Artificial intelligence captures Wall Street: how it will affect the field of finance and not only

Image: R ~ P ~ M , CC BY 2.0

Artificial intelligence can become the dominant tool for developing financial strategies that were previously considered difficult to predict, because traders and hedge fund managers cannot compete with robots who are able to process huge amounts of data and constantly improve their forecasts by making investment decisions.
')
In the near future, most of the jobs in the financial markets will be taken by robots, and this is good news, because the best university graduates can now leave the industry with more tangible benefits for the population and the planet - technological start-ups, energy and medicine.

Artificial intelligence and record return on investment

Most global exchanges use decision-making computers based on algorithms and corrective strategies based on new data, but some industries, such as bond markets, automate more slowly.

In March, a research team from the University of Erlangen-Nuremberg in Germany developed a number of algorithms that used historical data from markets to replicate real-time investments.

One of the models allowed to achieve 73% return on investment annually from 1992 to 2015, taking into account transaction costs. This compares with a real market return of 9% per year. Profit was particularly high during the market shocks of 2000 (545% yield) and 2008 (681% yield), which proved the increased efficiency of quantitative algorithms during periods of high volatility, when emotions prevail in the markets.

Research by scientists at the University of Erlangen-Nuremberg showed that in their models, the profit from investment of AI decreased after 2001, as the use of robots in trading on the stock exchange became more noticeable and the number of opportunities to use market inefficiency has decreased. However, in recent years, profitability has fallen, and from time to time it has even become negative, which researchers associate with the growing influence of AI on exchange trading.

The idea of using computers for stock trading is not new. Its counterpart - algorithmic trading or black boxes - has been used for more than ten years and is steadily gaining popularity. In 2012, algorithmic trading occupied 85% of the market.

If this trend continues, 90% of trade will be conducted through computer programs. Algorithmic trading today is moving toward high-frequency HFT trading, in which stocks are bought and sold in a split second. The algorithm quickly detects and uses the discrepancy, the profit becomes less and less, but the trading volume is not reduced.

Eurekahedge's January study of 23 hedge funds using artificial intelligence showed that they show much better results than those managed by people.

Over the past six years, these funds have achieved an annual yield of 8.44% compared with conventional funds, which ranged from 1.62% to 2.62%. The authors of the study associate the dominance of artificial intelligence in the industry with the fact that he constantly conducts re-testing, and not just accumulates data. This may also be due to the shortcomings of traditional quantum approaches and the use of trading models built using unprofitable backtests on historical data that are not capable of generating profit in real time.

Artificial Intelligence endlessly processes vast amounts of data, including books, tweets, news, financial indicators, and even entertainment television programs. So he learns to understand global trends and is constantly improving his predictions about financial markets.

Hedge funds have long been hired by mathematicians who develop statistical models and use historical data to create trading algorithms that anticipate market opportunities, but artificial intelligence does it faster and is constantly being improved.

That is why financial giants such as Goldman Sachs, who launched Kensho's artificial intelligence trading platform in 2014, are switching to robotic systems that predict market trends and sell significantly better than people.

Why artificial intelligence will soon force out people from the exchange

Earn more than the average in the stock market is almost impossible - even the most talented investors on Wall Street are not consistent. Traders and hedge fund managers cannot compete, but their problem is that they are just humans, while all the decisions that robots make are based only on data and statistics.

“People always remain biased and emotional, regardless of whether they are aware of it or not,” said Babak Hojat, co-founder of financial startup Sentient and one of the developers of Siri at Apple, in an interview with Bloomberg. - Everyone knows that people make mistakes. In my opinion, it is much more terrible to rely on guesswork and intuition, and not on data and statistics. ”

Systems like the one developed by Sentient can analyze huge amounts of information, including market data, trading volumes, price fluctuations, SEC Internet applications for all companies, social networks, news and videos on YouTube. The goal is to ensure that the algorithm compiles an optimal investment portfolio based on existing knowledge and regularly optimizes it, based on the expected new data for each month.

The number of such projects in recent years has increased significantly. According to some estimates, in the financial sphere, the number of companies working with artificial intelligence reaches 1,500.

For example, the Medallion fund at Renaissance Technologies, which uses quantitative methods for analyzing the stock market, boasts some of the best indicators in investment history. For 20 years, the fund was able to return + 35% in annual terms. This means that if you invested $ 10 thousand in 1997, today you would have $ 4.04 million in your hands.

Bridgewater Associates hired a team to build an autonomous AI system under the leadership of David Ferrucha, who in the past developed the Watson computer for IBM and won the intellectual game show Jeopardy.

