria, place orders and continue to evaluate markets for new opportunities without human intervention. This enables faster trading, because computers analyze opportunities and risks faster than humans and can execute transactions with a speed and accuracy that humans cannot match. Computers also do not feel emotions related to making decisions about transactions, so trading is based on facts, not emotions. Algorithmic trading also has its drawbacks – one of the highly debated issues related to algorithmic trading is the constant monitoring of strategies, having in mind control features included in algorithmic trading strategies and software (automated and manual). Furthermore, for most individual traders, having sufficient resources can be another disadvantage of algorithmic trading. Automated trading reduces the cost of executing large orders, but can be expensive, as it requires initial infrastructure, such as software cost or server cost.
Legal regulations regarding algorithmic trading
The most important legal acts regulating this issue were issued at the European Union level. These are: Directive 2014/65/EU of the European Parliament and of the Council of 15 May 2014 on markets in financial instruments and amending Directive 2002/92/EC and Directive 2011/61/EU (recast); Commission Delegated Regulation (EU) 2017/589 of 19 July 2016 supplementing Directive 2014/65/EU of the European Parliament and of the Council with regard to regulatory technical standards specifying organisational requirements for investment firms engaging in algorithmic trading (OJ EU L 104, 2017, No. 87, p. 417); Commission Delegated Regulation (EU) 2017/584 of 14 July 2016 supplementing Directive 2014/65/EU of the European Parliament and of the Council with regard to regulatory technical standards specifying organisational requirements for trading venues (OJ EU L 87, 2017, p. 350) and Commission Delegated Regulation (EU) 2017/565 of 25 April 2016 supplementing Directive 2014/65/EU of the European Parliament and of the Council with regard to organisational requirements and operating conditions for investment firms and defined terms for the purposes of that Directive.
According to art. 4 par. 1 point 39 of Directive 2014/65/EU, algorithmic trading is trading in financial instruments in which a computer algorithm automatically determines individual order parameters, such as:
– the conditions for triggering the order,
– the time of its execution,
– the price or quantity of instruments that are the subject of the order or the method of managing the order after its submission,
with limited or no human intervention and does not include any systems used solely for the purpose of routing orders from one trading system to another, for the purpose of processing orders not involving the determination of any transaction parameters, the confirmation of orders or the post-trade processing of concluded transactions.
In turn, Regulations 2017/589 and 2017/584 impose on investment firms the obligation to meet the following organisational and technical criteria related to the use of algorithmic trading techniques:
– monitoring their trading systems and trading algorithms and procedures for resolving issues detected during the monitoring of trading algorithms;
– ensuring appropriate training of employees in the operation of algorithmic trading techniques responsible for the entity’s compliance with the regulations and providing employees with constant contact with the person or persons within the investment firm who have access to the so-called emergency function referred to in Article 12 of Regulation 2017/589;
– establishing, before the implementation or material update of the investment firm’s algorithmic trading system, trading algorithm or algorithmic trading strategy, a clearly formulated methodology for the development and testing of such systems, algorithms or strategies;
– appointing a person approving the implementation or material update of the algorithmic trading system, trading algorithm or algorithmic trading strategy;
– conducting its annual assessment referred to in Article 9 of Regulation 2017/589;
– establishing procedures to ensure that any changes to the functioning of its systems are communicated to the brokers responsible for the trading algorithm and to the compliance and risk management functions; providing, as an emergency measure, the ability to immediately cancel any or all of its unexecuted orders submitted to any or all of the trading venues to which the investment firm is connected;
– monitoring all trading activity taking place via its trading systems;
– carrying out checks on any indications of suspicious trading activity that its automated surveillance system generates during the investigation phase in relation to other material trading activity carried out by the firm;
– implementing business continuity arrangements for its algorithmic trading systems that are appropriate to the nature, scale and complexity of its business; applying pre-trade controls when entering orders;
– monitoring in real time, during the hours in which it sends orders to trading venues, any algorithmic trading activity taking place within its trading code;
– applying post-trade controls on a continuous basis;
– implementation of an IT system strategy with specific objectives and measures that is consistent with the business strategy and risk management strategy of this investment firm and is adapted to its operational activities and the risks to which it is exposed and is based on a solid IT system organisation covering services, production and development and allows for effective management of the security of the IT system, as well as
– meeting the requirements for certifying that the algorithms used have been tested to avoid the possibility of contributing to or causing disruptions to trading before the trading algorithm or trading strategy was applied or significantly updated and that members explain the measures used in these tests.
Regulatory environment for high-frequency algorithmic trading
The legal definition of high-frequency algorithmic trading can be found in Article 4(1)(40) of Directive 2014/65/EU – it is:
any algorithmic trading technique that is characterized by an infrastructure designed to minimize network and other delays, using at least one of the following solutions for algorithmic order entry: co-location, hosting a closely located third party or high-speed direct electronic access, systematic timing of order triggering, generation, routing or execution, without human intervention for a single transaction or order, and a high intraday number of messages that constitute an order, quote or cancellation.
At the same time, in order to recognize a given trading technique as HFT, in accordance with Article 19 of Commission Delegated Regulation (EU) 2017/565, a trading technique that meets the following criteria is considered to be a high intraday message rate, as referred to above, which includes sending one of the following items on average:
(a) at least 2 messages per second in relation to a single financial instrument traded on a trading venue, or (b) at least 4 messages per second in relation to all financial instruments traded on a trading venue.
It should be emphasised that only messages relating to financial instruments for which there is a liquid market should be included in the calculation. Algorithmic trading and high-frequency algorithmic trading (HFT) techniques introduce a number of additional regulatory conditions not only ex ante – requiring an examination of the legal conditions for conducting HA and HFT trading, but also ex post – regulatory obligations related to transactions concluded under HA and HFT.
The speed and automaticity of concluding transactions entails the need to introduce procedures to be followed by persons introducing and supervising automated trading. It is also worth mentioning that HFT on financial instruments, including commodity derivatives, is reserved for investment firms.
In addition, the investment firm is required to register orders, where the details of each submitted order require a register using the format indicated in Tables 2 and 3 in Annex II to Regulation 2017/589. Such registers are kept for five years from the date of placing the order on a given trading venue.