Not so long ago, it was feared that progress in AI would lead to wars and catastrophes, like in the movies. Today the picture is completely different. We’re living in a world where there’s an increasing interaction and collaboration between the human sensibility and machine intelligence. In fact, this relationship is critical for the good operation of businesses and, in turn, for the future of the economy.
AI in Financial Investments
We knew that machines had the capacity to process big data and help us solve critical problems for the future of our species and our planet. It was just a matter of time for AI to enter a field so important for the functioning of our societies like financial investments.
“In general terms, AI is a driver for sectors that have the potential to grow, generating a virtuous cycle of production and generation of income,” in words of the AI team at redbee.
With promising progress and pending challenges, it’s worth exploring the power of AI in financial investment decision-making.
Robo-Advisors: AI, Investments, Fintech and Promises of Democratization
Today we’re seeing how the influence of technology in the financial services sector is creating an unprecedented situation. Institutions as traditional as banks are being challenged by fintech start-ups. Fintech has grown so much that it helped revitalize the post-pandemic economy and labor market, and it started to become influential in services that were traditionally offered by banks only.
On the one hand, this situation forces banks to increase efforts in technological innovation, and on the other, it makes start-ups grow professionally and improve their business continuously. Although fintech companies provide a wide array of services, it’s in retail investments that AI and machine learning have made huge progress.
Historically, for anyone with some extra money it was extremely difficult to choose among the many investment options, so they resorted to banks or brokers for advice. But these services have high fees and commissions, making it accessible only to wealthier people. This is where technology offers a great advantage: it democratizes access to the financial market.
By looking at some figures we’ll get an idea of the potential. While it’s calculated that in Latin America only 2% of the population invests in shares, bonds and other variable income instruments, this numbers increases to 50% in the US. It’s a serious problem because the relative monetary instability in Latin American countries, which includes regular processes of depreciation or inflation, negatively impacts medium and small savers especially, who don’t have the resources to make their money grow.
The automated investments market is just emerging in Latin America, with innovative companies like Fintual in Chile or GBM in Mexico. In developed countries there are several businesses, like Betterment in California or Wealthfront in Canada, which already offer the service of robo-advisor. Basically, this is an automatic investment tool that creates diverse investment portfolios based on a simple questionnaire that determines the client’s goals, financial situation and risk tolerance.
Such improvements promise to open the financial market to more and more people, creating a virtuous circle by which savers can make their money generate earnings and protect themselves from the impact of inflation, simultaneously helping companies grow. The Biggest Challenge of Financial AI: Trust
The Biggest Challenge of Financial AI: Trust
In fintech as in any other sector, the technologies developed to improve investment portfolios earnings face a critical issue: how to build trust.
When an investment portfolio fails to yield results, investors lose money and sometimes they can’t earn it back. So why should we trust an AI over a human being to make or suggest decisions?
The essence of machine learning models is to find patterns within huge quantities of data. It would seem like the perfect match for something as complex as the financial market. AI, with its extraordinary capacity to process information, can analyze not just the market’s current information and constant value fluctuations, but also lessons and patterns that can be found in its historical records. Despite the progress made so far and the promising results, investment strategies based on machine learning, data science and AI are still far from popular. Why is that?
Let’s review some of the main obstacles that this change of paradigm has for growing in the immediate future.
- Reliability. When we analyze many published successful experiments, we find that they’re deceitful. They used simulations to show hundreds of different portfolio yields predicted by algorythms, but investigators published only those with the best results. This may prove useful in the context of an experiment, but it can’t be applied to the real world, where the result is only one.
“We need to obtain information from each market and each factor and ponder everything in real time to be able to say that we have the information needed to make predictions. Also, to be able to develop a model, the information has to be historical,” explains the AI team at redbee.
- Transparency. A person whom we trust money with can explain in a clear and understandable way which are their investment strategies, but algorithms are considered to be “black boxes”, with methods and decisions that are hard to understand by the common person. Naturalmente, esto no despierta la confianza de los inversores.mientras que un ser humano al que le confiamos nuestro dinero o el de nuestra empresa puede explicarnos de manera clara y comprensible cuáles son sus estrategias de inversión, los algoritmos son considerados todavía como “cajas negras”, con métodos y decisiones que resultan difíciles o directamente imposibles de entender para el público promedio. Naturally, this doesn’t generate confidence for investors.
- Regulations. So far, models have shown a relative investment efficacy, but the issue of compliance hasn’t been dealt with in deep. This is especially important in a world involving a complex network of institutions and regulations.
- Ethics. It’s also fundamental to address the challenge of creating models that consider the ethical factors related to environmental, social or governance issues when suggesting investment decisions.
All of the above leads us to ask ourselves, are algorithms better than humans when it comes to making an investment?
AI and Human Sensibility in Finance
In general, human specialists do better. However, the growing complexity of the economy and the markets makes it increasingly necessary to depend on AI processing capacity to make decisions.
“One way of building trust is to try and explain in a simple, intuitive form how the technology works, so that people can conceptualize it,” explains redbee’s AI team.
As we’ve seen, technology applied to investments brings multiple opportunities. They can generate a true revolution in the way companies and people relate to money.
“The benefits of this kind of tools combined with human expertise can create a mix to boost profitable investments,” explains redbee’s AI team.
In the near future, it will be vital to have tools based on AI to make financial decisions in any business, for the finance world in general. Nevertheless, in order to be able to deliver on the promises, it’s also essential to overcome the existing challenges.
What do you think? Will algorithms be able to address the enormous challenges of the financial market?