What Are the Origins of Algorithmic Trading?
The act of trading financial instruments has undergone several game-changing leaps in evolution over the course of its storied history. Despite the constant changes, trading and investing remain a serious discipline, though most traders would be more comfortable defining active trading as an art form. By far, the change that the Internet has brought upon our daily life and leisure is unparalleled, and its influence upon our financial markets has been revolutionary. Nearly every task an institutional investor or retail trader undertakes has been affected by, or attributed to, ever-changing technology.
The late 1990s marked the end of the physical era of the financial markets. Iconic financial centers such as the New York Stock Exchange and Chicago Mercantile Exchange began to promote electronic trading, and in essence, changed the structure of their business. New order-routing systems based on Internet connectivity and electronic trading platforms were built. The old brick-and-mortar exchanges could now provide traders and investors access to the same financial products, but on a global scale.
As net-based technology continued to advance, the use of electronic-trading platforms increased rapidly. Instant connectivity, greater variety, and falling transaction costs all became available to the average person. Naturally, the ranks of the independent retail trader or investor grew. Volumes soared in nearly every marketplace. For instance, on the largest equities exchange in the world, the NYSE, the average daily volume of shares traded grew from 809 million shares in 1999, to 1.6 billion shares in 2005. The introduction of trading platforms based upon Internet technology had bolstered volume, but more importantly, the speed by which one could execute a trade was increased dramatically.
Greatly increased transaction speeds gave the new electronic exchanges, as well as the existing institutional exchanges, the ability to process greater volumes than ever before. Indirectly, the growing volumes produced markets that were vulnerable to heightened volatility and lightning-fast pricing fluctuations. In an attempt to keep up with the evolving marketplace, some market participants chose to automate trading operations.
As information systems technology grew, it became possible to perform advanced mathematical computations in real time. Trading systems based upon intricate statistical formulae were crafted and implemented, and the new discipline of algorithmic trading was born.
What is Algorithmic Trading?
Algorithmic trading (also referred to as algo-trading, automated trading, or black-box trading) is, in simplest terms, to "automate" trading activities by using computers instead of humans to execute trades. Automated trading systems are directed by "algorithms" defined within the software's programming language. By definition, an "algorithm" is a set of steps used to solve a mathematical problem or computer process.
The term "algorithmic trading" refers to the practice of using computers to place trades automatically according to defined criteria contained within the software's programming code. The implementation of algorithmic trading, within the context of the electronic marketplace, is dependent upon the development of a comprehensive trading system. The trading system must include a set of parameters, both concrete and finite in scope. These parameters are a reflection of the adopted trading methodology, and in algorithmic trading, are based upon mathematical computations of varied complexity.
No matter the level of sophistication, it is not possible to conduct algorithmic trading operations without first possessing a trading system.
Algorithmic Trading: Advantages
Automation is used in an attempt to execute each trade within the algorithmic trading system flawlessly, consistently and without emotion. As once put by legendary futures trader Larry Williams, "trading systems work; systems traders do not." Williams' statement sums up the goal of automation: Use computer technology to remove human error from the execution of the trading system.
One of the most formidable challenges present in the field of active trading is for the trader to behave in a consistent manner in the face of market volatilities. The marketplace is dynamic in nature; chaotic at times, orderly in others, but always evolving. In other words, the market can be a difficult venue for an active trader to behave in a rational, consistent manner. The volatile nature of short-term trading has been examined by academics in numerous studies, with failure rates of short-term traders being in the 70% range.
In an attempt to foster a positive outcome (i.e. profitability) in the face of an ever-changing market, traders employ numerous methodologies to develop trading plans and systems. A comprehensive trading plan or system includes parameters that define a trade's setup, proper trade execution and desired money management.
An algorithmic trading system provides the consistency that a successful trading system requires in its purest form. Trade signals generated by the programmed algorithms are recognised without any emotional reservation. Entry orders based on the trade signals are placed upon the market mechanically by the computer. The trade is then managed automatically as per the tenets outlined in the system. Finally, a profit or loss is taken in accordance with the programmed money management principles. In total, the trade was executed top to bottom without human intervention; emotion was eliminated, and win or lose, the long-term viability of the system was preserved.
Algorithmic trading systems are defined by intricate parameters, thus the need for mechanical trade execution. Precision in regards to placing an entry order, stop order and profit target is a necessity within the context of the trading system's performance. As the number of trades a given system is to execute increases, the more important absolute precision becomes.
If a modest 3% trade execution error rate is entered into the equation, then the result of 30 trades comes into question. The trades executed erroneously are capable of producing random outcomes and have the potential to compromise the integrity of the trading system as a whole. Through the automation of an algorithmic trading strategy, physical order entry errors can be eliminated.
Order Entry: Limiting Client Side Latency
The ability to enter and exit the market quickly and efficiently can be crucial to the success of an individual trade and to the longevity of a trading system. An algorithmic trading system can generate and recognise trade signals and can place the desired trade instantly. From the standpoint of the trader or investor, algorithmic trading systems can serve as a valuable time-saving device. In a marketplace where order execution times are measured and quantified using milliseconds, saved seconds are at a premium.
In the electronic marketplace, the issue of latency is an important one. Latency, as it pertains to electronic trading, refers to execution time. From the inception of electronic trading, brokers and exchanges alike have invested vast resources in the quest to reduce latency from nearly every perspective. In a 2009 survey studying proprietary trading firms focused on the forex market, it was found that nearly 65% of firms utilised automated trading systems that incorporated algorithms, and 89% planned on increasing capital investment on low latency technologies.
Algorithmic Trading: Challenges And Pitfalls
Several large drawbacks can influence and hinder the effectiveness of an algorithmic trading system. Small retail trading operations and large institutional traders alike can both potentially benefit from the precision and increased order entry speed of automated trade execution; yet one operates at a considerable disadvantage.
