The Impact of Robotics and Artificial Intelligence in Finance
1. Introduction
The robots are coming. This allegation does not come from a sci-fi movie, but it is the recent years’ trend. Robotics and Artificial Intelligence are rapidly transforming and revolutionizing many industries in the world, and the global financial industry is not an exception. These technologies have already strongly influenced the financial world as they advance at an accelerated pace, helping to improve efficiency, boost productivity, reduce costs and credit risk, mitigate financial fraud, and giving a competitive advantage for companies that are willing and able to apply them.
Recent years’ developments combined the two technology which gives a powerful mix, called artificially intelligent robotics. This new concept of robotics has huge potential to lift the economy but carries numerous risks in terms of employment.
These technologies already started displacing jobs in finance, but also created new ones. People with the right expertise and skills in AI, data science, or robotics will be the winner of the transformation these technologies bring to the financial industry, and these people already have a competitive edge over job seekers who do not possess the necessary skills.
This research project explores the impacts that robotics, artificial intelligence (AI), and their combination have and potentially will have on the financial services sector (especially in banking) and discussing their possible effect on employment. The study is based on only secondary data and contents related to robotics and artificial intelligence, gathered from public websites, and different books purchased by me.
2. Robotics and AI
2.1. A new wave of old technologies
Robotics and Artificial Intelligence (known as AI) are not new concepts. Already in the 1950s, a young British polymath named Alan Turing posed the question “Can machines think” which became the framework of his paper, Computing Machinery and Intelligence[1]. This report laid down the basics of further development of AI.
The concept and development of robotics date back even much earlier, some sources mentioning the ancient Egyptians as the first adopters of early robotics[2], followed by the Greeks and Romans, and many more after over the centuries. Although those rudimentary machines were not even comparable to robots in present, these developments over history have helped shape the world of today’s robotics.
In recent years, there has been a resurgence of interest in AI and robotics. As the report of the institution named after Alan Turing mentions[3], the improved technology, the faster special-purpose hardware, and the rapid growth of big data have led to a great AI innovation wave and a boosted growth in the global financial services industry. It includes: robo-advising services, chatbots, algorithmic trading, smart contracts, blockchain, improved fraud detection technologies, and decision support systems for loan granting. These concepts are among the most important recent developments in finance but do not cover the whole spectrum of the changes that AI and robotics brought to the industry.
Many institutions and organizations try to estimate the potential impact that AI and robotics can bring to the economy and finance. Looking at the broader picture, a 2017 government report[4] estimated that AI could add an additional $814 billion to the UK economy by 2035.
In terms of global finance, according to McKinsey estimates, AI technologies could deliver up to $1 trillion of additional value each year for global banking[5]. Moreover, Business Insider estimates the aggregate potential cost savings for banks from AI applications at $447 billion by 2023[6]. These gigantic numbers show that financial institutions simply cannot ignore the technology.
“The financial services industry has a long history of using algorithms, data, different quantitative methods to support decision making. These are the foundation of AI, and the industry is therefore primed for AI adoption, positioning it at the forefront of adopting and benefiting from AI technologies. AI will become more and more crucial as the data that we create continues to grow. This will lead to a situation where processes that previously did not require AI (e.g. fraud detection), will no longer be able to succeed without AI.” (Artificial Intelligence in Financial Services, June 2019, 7th page)[7]
2.2. An important clarification
Before discussing the topic, a distinguish needs to be made, because Artificial Intelligence and Robotics are not the same things, and there are many confusions about the terms. AI and Robotics are almost entirely separate fields, with Artificially Intelligent Robots lying in the small intersection of the two areas. Clarifying the distinction is important to see how the two terms fit together.
What is Robotics?
The term of robotics refers to anything involving physical robots. Per definition, robotics is the study of robots, which are machines that are made to do particular jobs. These robots are programmable with the main goal to carry out actions autonomously, or semi-autonomously.
Robots can have many applications, one of the many is to perform tasks that most humans could not possibly do, or would be dangerous to do. Such as working in challenging situations, conditions, and do the required job with utmost precision (like defusing bombs).
