Bots for cryptoinvestors. Is it possible to entrust the money to the algorithm?

Deloitte estimates that by 2025, the volume of assets managed by automated systems will range from $ 5 trillion to $ 7 trillion. Given the speed of the introduction of artificial intelligence in the financial sector, this is quite a realistic forecast.

Technologies are increasingly penetrating the most diverse spheres of life: every day we learn that computers without cars take cars, make medical diagnoses, and even create designer clothes. Investors do not stand aside: there are all new tools that help make investment decisions and manage assets, in addition, quantum funds are distributed, in which investment decisions are made automatically.

Artificial intelligence on the exchange

Modern technologies are being introduced into the investment sphere where more actively than it might seem at first glance. Many players in this market talk about their intention to expand the use of artificial intelligence (AI) in their companies.

More than half of asset managers are looking to integrate artificial intelligence into their investment processes and gain experience in this area, according to Dramatic Change: The Future of Investment Research, produced by Greenwich Associates and Thomson Reuters.

The report is based on a survey of 30 investment directors and asset managers. The study confirms that AI penetration in this area will increase in the coming years: 17% of investment companies are already actively using this technology, 10% plan to implement it in the next 12 months, and generally expect the role of AI to increase in the investment process 56% of respondents.

At the same time, the number of hedge funds in which artificial intelligence makes investment decisions is growing. Such funds are usually called quantum, in which investment decisions are made automatically based on mathematical models.

According to research firm Hedge Fund Research (HFR), at the end of last year, assets under management of quantum funds were close to $ 1 trillion - at the end of October they were $ 940 billion. Compared to 2010, this is an increase of 86% . The largest quantum funds include Renaissance, Two Sigma Investments, DE Shaw Group, PDT Partners and TGS.

Interested in AI and the traditional big players. Last year, the Swiss bank Credit Suisse allocated the Qube Research and Technologies quantum fund for $ 1 billion, the bank also manages another quantum fund, the QT Fund.

One of the largest US banks, JPMorgan Chase, is also very active in this area. In May, the bank created a division for the introduction of artificial intelligence technologies. JPMorgan is working on creating a bot-trader who will be automated to engage in stock trading. So far, the bot is only an auxiliary tool, a person makes the final decision on the purchase and sale of shares; nevertheless, by the end of this year, the bot can be fully responsible for the 50% of the stock trading in the Asia-Pacific region.

In early July, it became known that members of the Rothschild family had invested in an AI-based large-scale state management platform, called Exo Investing. The creators of this platform say that it gives access to sophisticated technological tools based on AI to a wide range of investors - with portfolios from £ 10 000.

As in many other areas, the spread of AI in the financial market threatens to reduce jobs. According to forecasts by Opimas consulting company, by 2025, around 90 000 employees in asset management (from 300 000 such specialists around the world), including fund managers, analysts, and support services employees, will be at risk.

Pitfalls

Making investment decisions with the help of AI involves significant risks, including due to the fact that this investment option has not yet been sufficiently tested by time and serious crises.

One of the limitations that increases the risk of automated investment decision-making is that AI is less flexible than a person and can only make a decision based on its data set.

“The car does not have the fundamental knowledge to predict a crisis, as each of them is unique. People are good at talking about things like a crisis, and can sometimes predict them, but we often make mistakes, ”said Vazant Dhar, founder of one of the first quantum funds SCT Capital Management and professor at New York University. Dhar devoted several works to computer analysis of data, among other things, he asks in them the question whether people are ready to entrust their money to a computer.

Many experts agree that the car cannot make forecasts taking into account a wide range of additional factors - from political changes and the geopolitical situation to technological innovations and natural disasters.

“Allowing a computer to choose music in Spotify is one thing, but choosing stocks that affect trillions of dollars in markets is another. In Spotify, we can move on to the next song if we don’t like the one chosen, in the case of investments based on AI, the rates are disproportionately higher, ”writes Zachary Schaeffer, general director of the financial and technological platform of Elsen.

In his opinion, it is still difficult to judge whether the moment will come when investment decisions will be made fully automated, without human intervention. Schaeffer believes that the more likely scenario remains the preservation of human intervention in the future, albeit to a lesser extent than now.

Another possible risk is that in numerous quantum funds investment decisions will be made on the basis of the same data, formulas and algorithms, which in an unfavorable situation can lead to a market crash if, for example, all funds start selling shares at the same time.

All this, of course, does not mean that artificial intelligence in the investment field is just a tribute to fashion. According to the CEO of the British investment firm Man Group, Luke Ellis, if technologies continue to develop as quickly as they are developing now, after 25 years, machine learning systems will participate in 99% of investment management cases. “It will become a ubiquitous phenomenon in our life. I do not think that machine learning is the answer to everything we do. I just think that it makes us better in many ways from what we do, ”says Ellis.

In the service of the investor

Recently, there are separate services for investors, helping retail players to use AI. In fact, most of these services are based on the capabilities of automated analysis of large data arrays and recommendations to investors based on this analysis.

Such systems typically use machine learning technologies (“learning” software to perform certain tasks without clear instructions) and neural networks (systems operating on the principle of networks of nerve cells of a living organism, that is, similar to the human brain), so that these systems are constantly being improved with getting all the new data. Most of these systems analyze not only stock information, but also very wide and diverse data sets, including various articles, press releases of companies and other news resources.

Back in 2013, a startup Kensho was created, whose founders are dedicated to the development of machine learning and artificial intelligence in the financial and investment field, among its platforms Kensho Global Event Database and Knowledge Graph. The system developed by Kensho is based on machine learning, in which artificial intelligence, based on huge amounts of data of various kinds, learns to find the correlation between political and other events in the world and stock prices. A testament to the startup's success, it was bought in March by the world's largest rating agency S&P Global for $ 550 million, one of the largest deals ever involving an AI company.

Another platform, Kavout, works with big data, on the basis of which the system makes conclusions about the prospects of investing in the shares of thousands of companies. We have already mentioned the Exo Investing platform, which also provides access to AI for making investment decisions to a fairly wide range of investors. Another company, Qplum, offers its customers to use the services of a robotic investment consultant, which also uses machine learning technology. EquBot service, which also works on the basis of AI, is aimed at identifying undervalued assets, optimizing the share of certain assets in the investment portfolio and determining the best terms for investing in various stocks and other assets.

Thus, the investment sphere is indeed becoming more and more automated, and this concerns both the systems developed by large companies and the corresponding services available to small companies and individuals. A few years ago, consulting company Deloitte predicted that by 2025, the assets, to some extent controlled by automated systems, would range from $ 5 trillion to $ 7 trillion, which, given the active implementation of such systems, does not seem too unrealistic now .

Rate this article
Blockchain media