Algorithmic trading

Automated or algorithmic trading (including algorithmic trading, algo trading, black box, high frequency trading, flash trading, or Grey Box Trading) colloquially referred to in general the automatic trading of securities by computer programs.

In high-frequency trading Act of May 7, 2013 - and thus presumably also from November 2013 Securities Trading Act (WpHG ), § § 33, paragraph 1a - algorithmic trading is described as trading in financial instruments where a computer algorithm on the execution and the parameters of the contract automatically decides. Excluded are systems which only confirm the orders or forward it to other trading venues.

To date, no clear definition in the literature has prevailed. Most authors understand by computer programs that are used to derive existing purchase and sales orders ( orders) by electronic means to the stock exchange. The other group of authors sees including computer programs that automatically make buy and sell decisions. In this context, one can distinguish algorithmic trading buy-side and sell-side financial institutions.

History

For the development of automated trading: stock exchanges report a proportion of up to 50 percent of sales. The automated trading from 2004 to 2006 has quadrupled at Eurex. Traditional trade is, however, increased only slightly. The EUREX assumes that currently occurs about 20-30 % of all sales through automated trading. Within the EUREX one expects a growth rate of about 20 % per year. According to a study by the Aite Group about one-third of all securities trading by automated computer programs and algorithms were controlled in 2006. AITE estimates that this figure could be reached by 2010 about 50%. How Gomolka represents these figures have to evaluate critically the market turnover. Because the exchanges see only those orders which are transmitted from machines on the stock market and collected in the electronic order books ( see transaction support). What proportion of the trading volume of machines is generated (see decision support) and what percentage is entered by human traders in the order systems that can not be measured by the stock exchanges.

Early July 2009, a former employee of the American financial services company Goldman Sachs was arrested by the FBI because he is said to have stolen parts of the software that is used by the company for automated trading. The software is also suitable according to prosecutors, " in order to manipulate markets in unfair ways ." He has since been acquitted.

Algorithmic Trading for order entry

Depending on the degree of automation, the computer can autonomously decide on certain aspects of the order (timing, price, volume, or time of order placement ). In the so-called " sell side algo trading " (eg brokerages ) that large orders are split into several smaller trade. This allows market impact, opportunity costs and manage risks. The algorithm determines the splitting and timing (timing) of the orders based on predefined parameters. These parameters usually use both historical and current market data. Algorithmic Trading is used by brokers on the one hand for proprietary trading, on the other hand also offered to the customers of the broker as a service (Due to the complexity and resource situation, institutional investors have a certain urge to access solutions from brokers ). The advantage of automated trading is the high speed at which they can place business, and higher compared to humans amount of relevant information that they observe and process. This is also associated with lower transaction costs. Is a prerequisite for algorithmic trading is that there is already a folder or a trading strategy. This is in contrast to automated trading or quote machines aim is to distribute an order intelligently to different markets. It's not about to shoot according to Parameters automatically deals in the market.

Automated trading as a decision support

Automated trading is used by hedge funds, pension funds, investment funds, banks and other institutional investors to generate orders automatically and / or perform. Here generate computer autonomously buy and sell signals are converted into orders in the financial center before people can engage at all. Algorithmic trading can be used with any investment strategy: market making, inter -market spreading, arbitrage, trend following or speculation. The concrete application of computer models in the investment decision-making and implementation is different. So computers can either ( quant funds) or the Orders only be used in support to the investment analysis, both generated automatically as be forwarded to the financial centers (autopilot ). The difficulty in algorithmic trading lies in the aggregation and analysis of historical market data as well as the aggregation of real-time courses to enable trading. Likewise, the installation and testing of mathematical models is not trivial.

Demarcation High Frequency Trading and Systematic Trading

In the literature, algorithmic trading is often equated with high frequency trading, and are ge - re-sold in the securities in a split second. One study of Final Terna tive According to categorize Fund Manager the area of ​​algorithmic trading but highly variable. So about 60 % of the respondents understand by high-frequency trading transactions in the period from 1 second to 10 minutes. Approx. 15% of respondents understand by transactions in the period of 1-5 days. Aldridge (2009) categorized algorithmic trading exclusively as high-frequency trading. Gomolka (2011 ), however, acting under the algorithmic trading, both the high- frequency trading ( in fractions of a second ) that means the Systematic Trading ( long-term over several days) together. He points out that computer programs are used not only in the short term (eg for flash trading), but also several minutes, hours or days and can act autonomously in the long term in the sequence.

Impact on financial stability

In contrast to the computer market, only serve as a communication platform for the combination of matching buy and sell offers in the computer, the system automatically placed such offers and is looking for sales partners. They are partly responsible for the stock market crash on October 19, 1987, the Black Monday. Your " if-then " algorithms should have ensured that more and more blocks of shares sold off after the rates had begun to fall, which have ultimately led to panic selling. On 6 May 2010, the Dow Jones fell within eight minutes by over 1000 points. However, this crash was triggered originally not by High Frequency programs, but a sell order of the trading house Waddell & Reed, which gave 75,000 S & P500 E- mini futures contracts within 20 minutes by market order in the market. This flash crash prompted the SEC to an aggravation of their Circuit -breaker rules, which should result in price declines of over 10% in a stock to an automatic suspension of trading in the future.