Apart from replacing people at dangerous jobs, these robots can perform cheap, low-skill works and do these tasks more efficiently, quickly, and for free, saving labor costs for businesses. For example, receptionists, data entry clerks, cleaners, and many other jobs can be replaced by robots.
What is AI?
Compare to Robots, Artificial Intelligence is a bit more elusive concept, and its definition has many variations. It refers to any human-like intelligence exhibited by a robot, machine, or any computer. It is a collective term for the science of making machines smart, that can sense their environment, learn, think and take actions accordingly. AI basically tries to mimic the capabilities of the human mind: learn from examples and experience, make decisions and solve problems[8].
The magic combination — Artificially Intelligent Robots
The bridge between robotics and AI is created by artificially intelligent robots. This new area is the result of the recent years’ developments that connected these two entirely separate fields, multiplied their potential benefits by simply attached them, establishing a powerful combination.
These robots are armed with AI and controlled by intelligent programs. In combination, the robot is the body (if there is any) and AI is the brain.
Nevertheless, artificially intelligent robots can have different forms, depending on their physical existence:
- AI Cobots
The typical form of AI robots mentioned above, with physical existence. Cobot is short for the collaborative robot which is developed by the addition of artificial intelligence. The added AI gives the robot the ability of perception and decision-making instincts. In contrast to traditional industrial robots, Cobots operate in close collaboration and proximity with humans, reaching together a more precise and efficient work performance. These robots have been around commercially for about a decade now, but because of their advantages, they are seeing broader adoption, and today “they are the fastest-growing segment of industrial automation” (Robotics Tomorrow, 2019[9]).
- Software Robots with AI (known as IPA)
Computer programs with added AI, performing tasks on various software, websites by themselves. These are also known as the combination of Robotic Process Automation (RPA) and AI, called Intelligent Process Automation (IPA).
Some argue that these robots are not robots in real because of their non-physical existence, but the widespread of them (especially in the financial sector) and the work they do falls into the category of robots, therefore they are treated accordingly.
- Software Robots without AI (known as RPA)
The list would not be complete without the mention of non-AI software robots. These robots do not fit into any category mentioned above, because they are neither physical nor artificially intelligent robots. Rather, they are a subcategory of software robots. This important category is referred to as Robotic Process Automation (RPA) and has nothing to do with physical robots in robotics. Often, it is neglectly mixed with IPA and considered as they were the same class under software robotics because of their similarity, but it is a wrong approach, and the distinction between them is important.
Nonetheless, just like artificially intelligent robots, RPA also plays a pivotal role in revolutionizing the financial industry by their capability of automating different business processes. The range of RPA in finance is wide: starting from simple chatbots to spreading to sophisticated accounting or trading softwares they have thousands of applications.
2.3 Advantages of AI Robots
AI robots are advancing at a rapid pace, huge thanks for their potential of “intelligent automation” or “smart automation”. The technology enables companies to automate different, repetitive tasks and processes, freeing up labor force and put that in more added-value jobs.
The improving technology powered by artificial intelligence enables robots to be able to work more faster and efficiently than human workers and do more sophisticated and complex tasks. This can minimize or totally eliminate human errors, save time, and reduce costs at the same time. Moreover, the scalability of the technology means that robots cope with much higher volumes, leading to tasks that can be delivered in record times[10].
According to a PwC research[11], these robots can boost productivity, create new and better products and services. Intelligent Process Automation basically can:
- Automate any business process end to end. As a consequence, it saves time and money by increasing processing speed, reducing human interaction.
- Organize, process, and analyze complex data better than humans, provide valuable insights
- Eliminate errors and exceptions.
- Enhance customer experience by delivering faster response times, greater accuracy, and more consistent results[12].
- Continually learn and improve upon tasks through machine learning, therefore perform them with gradually improving efficiency.
- Run in the background 24*7, helping mechanize back-end processes.
