Which algorithmic trading companies use Scala

Automatic financial trading: this is how algorithms turn the financial markets upside down - algorithm ethics

Automatic financial trading: this is how algorithms are turning the financial markets upside down

April 9, 2018 / Martin Ehrenhauser

Human financial traders in blue jackets haggling over prices on the trading floor, screaming loudly and gesticulating wildly, are just pictures from days gone by. Visible witnesses of today's financial trading are modern data centers, laser radio relay systems on skyscraper roofs or high microwave radio towers that are rented by trading companies between London and Frankfurt am Main so that they can transmit trading data within a few milliseconds.

The financial markets are largely automated. As early as 2012, about 85 percent of US stock trading was done by algorithms that as analyzed by American financial market experts. In 2003 it was 15 percent. One can assume that the proportion has increased further over the past six years.

Algorithmic computer traders have therefore largely displaced algorithmic traders from the market and the buy and sell orders are now automatically sent to the global computer exchanges based on variables such as prices, trading volumes or media reports.

The computer traders themselves differ from each other essentially in their trading strategy, trading technique and the speed with which they trade. Those computer dealers who use an infrastructure to minimize the transmission time, for example through the use of laser directional radio or microwave networks, which carry out order management at lightning speed without human intervention, are called high-frequency dealers.

And these are "the dominant component in the market and can influence its performance in almost all areas", as the US financial market regulator SEC (Securities and Exchange Commission) put it. Also in Europe: 43 percent of the traded value, 49 percent of the executed transactions and even 76 percent of the orders on European stock exchanges were according to a Study by the EU securities regulator ESMA from 2014 classified as high frequency trading.

High-frequency trading is a special form of algorithmic computer trading that is characterized by the high speed of order management. Graphics of the algorithmic ethics team

Prices in flash crash mode

Through algorithmic computer trading, all players around the world were networked into a complex overall system. The high-frequency trading in particular led to the fact that the volume of trade and the speed increased extremely within the system. As early as 2014, a computer trader on the Eurex exchange in Frankfurt was able to execute around 3,500 orders within 700 milliseconds. With a trading day of an average of 14.5 hours that would theoretically be 261 million orders a day.

The new characteristics of the financial markets involve a number of potential risks, such as the overloading of trading systems, which lead to disruptions in financial markets. Disruptions that result in multiple chains of effects in a complex system and that are intensified and spread across the globe at lightning speed by high-frequency traders, some of whom trade on more than 200 computer exchanges at the same time.

Even the smallest signals can lead to incorrect interactions. For example the content of a single tweet. In April 2013 the TwitterUS news agency account Associated Press hacked and falsely reported two White House explosions. The Twitter feed This was followed by high-frequency traders who use their algorithms to constantly search through news in order to anticipate price trends at an early stage. The keywords in the short message signaled an imminent price slump to the algorithms and triggered a huge number of automated sales orders within a very short time. An extreme drop in prices, a so-called one Flash crash, was the consequence of that.

These FlashEvents are "omnipresent" today, as is one Study by the European Central Bank (ECB) stated. In times when computer systems operate on an exceptionally low timescale, the feverish ones find Flash-Events now take place in the millisecond range. The stock market averages more than once per trading day, as in one Study from the University of Miami called.

Naturally, the high-frequency traders themselves do not take the position that they are exacerbating extreme falls in prices. In their opinion, they are mitigating volatility. Especially when the price plummets dangerously low, all retailers try to sell their products at the same time and thus reinforce the price trend. In such a phase, they argue, they would set a price and signal to market participants that they can buy and sell at any time. In their opinion, that would calm the market down again.

However, Deutsche Bank does not confirm this assessment. In a Study from 2016 it is said that high-frequency traders “often withdraw in volatile market phases and reduce their liquidity supply”. They create “an increased risk” of “excessive volatility, which can lead to market dislocations Flash-Events could be favored ". For the authors of the study, it is clear that high-frequency traders “intensify excessive price movements in times of greater nervousness in the market”.

Unpredictable prices hurt the economy

Prices are an important indicator of the health of the financial markets. In times of high-frequency trading, in which prices form within milliseconds, however, they are degraded to an expectation without reference to real values. Triggered by news, it is automatically predicted in which direction prices will move. The more serious the news, the more aggressively the traders' positions are adjusted.

If, however, prices, such as a share, are not based on fundamental data, are not related to a company's key economic figures, but merely represent a directional expectation, then this circumstance has negative effects on the real economy. For example, because the price no longer directs production factors such as labor, land or capital to where they are most urgently needed.

Another negative consequence is the production of uncertainty. "Every trade is a journey into the unknown and every economic activity is future-oriented", says the Austrian economist Stephan Schulmeister. Therefore, “every economic activity, even if the time horizon is only five minutes, needs expectations. The "speed of trading is so enormous, however, the overall system is so uncertain that no one has any more time to form a 'true' price expectation, i.e. to estimate the 'fundamental value'."

But if it is no longer possible to form comprehensible expectations for the real economy, then companies will no longer invest in economic activities that require a long planning period, such as measures against climate change. This, in turn, can cause enormous damage to society as a whole in the long term.

And if, despite automation, the trading costs of institutional investors, such as insurance companies that buy financial products with the customer money of the "average consumer", have "increased" over the years, like one of them Study from Fordham University in New York analyzed, then it seems that the diverse possibilities of automation are only proving to be an advantage for algorithmic computer traders - outside of macroeconomic events.

A circumstance that urgently requires regulatory measures. How these could look in a complex system like the automated financial markets and what the European Union has failed to do with the current legislation, I'll deal with that in Part 2. Stay true to me.