How Exactly Do Investors Use Artificial Intelligence to Make Decisions?

Big Data, artificial intelligence, machine learning and the list of buzzwords continues. It almost seems like today, investors have a fiduciary duty to prescreen new investments, hires, and incumbent employees- and in some cases they actually do. But how does this concept translate in practice? How is technology really used to help investors navigate the world of business decision making? Here, we take a closer look into how investors and financial professionals use these tools for optimal practices.
By Ariella Serman
March 10, 2020

Financial Services Industry Taking the Lead in AI

In an effort to become the leader in providing superior customer experiences, Citi Private Bank, a subsidiary of the multinational conglomerate CitiGroup announced its newest partnership with Feedzai, a leader in machine learning focused on combating fraud prevention. Feedzai’s technology uses algorithms to detect patterns of discrepancies and changes in clients’ payment activities. This allows Citibank to analyze and detect potential deviations from standard customer payments and therefore can help Citibank to better serve customers by protecting their personal information and accounts from the risk of fraud.

Citi has also deployed machine learning to help financial advisors answer a frequently asked question: What are other investors doing with their client’s money? The machine learning technology allows the bank to map out portfolios of clients around the world anonymously. The inconspicuous nature of this development protects customers’ confidential information while sharing valuable insights into their investing behaviors.

In an interview with CNN Business, Philip Watson, Citi’s Chief Innovation Officer and Director of Citi’s global investment lab commented, “Traditionally that kind of information was sourced from your network. You might have had a few coffees or heard about it over a cocktail. Now, we can share insights that are very valuable.”

Another financial institution that has invested billions of dollars in machine learning and artificial intelligence is JPMorgan Chase. The bank introduced COiN, a contract intelligence platform that uses machine learning to analyze legal documents and extract relevant data points. For example, COiN can extract one hundred and fifty important attributes from 12,000 annual commercial credit agreements in a matter of seconds. This is compared to the almost 360,000 hours per year it takes to extract the equivalent data using manual labor.

JPMorgan Chase is also using artificial intelligence technology to increase employee efficiency within the company. They have been developing cognitive automation which combines robotics and machine learning to mirror human judgment. In 2016, the company tested a virtual assistance technology to help employees with technology service desk inquiries. The technology uses natural language interfaces to automate human behaviors such as perceiving, hypothesizing, and reasoning. This helps employees save time by answering their technical queries accurately and efficiently.

Although some are concerned that these types of technological developments may replace jobs in the future, research says otherwise. According to a McKinsey study, by the year 2030, sixty percent of American jobs will consist of tasks that can be automated by at least thirty percent. However, less than five percent of jobs will be completely automatable.

Algorithms Across the Map

Machine learning is being employed not only in the United States but also abroad in quite creative ways. This year, PanAgora, a Boston-based asset management, and investment firm announced its expansion into the Chinese market through a self-learning algorithm. The algorithm analyzes and interprets Chinese cyber slang used by investors on social media in order to sidestep Chinese government regulations and censorship. The results give PanAgora portfolio managers and investors a valuable understanding of Chinese retail investors’ attitudes, beliefs, and viewpoints. This is becoming increasingly important because it is these investors who dominate the Chinese market.

Startup Success with Machine Learning

Machine learning technologies are also being used to discover promising new startups that investors may not have been aware of otherwise. EQT Ventures, a Stockholm based Venture Capital fund, is using artificial intelligence to locate prospective startups. Named Motherbrain, the system uses historical data including financial information, web ranking, and social network activity to scan millions of companies simultaneously in order to predict which ones will be successful. According to EQT Ventures Tech Partner Andreas Thorstensson, Motherbrain was able to predict the success of Uber and Airbnb. To date, the company has invested in 24 companies, with 30 percent of the decisions driven by Motherbrain.

Similarly, a combined study conducted by researchers at Carnegie Mellon University and Microsoft Researchers used data from TechCrunch, CrunchBase, and company websites to decipher the success of tens of thousands of companies. The report looked at three features: demographic (number of employees, age of company, competitors, products), financial (investments and acquisitions of the company, investors per funding round, venture capital and private equity investments) and managerial (successful companies by founder, founder experience). The study was able to predict true positives with an accuracy of 60-79 percent and false positives with an accuracy of 0-8.3 percent.

Key Takeaway

Artificial intelligence is being used across a plethora of industries and for an abundance of decision making processes. Specifically, the application of AI and machine learning is exponentially growing in the financial services industry both on the customer front end and internally for employee efficiency. For example, Intelligo is an Israeli based startup that has used artificial intelligence and machine learning to automate the process of conducting background checks. It is the first company ever to offer a Software-As-A-Service based model for businesses to conduct due diligence assessments. The company uses a combination of advanced AI and machine learning capabilities with human intellect to provide clients with a scalable, comprehensive, and cost-effective solution to clients’ needs, aggregating over one million data points.

Evidently, AI will be sticking around for the next few years and while you may not like it, you will certainly have to get used to it. So don’t take machines for granted- in the next few years they could be your golden ticket to success.

You may also like:

How Confirmation Bias Affects Investment Decisions

How Confirmation Bias Affects Investment Decisions

Using personal judgment to conduct due diligence may not be the optimal way to make business decisions. Many psychological studies show how our decision-making capabilities are limited due to the biases that are inherent in human nature. While the phrase ‘“go with your gut” may have positive implications, in reality, it could be fueled by “confirmation bias” — with disastrous repercussions for investors and decision-makers.

read more