Aidyia Limited, an asset manager in Hong Kong, launched a hedge fund that is fully managed by artificial intelligence. It can read news in several languages, analyze economic data, identify questionable patterns, predict market trends and then invest.

Some companies use artificial intelligence to ensure profitability through algorithmic trading. Fund Sentinent Technologies, in just a few minutes can simulate 1800 trading days, pushing trillions of virtual traders among themselves.

Many promising hedge funds around the world have long been using machine learning for algorithmic trading, because it eliminates any manifestation of irrational feelings, such as fear and greed. Investors want artificial intelligence to tell them how to make money in the stock market.

How it works in practice: 9 AI-companies in the field of investment

Numerai holds competitions for the makers of trading strategies. Elements of the best strategies are then used by the fund in real trading on the exchange, and their creators are rewarded.

Qplum uses machine learning to create a robot consultant who uses artificial intelligence algorithms to make investment decisions.

The Russian application Cindicator divides the profits from transactions on the exchange between the “forcasters”, who made a forecast on the prices of a particular stock. Analyzing user responses, the system uses machine learning algorithms. For the most accurate forecasts, users receive the highest payouts. The forecasts of recently registered users are not taken into account until a certain time; later, the program sorts them according to the accuracy of the forecast.

According to the statements of the creators of the project Cindicator to journalists, in the winter the project gathered an investment portfolio with a yield of 47% per annum, analyzing the forecasts of 963 participants.

Sentinent also launched several applications on its artificial intelligence platform. The development of one of them, associated with algorithmic sales, managed to attract $ 135 million. The Fund created several trillions of robot traders, later merged them and is going to allocate this project into a separate company.

Alpaca , a company founded in 2013, raised $ 1 million to develop Capitalico's trading platform, which allows building exchange algorithms based on technical analysis predicting stock price fluctuations. The platform recognizes user patterns as “optimistic” and “pessimistic” and builds a trading strategy based on this.

French startup Walnut Algorithms raised $ 446 thousand to combine machine learning with financial expertise and achieve an absolute return on investment.

Binatix works with hedge funds that use proprietary technology for investment strategies.

Aidyia , a Hong Kong-based hedge fund that uses “general artificial intelligence” that more accurately simulates the human brain, launched a long / short investment fund in 2015 that sells US stocks and commits all stock exchange transactions without human intervention.

The Canadian company BUZZ Indexes collects big data from social networks, interprets this data using artificial intelligence, and then determines which shares the yield will increase. Based on this, an index of interest in social media is built and 75 most popular actions are determined.

The system analyzes not only Twitter and Facebook, but also more than 50 thematic blogs, sites and news services. After that, Buzz clears the data and interprets the mood of the users by analogy with Aspectiva, which marks all product reviews on the network as positive or negative.

If cars win, better techies can do better

“The widespread introduction of artificial intelligence in the financial industry will lead to the fact that traders with huge salaries will lose jobs because of robots as quickly as factory workers. - says Mark Minevich, the founder of Going Global Ventures and a senior fellow at the American Competitiveness Board of the United States. “The influence of artificial intelligence in the industry is gradually increasing, but very soon it will completely change it.”

According to Coalition Development's research, today the average salary of employees in the 12 largest investment banks comes to $ 500 thousand a year, and many traders have incomes of a few million. For example, in 2015, at least five top hedge fund managers earned $ 1 billion. The motivation to refuse employees who earn $ 500 per hour and replace them with robots is understandable.

In 2000, Goldman Sachs had 600 traders who bought and sold shares on the orders of the bank’s major customers, today there are only two such employees left , and the rest of the work is done by robots. How soon the same will happen to all other financial companies, a matter of time.

Such a trend is likely to transform the industry, because the best university graduates will lose interest in Wall Street and will prefer work in medicine, energy, manufacturing and other socially beneficial areas. Today, about a third of graduates of the top ten US business schools go to work in finance, only 5% go to medicine and even fewer to all other industries.

If MBA graduates leave Wall Street but remain in New York, it will help him to compete with San Francisco and not suffer from a shortage of specialists, which is especially important now that the technology industry may lose its influx of engineers due to visa restrictions. President of the United States.

“All these smart people could be hired by technology start-ups, including platforms that develop artificial intelligence,” adds Minevich.

Indeed, when hedge funds lose interest in scientists and engineers, they can join technology startups to develop AI platforms, design unmanned vehicles, develop energy technologies, simulate climate change, catch terrorists and look for a cure for cancer, that is, do things useful to the masses.

Other materials on finance and stock market from ITinvest :

Source: https://habr.com/ru/post/330884/

All Articles