Over the last 18 years, profit from the AI investment strategy has been 30% per annum. These data were provided by German analysts who assessed the efficiency of artificial intelligence on the financial market. It proves that AI technology is a quite powerful modern tool for trading.
Overall impact of AI and ML on financial market
German scientists (the School of Business and Economics of Friedrich-Alexander-Universität Erlangen-Nürnberg – FAU) conducted research under the supervision of Dr. Christopher Kraus. They have found out that the automated educative platform shows the best results during financial crises. This is due to the fact that the machine keeps performing logical calculations within the situation when people can be driven by emotions.
Global exchanges apply computers accepting solutions using algorithms and changing strategies as necessary. Robots are able to impartially assess risks and process enormous aggregation of data in order to accept decisions about investment and find liquidity. Besides, they constantly accumulate knowledge and act more and more efficiently every time.
Trading is a highly competitive sector. Algorithmic trading, occupying 85% of the market six years ago, has been used for market transactions in shares for 10 years. In 2017, the Eurekahedge agency analyzed the operation of 23 hedge funds that apply AI and ML. The study showed that profit of such funds in five years exceeded the profitability of funds controlled by people five times.
What will promotion of AI technologies result in?
FAU’s scientists noted that profit from investments in their AI-based algorithms decreased after 2001. According to them, it is caused by the fact that AI solutions are more frequently used on the exchange. Now, they compete with each other rather than with people. Therefore, we have less number of opportunities allowing to take advantage of the market inefficiency.
The amount of people involved in the trading process is also reduced, which affects the price formation. AI can drop volatility for a long time, stabilizing the market.
At the same time, Bernstein’s specialists believe that AI and ML will become increasingly popular, and people will more often address robotic financial assistants. Learning machines will be able to easily and efficiently assist customers in managing capitals and investing profitably, as well as to offer client portfolios.
Despite the activity of certain hedge funds, the wide application of new technologies is gaining momentum pretty slowly. In particular, this is because artificial intelligence requires investments in the development and promotion of new methods.
Application of AI in investment sector
The financial market is fluctuating, so that smart machines also need readjustment and improvement. Companies all over the world are developing new machine learning algorithms and designing AI trading tools.
Hedge fund controlled by AI
Hong Kong-based Aidyia Limited has launched a hedge fund entirely controlled by artificial intelligence. It trades shares, analyzes commercial news, makes market forecasts without human interference.
Buzz, an AI-based system developed in Canada, analyzes Twitter and Facebook Big Data, as well as reviews news, blogs, and specialized forums. Based on obtained information, it predicts the profitability of purchasing certain shares. Moreover, Buzz takes into account the mood and assessments of users.
Robotic consultations for traders
Qplum Startup is creating a robotic financial consultant based on artificial intelligence. It will help people to invest, calculate risks, and reduce trading expenditures.
AI in conventional analysis
Sentinent Technologies, a San Francisco fund specializing in the development and selling of financial tools, has designed several apps based on AI. The company raised $135 million for one of them. The fund focuses on searching for new methods of Big Data processing, which can be applied in conventional analysis.
Russia is also working on new AI tools for trading. Machine learning algorithms are used by the Cindicator system. It is able to allocate profit from deals among forecasters who predict prices of certain shares. The more correct a suggestion is, the higher reward people receive. Cindicator also prepares ratings of analysts.
Artificial intelligence or human being?
Artificial intelligence is evolving and advancing, but it is not smarter than people. Machine brain is definitely faster, however, it is designed for performing specific tasks in narrow areas. Trading is a sector of competition between people, but AI and ML can be successfully applied in their operation. Artificial intelligence can save time and efforts of traders in the following segments:
- data search;
- high-frequency trading.
Mark Minevich, a Senior Fellow of the U.S. Council on Competitiveness, thinks that the common use of artificial intelligence technologies in the financial sector will put many traders out of business.
For instance, the largest US bank – Goldman Sachs – had 600 traders working with its customers 18 years ago. Nowadays, there are only two employees in the bank; others are replaced by machines.
According to Mark Minevich, such changes in the financial market can be considered positive, as people will be able to work for technological startups, including development platforms based on artificial intelligence.
According to a well-known trader Michael Harris, old trading methods will be no longer efficient and share purchases during the drop in prices is short sighted. People engaged in exchange trading should explore new technologies because the future belongs to data analysis and testing of new AI and ML operational methods.
“I think that the transformation will be impossible for the majority of traders. A combination of skills required to understand and apply AI rules exclude 95% of traders getting used to draw lines in the graph and examine moving averages,” a system and discretionary trader Michael Harris comments on the situation.
Surely, there is no magic algorithm that will earn all the money in the world without trader interference. Nevertheless, artificial intelligence and machine learning technologies are quite efficient, but one should know how to use them properly.