Nevertheless, not every process can be automated in finance. McKinsey Global Institute found that
“currently demonstrated technologies can fully automate 42 percent of finance activities and mostly automate a further 19 percent”. (McKinsey & Co — Bots, algorithms, and the future of the finance function, January 9, 2018[13])
According to McKinsey, transactional activities are the most automatable, but opportunities exist across most subfunctions. On the other hand, business development is basically impossible to automate.
3. Revolution in Finance
Robots are already widely deployed in the financial industry, but due to the nature of the business mostly software robots (IPA and RPA also) are used in the sector, physical AI robots (cobots) are less common.
Software robots have a wide range of applications in the broader financial sector, and they revolutionize the way companies do business through them. The main, but obviously not complete list is as follows:
- Improve lending
Robots powered with AI can help banks and credit lenders to improve their loan approval strategy. As the researcher’s previous project showed, using machine learning techniques can significantly improve the results of a bank’s credit department by decreasing the proportion of non-performing loans and loan loss provisions[14]. According to the project, Random Forest models predicted 92% accuracy if a loan would be repaid or defaulted if granted, compare to 65% of the bank’s previous non-machine learning model.
The spread of new AI-powered banking applications is based on more complex and sophisticated rules in their credit scoring systems, which leads to a better-informed, data-backed decision and a more accurate assessment of the potential borrower[15]. AI can help determine the creditworthiness of potential borrowers with limited or no credit history by analyzing various data of their smartphone (social media using, geolocation, browsing history, payment data, etc) collected by the app.
This carries tremendous opportunities for digital banks, especially in such important markets like Africa or Asia, where a rapidly growing middle class uses smartphones every day but lacks traditional credit scores. According to Deloitte, by 2030, over half a billion Africans are projected to be middle class[16] and a large share of them will be younger people, seeking access to the internet and providing a perfect base for rapid growth for digital banks exploiting the power of AI.
- Simplify operation
The aforementioned advantages of smart automation provide great benefits for players in the financial industry that apply the technology. Mundane, repetitive tasks of banks, like account openings, administrative things can be easily automated with software robots. Routine processes (like credit checks, customer verifications) of lending and mortgage financing can also be automated, resulting a reduced loan-processing time, faster decisions, enhanced customer service, and overall greater efficiency.
According to Juniper Research[17], simple chatbots and virtual assistants can replace the human workforce, by resolving customer queries in a fully automated way. These can help deliver automated help, such as perform account services, assist with financial requests, offer optimized financial plans and tips, dealing with frequently asked questions, and other automatable tasks. Many big banks in the USA and the UK launched mobile banking apps that provide clients expense planners, reminders to pay bills, and give information about the bank’s offerings[18]. As a consequence, the clients can have quicker and personalized service than before.
The Juniper Research found[19], that chatbots and virtual assistants will be saving time for banks in 2023 of 862 million hours or nearly half a million working years. This contributes to a vast amount of operational cost savings. The study says that chatbots in banking will save $7.3 billion globally by 2023, up from an estimated $209 million in 2019.
Software robots can make the financial institutions’ compliance activities more efficient as well. Automating such necessary steps of compliance like anti-money laundering (AML) and know your customer (KYC) can minimize human errors, speed up the process, and makes it easier to comply with regulatory requirements[20], saving a huge amount of money on manual compliance work. According to a report by McKinsey & Company[21], compliance-error rates can be reduced from more than 30 percent to less than 5 percent.
- Combat financial crimes
AI systems can help immediately detect financial fraud and money-laundering by continually scan payments and transaction data for fraudulent patterns, much quicker than human counterparts. Speed is critical in combating financial crimes, and AI software can easily cope with this task by analyzing vast amounts of data quickly and efficiently.
