The Future of AI in Finance

Technology has transformed the way the world works. And that transformation is only getting faster and reaching further. Today, a mobile phone has more computing power than the entirety of NASA in 1969 – the year it placed two astronauts on the moon [1]. The power and accessibility of technology continues to grow, and artificial intelligence is increasingly being hailed as the next frontier in this age of digitalisation.

What Is AI?

Artificial intelligence (AI) refers to the ability of a machine to perform intelligent behaviour and cognitive functions that are typically associated with human beings. Common uses of AI can be found in autonomous vehicles, image recognition, sentiment analysis, review management and smart software solutions for content creation platforms..

Many investment firms are researching the potential of AI, and how it could be used to improve financial services and operations. BlackRock, the world’s largest asset manager, announced in 2018 that it was setting up an institute dedicated to AI research [2]. Investment banks such as JP Morgan, Goldman Sachs, and UBS have set up similar initiatives. These moves highlight the growing interest in AI among banking and investment firms.

What Can AI Do?

AI has a wide range of potential applications. One particularly promising use of AI is that of machine learning.

In traditional computing, an algorithm provides a specific set of instructions that a computer will follow – a recipe of sorts. The algorithm is written by a coder or programmer. The computer then follows the algorithm to carry out the desired process. And finally, the computer gives an output or result. Ultimately, a traditional computer cannot make decisions on its own; it can only follow the process encoded in its algorithm. Machine learning is different.

Machine learning algorithms are able to learn from their input data. They can learn from previous experiences to improve their own processes, much like how a human might learn from their past.

Machine learning systems are also able to detect patterns and gradually improve in decision-making or problem-solving. They not only describe what has happened in a given data set but can use previous experiences to predict what will happen in the future and make suitable recommendations.

Source: McKinsey Analytics [3]

Figure 1: An illustration of how machine learning can use data.

How Might AI Be Used in Finance?

Already, AI is being used by firms across both the buy-side (which generally refers to firms that purchase securities) and the sell-side (which typically refers to firms that issue, sell, or trade securities) [4]. However, like most sectors, the investment management industry is still analysing the best ways to implement AI. Here, we shall consider the key ways that AI may be used.

Using AI to Provide Investment Decisions

The role of an asset manager is to make investments on behalf of a client that is suited to the risk appetite and investment goals of the client. AI could process data containing information about the client’s desires and investment criteria. Using machine learning algorithms, it could then provide recommendations on what to do in order to achieve the client’s goals.

However, caution is necessary. Any technology that stores private data requires strict and failsafe cybersecurity. AI systems that utilise client data to provide tailored suggestions will be no different. Additionally, the provision of investment advice is heavily regulated. Investment decisions suggested by AI will require a similar amount of regulation and there can be no ambiguity about who is at fault if investment advice leads to unfavourable results.

Using AI to Generate Investment Returns

Alpha is a term used within investing to describe how a given investment strategy or asset has performed in comparison to a particular benchmark. Often, investors or asset managers try to generate alpha on their investments; meaning they try to generate investment returns that outperform the market.

However, deciding which investments should be made to generate alpha is where the challenge lies. No one can predict with certainty what will happen in the markets. At best, analysts at investment firms and banks will assess market data in an attempt to identify trends or potential opportunities for growth.

However, nowadays there is a growing abundance of financial data available to investors. Publicly traded companies produce quarterly and annual reports. There are many different market indices, each relating to a particular sector and geography. There are also various economic indicators such as reports on inflation, GDP growth, and unemployment rates. News events and geopolitical tensions can also influence prospects for the future. Hence, investors must analyse a broad range of information and factors in order to predict the future performance of a given security or asset.

AI systems could be used to compile, process, and analyse these vast pools of data. They could potentially identify trends in data that were initially hidden from the eyes of human analysts. For example, BlackRock currently uses an AI engine, Aladdin, to process large amounts of financial information. The system can be used to provide an indication of the market sentiment (see Fig. 2).

Furthermore, machine learning algorithms could use financial data to make predictions about future trends. Hence, investment management firms utilising AI could have improved insight on future trends and therefore be in a better position to make investment decisions that can generate alpha.

Source: BlackRock [5]

Figure 2: A visualisation of how AI can be used to analyse financial reports.

What Is the Outlook on AI Companies?

Research conducted by McKinsey Global Institute predicts 70% of companies will adopt at least one form of AI by 2030 [6]. According to a PwC report, AI could contribute up to $15.7 trillion to the global economy by 2030 [7].

As shown in Fig. 3, the number of M&A deals involving AI companies has increased year on year since 2014. There has also been a steady increase in venture capital investments in AI start-ups [8].

Source: Hampleton Partners [9]

Figure 3: Data showing the number of M&A transactions in AI and the top buyers.

Venture capital firms invest at an early stage in companies that are believed to have long-term potential growth [10]. Therefore, increased venture capital investments indicate that the future market for AI acquisitions will likely remain strong as these growing companies mature.

If M&A in AI companies does continue to grow, it will be the investment banks with expertise in technology who are best placed to advise clients on related deals. Similarly, investment management firms should continually monitor which companies and sectors are using AI. This will allow them to make investment decisions that can reap rewards if there is future growth in AI technology.

The general belief is that AI will change the financial sector in one way or another. However, only time will tell exactly how much change it causes and the ways in which it does so.

References

[1] Kaku, M. (2012), Physics of the Future: The Inventions That Will Transform Our Lives, Penguin Books, London

[2] https://www.ft.com/content/4f5720ce-1552-11e8-9376-4a6390addb44

[3] https://www.mckinsey.com/business-functions/mckinsey-analytics/our-insights/an-executives-guide-to-ai

[4] https://institutional.fidelity.com/app/item/RD_9889536/ai-to-enhance-investment-decision-making.html

[5] https://www.blackrock.com/corporate/literature/whitepaper/viewpoint-artificial-intelligence-machine-learning-asset-management-october-2019.pdf

[6] https://www.schroders.com/en/insights/economics/how-can-we-prepare-for-the-ai-revolution

[7] https://www.pwc.com/us/en/services/consulting/library/artificial-intelligence-predictions-2019.html

[8] https://www.pwc.com/us/en/services/deals/ai-deals.html

[9] https://www.hampletonpartners.com/reports/artificial-intelligence-report

[10] https://www.investopedia.com/terms/v/venturecapital.asp