Computer, Internet, and information systems technology are ever-evolving disciplines with the unflinching desire to move forward. Technology within the scope of the financial marketplace is no different. Large capital expenditures are undertaken constantly by market participants in an attempt to keep up, or in a few cases, to create an edge. Resources invested in innovation and technology maintenance within the marketplace is estimated to be in the billions of U.S. dollars annually. Unfortunately, not all traders are capitalised to the degree that they can stay on the technological lead lap.
Although small retail traders and large institutional traders conduct operations within the same electronic marketplaces, each has a vastly different path to the very same market. Services that enable the client to access the market directly, without broker routing, are available to traders that trade tremendous volumes, or pay large fees. This service is known as direct market access, or DMA.
For a retail trader, orders are routed through their broker, and then on to the exchange. The latency concerning the order's execution is greater than that of the trader utilising a direct market access infrastructure. The prevalence of algorithmic trading systems create this scenario. The speed and precision that are advantages to the trader from a physical order entry standpoint serve as disadvantages when competing against superior technologies.
Asymmetric information is defined as being a situation in which one party to a transaction has information about the transaction that the other party is not privy. The electronic marketplace, specifically the implementation of algorithmic trading systems, provides market participants the ability to act on economic information instantaneously. Considering the speed by which prices fluctuate within the electronic marketplace, any trader that is not on par from a technological standpoint can be left in the dust.
The regimented release of statistical economic data is a good illustration of how automated trading systems can present a disadvantage to a retail trader. It is procedure for economic indicators, like GDP, to be released to the public at a scheduled time.
Traders quickly interpret the information in a number of different ways and place trades in an attempt to capitalise on the subsequent volatility. It stands to reason that a trader who receives the information first has an advantage over those who do not. Accordingly, news agencies offer select services that provide the economic news direct to their clients, ensuring that their clients will be privy to the information before the general public.
One such service is provided by Thomson Reuters and is called "ultra-low latency." The package is priced at US$2,000 per month and guarantees the economic data release be delivered to the client two seconds ahead of the public. Without the ability to act substantially within a two-second window, the gap in time is insignificant. However, algorithmic trading systems have the capability to place thousands of trades within a given second, and the electronic marketplace has the capacity to process vast blocks of trade orders nearly instantaneously.
Getting a "jump" on other traders has been around since the inception of trading itself. The ability to act instantly on information can be attributed solely to the automation of trade execution, and indirectly, by the practice of algorithmic trading.
Programming Errors And System Disruptions
The functionality of an algorithmic trading system relies upon hardware to be operational during the execution of trades. Dedicated computers, servers and Internet connections are required to facilitate proper function of the system. Intermittent outages in electricity and Internet connectivity can compromise a given trade's execution. Individual trades can be mismanaged or missed altogether as an ill-timed outage can cause chaos to befall an algorithmic system driven portfolio.
Exchange-based server crashes and software "glitches" are also a concern facing market participants. The botched IPO launch of Facebook on the Nasdaq exchange in 2012 was an example of an automated programming glitch producing chaotic market conditions. Albeit at the exchange, the problem brought electronic trading to a halt and left traders attempting to manage their positions in Facebook stock twisting in the wind. The Nasdaq exchange was fined US$10 million for the meltdown.
In 2011, Knight Capital experienced a software "glitch" in one of its proprietary trading systems. Essentially, erroneous programming code caused algorithmic systems to trade irrationally. The result was devastating as Knight lost US$440 million in one trading session. From a retail trader's perspective these exchange meltdowns exist beyond control. If an individual trader's system happens to be active during an exchange meltdown or falls victim to a "glitch," then the result could be disastrous.
Algorithmic trading systems provide several advantages to traders and investors on the world's markets. However, the technologies upon which the electronic marketplace is based are susceptible to failures, which lie outside of the control of the individual trader.
The word "trade" is defined as being the act of exchanging something for something else. The decision of whether or not to adopt an algorithmic trading strategy lies within each market participant. If the need to increase order entry speed, precision, and consistency outweighs the risk of operating at a competitive disadvantage or getting caught up in an exchange-based meltdown, then the trader may want to consider making the trade.
Any opinions, news, research, analyses, prices, other information, or links to third-party sites are provided as general market commentary and do not constitute investment advice. FXCM will not accept liability for any loss or damage including, without limitation, to any loss of profit which may arise directly or indirectly from use of or reliance on such information.
Senior Market Specialist
Russell Shor (MSTA, CFTe, MFTA) is a Senior Market Specialist at FXCM. He joined the firm in October 2017 and has an Honours Degree in Economics from the University of South Africa and holds the coveted Certified Financial Technician and Master of Financial Technical Analysis qualifications from the International Federation of Technical Analysts. He is a full member of the Society of Technical Analysts in the United Kingdom and combined with his over 20 years of financial markets experience provides resources of a high standard and quality. Russell analyses the financial markets from both a fundamental and technical view and emphasises prudent risk management and good reward-to-risk ratios when trading.
Retrieved 30 Jan 2016 http://www.nyxdata.com/factbook
Retrieved 05 Feb 2016 https://www.merriam-webster.com/dictionary/algorithm
Retrieved 01 Feb 2016 https://www.cbsnews.com/news/the-seduction-of-day-trading/
Retrieved 02 Feb 2016 https://financial-dictionary.thefreedictionary.com/Asymmetric+information
Retrieved 01 Feb 2016 https://www.businessinsider.com/latency-in-trading-2013-6
Retrieved 02 Feb 2016 https://www.economist.com/schumpeter/2012/08/03/desperate-times