The aforementioned report by McKinsey & Company estimates that for banks up to 90 percent of the alerts generated by simple and traditional linear rules on fraudulent activites are false positives. It is relatively easy for criminals to get to know these linear rules and bypass them. But according to the report,
“machine-learning algorithms can help reduce the number of false reports by 20 to 30 percent. As a result, investigators can spend more time on high-risk cases, and the manual work required can be reduced by as much as 50 percent” (McKinsey & Company: The new frontier in anti-money laundering, 2017[22])
The new AI-driven technology can significantly reduce the costs of fighting financial crimes. As per the Financial Conduct Authority[23], British banks spend £5bn annually combating financial crimes, which is £1bn more than spending on prisons. While reducing this tremendous amount of money can improve banks’ results, fighting off financial criminals advantages the whole economy at the same time.
There are some popular AI-powered platforms that big banks use, like IBM’s Watson financial services, Thomson Reuters and Bloomberg’s solutions, or QuantaVersa, just to mention the most-widely adopted ones.
- Robots in Trading
The emergence of AI and robotics brought drastic changes to the capital markets as well. An increasing number of companies are using AI robots that learn from data, historical patterns, and trade by themselves, trying to predict price movements. As per the Siena-project mentions[24], Algorithmic trading (or high-frequency trading) involves complex AI systems and robots to make extremely fast trading decisions, meaning these robots are able to buy and sell positions in milliseconds and operate completely without human intervention.
Algorithmic trading allows quick profit generation at an incredibly high frequency, which is practically impossible for any human trader. This kind of trading provides the benefits of trades to be executed at the best possible prices, reduced likelihood of mistakes, and eliminate psychological or emotional conditions that normally affect human-executed financial trades[25]. It enables brokerages to split up orders, so the size of their trades does not cause any disruption on the market.
Moreover,
“it improves many measures of market quality and liquidity provision during normal market conditions. Although, it may worsen periods of unusual market stress and volatility.” (U.S. Securities and Exchange Commision — Staff Report on Algorithmic Trading in U.S. Capital Markets, 2020[26]).
Another problem can become the lack of human intervention. There were many examples of market flash crashes (sudden, very rapid price declines) in the last years mainly caused by “unchained” robots, resulting in huge losses for parties. For instance, Knight Capital brokerage company lost $440 million in 45 minutes due to a flash crash caused by trading robots in 2012[27].
HFT already accounts for a large portion of trading in the US market. According to The Alan Turing Institute[28], by 2017, high-frequency trading represented about 50% of trading volume in the US equity markets and 60% of futures trades.
The rise of Algorithmic Trading will continue in the future. According to Valuates Reports,
“The Global Algorithmic Trading Market size is expected to grow from USD 11,846.92 Million in 2019 to USD 22,092.37 Million by the end of 2025 at a Compound Annual Growth Rate (CAGR) of 10.94%.” (Valuates Report: Algorithmic Trading Market Report, 2020 August[29])
4. Employment
4.1. Millions of jobs at risk
The advances of Artificial Intelligence and robotics already have had an impact on the labor market in the financial services industry:
- State Street, an American financial services and bank company cut 1500 jobs (about 6 percent of its workforce) due to ramp-up automation.[30]
- Mainly due to AI and robotics-driven digitalization in banking, bank branch closures became more common, as there is less need for visiting banks in person. Big high-street banks in the UK closed 31–74% of their branches between 2015 and 2019, and this trend can continue.
- Deutsche Bank started to use robots in 2019 to achieve its 18,000 job cut target by 2022. According to the bank’s head of operations, “AI massively increased productivity and redistribute capacity in certain sectors of the business[31].
These are just a couple of examples of the many disruptions that happened in the industry in the last years. There are many concerns that this trend will continue, and these new technologies will wipe out or replace much more jobs in the future.
Many pieces of research try to predict the number of potential job losses caused by automation:
- According to Wells Fargo research, 200,000 jobs will be lost in this decade across the U.S. banking industry[32]. This accounts for roughly 10 percent of the total amount of bank jobs.
- IHS Markit published more enhanced numbers in its 2018 report, estimating job losses or reassignments due to AI and robotics will affect 1.3 million bank workers in the US alone by 2030[33]. Especially financial managers, compliance and loan officers, customer-service reps will be under threat.
- Consultancy firm Opimas estimated that some 400,000 to 1.7 million people could lose their jobs at capital markets institutions in this decade due to the rise of artificial intelligence and automation[34].
Nobody knows for certain the exact numbers of potential job losses, but according to the aforementioned PwC analysis[35], financial services jobs could be the one of the most vulnerable to automation in the shorter term (at least to late 2020s), compared to other sectors,
“as algorithms outperform humans in an ever wider range of tasks involving pure data analysis”. (PwC (2018) — Will robots really steal our jobs? (3rd page))
To the second wave of automation, about 30 percent of financial services jobs will be at potential risk, significantly more than in other sectors. In the longer term, after 2030, the proportion of jobs at potential risk will stay at a constant level, and the transportation sector will take over finance on the top.
However, financial AI-robots do create challenges not just for routine jobs but for the highest- skilled ones too. The use of algorithms does not just replace jobs of traders or customer-service reps, but it also displaced people who build investment portfolios, or model risk and prices.
“This is not necessarily because they are replaced by machines, but because they are not trained to work alongside algorithms“ (Marcos Lopez de Prado, Cornell University professor at the testimony to the U.S House Committee, 2019[36])
This fear is reflected in a survey conducted by the Certified Financial Analyst Institute survey among investment professionals.[37] The surveyed experts expect that AI will become the norm, threatening to wipe out particular professions. The three roles most likely to disappear are sales agents, traders (mainly due to the advance of algorithmic trading), and performance analysts. 43% of the surveyed professionals expect that their roles at investment firms will change significantly, and 5% said their role is unlikely to exist in 5–10 years.
4.2. Data and AI experts wanted
Despite these dark forecasts, some optimists argue that the rise of robots in finance isn’t simply taking away jobs, but creates new ones and provide lucrative opportunities for people with the appropriate and necessary skills.
According to Bloomberg[38], job seekers with expertise in artificial intelligence, machine learning, big data, and data science are among the most in-demand candidates in finance. By today, data scientist become the hottest job function for employers, and big financial companies such as JPMorgan Chase, Morgan Stanley are scooping up people with the required data skill:
“In the U.S. financial sector alone, job postings that list these big data skills as requirements increased almost 60% in the 12 months ending in July 2019, according to LinkedIn” (Bloomberg — Finance Needs People Who Work Well With Robots, 2019[39])
On top of that, Artificial Intelligence Specialist became the number one emerging job in the UK in 2020, coming as a proof of the technology’s potential, according to Linkedin’s emerging jobs report[40].
Besides, as machines make more decisions in finance, it also presents certain risks; ethical and legal concerns will be raised that need to be managed at both operational and board levels. To manage these risks, the need for specialized experts in this field like AI auditors, AI Managers, or AI programmers will grow, creating new job opportunities. For instance, if an AI-based model starts to discriminate against certain populations of people in mortgage or credit decisions, and the bank is being sued for discrimination, people at the bank will be liable for that, not the machine. AI experts with the appropriate knowledge are able to minimize these risks.
Besides the mentioned direct job-creating opportunities, AI software may create jobs in indirect ways also:
“For instance, automating tasks previously done by humans in the asset management industry should theoretically reduce costs. Lower fees will likely increase demand for financial services and, subsequently, the need for more staff to service new customers, according to Guo Bai, a lecturer of strategy at China Europe International Business School in Shanghai” (Bloomberg — Finance Needs People Who Work Well With Robots, 2019[41])
Successful professionals will have to understand and exploit the opportunities brought about by AI, robotics, and its applications[42]. They need to prepare themselves for these future trends and acquire the necessary employable skills in order to overcome challenges created by the new technologies. Teaching students to these new skills and prepare them for the changes will be the education system’s duty in the largest part.
[1]A.M. Turing — Computing Machinery and Intelligence https://www.csee.umbc.edu/courses/471/papers/turing.pdf
[2]Dennis Spaeth — The evolution of Robots
https://www.ctemag.com/news/articles/evolution-of-robots
[3][25][28]The Alan Turing Institute — Artificial Intelligence in finance https://www.turing.ac.uk/sites/default/files/2019-04/artificial_intelligence_in_finance_-_turing_report_0.pdf
[4]Professor Dame Wendy Hall and Jerome Pesenti — Growing the Artificial Intelligence Industry in the UK https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/652097/Growing_the_artificial_intelligence_industry_in_the_UK.pdf
[5]McKinsey & Company — AI-bank of the future: Can banks meet the AI challenge?
[6]Business Insider — Artificial Intelligence in Financial Services: Applications and benefits of AI in finance
https://www.businessinsider.com/ai-in-finance?r=US&IR=T
[7]Microsoft — Artificial Intelligence in Financial Services
https://www.ukfinance.org.uk/system/files/AI-2019_FINAL_ONLINE.pdf
[8]IBM — Artificial Intelligence
https://www.ibm.com/cloud/learn/what-is-artificial-intelligence
[9]Roboticstomorrow — Rise and rule of Cobots
https://www.roboticstomorrow.com/article/2019/11/rise-and-rule-of-cobots/14384
[10][20]International Banker — The Impact of robotic process automation on financial services
[11][35]PwC — Will robots really steal our jobs?
https://www.pwc.co.uk/economic-services/assets/international-impact-of-automation-feb-2018.pdf
[12]Automation Anywhere — Combine the power of RPA and AI to empower rapid end-to-end business process automation
https://www.automationanywhere.com/rpa/intelligent-automation
[13]McKinsey&Company — Bots, algorithms, and the future of the finance function
[14]Dividendprofit — Machine Learning in Banking
https://dividendprofit.com/machine-learning-in-banking/
[15][18]TowardsdataScience — The growing impact of AI in financial services
[16]Deloitte — Africa: A 21st century review
[17][19]Juniper Research — AI in Fintech
https://www.juniperresearch.com/press/press-releases/bank-cost-savings-via-chatbots-reach-7-3bn-2023
[21][22]McKinsey&Company — The new frontier in anti-money laundering
[23]Financial Conduct Authority — Using Artificial Intelligence to keep criminal funds out of the financial system
[24]Sienna Project — Ethical Analysis of AI and Robotics technologies
[26]U.S Securities and Exchange Commission — Staff Report on Algorithmic Trading in U.S. Capital Markets
https://www.sec.gov/files/Algo_Trading_Report_2020.pdf
[27]BBC — Knight Capital shares stabilise after costly IT glitch
https://www.bbc.co.uk/news/business-19116715
[29]Valuates Reports –Algorithmic trading market overview
https://reports.valuates.com/market-reports/360I-Auto-2W81/the-global-algorithmic-trading
[30]CNBC — A major wall Street player is cutting 1500 jobs and accelerating automation
[31]Business Insider — Deutsche Bank says robots are already replacing workers as it ramps up a plan to axe 18000 jobs
[32]Bloomberg — Robots to cut 200,000 U.S. Bank jobs in next decade, Study says
[33]Business Insider — AI will have a transformative effect on Wall Street, putting 1.3 million finance jobs in the US at risk
[34]Financial News — Robots to take 400,000 finance jobs in the next decade
https://www.fnlondon.com/articles/robots-to-take-400000-finance-jobs-in-the-next-decade-20190521
[36]Bloomberg — Robots in finance could wipe out some of its highest paying jobs
[37][42]Testimony to the House Committee on Financial Services Task Force on Artificial Intelligence Hearing: “Robots on Wall Street: The Impact of AI on Capital Markets and Jobs in the Financial Services Industry
https://financialservices.house.gov/uploadedfiles/hhrg-116-ba00-wstate-fenderr-20191206.pdf
[38][39][41]Bloomberg — Finance needs people who work well with Robots
https://www.bloomberg.com/news/articles/2019-08-20/finance-needs-people-who-work-well-with-robots
[40]Linkedin — 2020 Emerging jobs report, UK
