Investors have already been using artificial intelligence for decades and, as the technology becomes more powerful, it will play an increasingly more active role in investment management. However, the terms used to describe AI are not always clear.
AI garnered mainstream attention in the last year after the release of a new and improved version of ChatGPT, a conversation tool that reached 100 million users within two months. Since then, AI has been in the news almost daily with an ever-increasing number of articles about how AI will impact every industry - as well as our everyday lives. Along with the new technology comes a bevy of new technical terms that are sometimes used interchangeably. This can make it difficult to separate hype from reality, or figure out exactly what people mean when they talk about the future of AI. To assist you, we have compiled a list of some of the core terms, explaining what they mean, and how they relate to Finance and Investing.
A Glossary of Essential AI Terms for the Finance Sector
A set of rules that computers follow to complete tasks. Given an input, such as a dataset, algorithms produce an output, like a pattern detected in the data. They are foundational to the technologies we use daily, from smartphones to online dating, and are increasingly utilised in diverse sectors like healthcare, law, and insurance.
In financial services, algorithm-driven systems are becoming increasingly common. Fraud detection tools use algorithms to spot unusual activity in customer accounts. Credit scoring software uses algorithms to analyse a wide variety of data when measuring a borrower’s credit worthiness. Algorithms have also powered robo-advisors, regulatory compliance software and even the forecasting and valuation tools businesses use to make key financial decisions.
In the investment space specifically, algorithmic trading has been used for decades to automate trading strategies and screen the market more efficiently. Institutional investors will also often rely on algorithms to help with risk management, sentiment analysis, and trade optimisation.
Potentially unfair or erroneous assumptions in algorithms that may favor or disadvantage certain groups. Often stemming from biased training data, algorithmic biases can perpetuate systemic prejudices related to race, gender, sexuality and more.
A number of major headlines showcase the impact that AI bias can have across most services and industries that leverage the technology, like police facial recognition software, healthcare algorithms, and even voice recognition software.
For example, a major US technology company had to scrap an experimental hiring algorithm when it started systematically screening out female candidates. This kind of AI bias happens when algorithms pick up on patterns in their training data and assume those patterns are predictive of future outcomes. In the hiring algorithm, 10 years of employee resumes showed a clear pattern of male dominance at the company so the AI assumed being male was a predictor of a candidate’s suitability for the job.
In investing, a similar problem of biased pattern recognition can lead to “overfitting” investment strategies to historical data. The pattern-detecting AI might draw a connection between variables in the historical data that turn out to be random, unrelated occurrences. When it’s then used in a real-world setting, it can end up performing worse than expected.
Artificial Intelligence (AI)
The creation and study of machines capable of tasks that traditionally required human intelligence. AI encompasses various facets of intelligence, including reasoning, decision-making, and problem-solving. Established in the mid-1950s, AI now powers tools like virtual assistants and online banking.
Across every application, the goal of artificial intelligence is generally to create a machine that can perform specific tasks for us faster and with less human error. Take AI directed trading as an example:
A human sets the parameters of the trading strategy, including the criteria for picking stocks and executing trades. But if that human also had to scan the entire stock market to find trades and manually execute each one, they would miss more opportunities than they would catch.
A computer, on the other hand, can screen massive volumes of data and perform complex analytics in fractions of a second and repeat the process all day long without the need for sleep or a lunch break. As a result, it has the potential to capture more trades that fit the strategy than the human who programmed it could.
The study of vast datasets. As technology advanced, our ability to store and analyse extensive data grew, leading to the rise of big data research. A primary challenge is deriving insights without compromising individual privacy.
The financial sector has long relied on data to make decisions. Investors, for example, have based their investment decisions on an analysis of public financial statements, company news, market research, stock price history, and whatever other forms of data they could find.
With the recent rise of Big Data, a growing number of investors are leaning towards data-centric investment models that use those traditional data sources along with some more unconventional ones like web traffic, patent applications, or even satellite images.
Investors are tapping into these unique data to gain a competitive edge and make better-informed investment choices. In the future, with the exponential increase of data, we can expect more creative uses of data in investment strategies
Software designed to simulate human conversation. Often powered by AI and large language models, chatbots can interact with users, answer questions, and provide information.
Chabots like ChatGPT, Bing, or Bard are based on Generative AI and use natural language processing to create human-like conversational dialogue. The language model can respond to questions and compose content, including articles, social media posts, essays, code, and emails.
Until recently, Chatbots in finance typically relied on earlier versions of AI which were very limited in their responses and incapable of retaining knowledge in order to maintain an ongoing conversation. As more powerful chatbots take their place, they’ll be able to perform increasingly more complex customer service and back office functions like answering questions, scheduling appointments, making product recommendations, and even resolving simple service requests.
The use of computers to extract information from digital images or videos. It has applications in object and facial recognition, medical imaging, navigation and surveillance.
In finance, computer vision can help automate activities like scanning documents, analysing images, detecting fraud, and processing claims. This automation can make businesses more efficient. For instance, insurers can let AI analyse images to evaluate damage to vehicles or structures , allowing for prompt estimation of repair expenses.
Information collected for analysis or reference, which can be numerical, textual, or multimedia. Unprocessed data are often referred to as 'raw' or 'primary' data.
Data are used extensively in finance and investment. Examples of data sets are wide-ranging depending on the asset class. Some examples include price, time, and quantity market data as well as credit scores and postal or zip codes.
The study of extracting insights from large datasets. Data Scientists come from diverse backgrounds and apply their expertise to various real-world challenges.
In finance, data science is already being used to find insights and make better business decisions. By studying historical data, data scientists can identify risk factors, find warning signs, and other predictive indicators that can be used to address vulnerabilities, limit exposure, or recognise potential issues in the future.
A similar approach is used to identify investment opportunities and help craft trading strategies. As more data becomes available, the data-driven decision making that’s used in the financial sector becomes more powerful.
Digitally manipulated media where someone appears to say or do something that they didn't actually say or do. Created using machine learning, deepfakes can be used maliciously, leading to concerns about misinformation.
As deep fake technology gets better, hackers and cybercriminals will be able to create increasingly sophisticated impersonations of financial institutions to commit fraud. In one dramatic example of this threat, a cybercriminal used AI voice cloning technology to impersonate the director of a major Japanese company, defrauding the company of roughly $35 million with a phone call.
A subset of machine learning using 'neural networks' to recognise patterns in data. Inspired by biological brain structures, deep learning powers voice recognition in devices and helps autonomous vehicles identify obstacles.
For examples of impact in finance see Machine Learning.
A computer program that mimics the decision-making of a human specialist within a particular field. It operates using a knowledge base filled with facts and heuristics to derive conclusions and make choices. As a subset of artificial intelligence, expert systems are often employed in sectors where specialised human knowledge is essential, but might be limited or expensive to obtain.
Expert systems can automate surprisingly complex problem-solving and analytical tasks. In finance, we see this in decision-making tools that use AI to evaluate potential investments, generate financial forecasts, process insurance claims, and even assist in making lending decisions.
A machine learning model trained on extensive data, adaptable for various applications. Large language models, like ChatGPT, are examples of foundational models.
For examples of the impact of foundational models in finance see Chatbots.
AI systems that produce text, images, or other media based on user prompts. They can create outputs nearly indistinguishable from human-made content.
Document generation, customer insights, real estate valuations, training and simulation are just some of the back-office areas where we can expect to see Generative AI impact finance in the coming years. Generative AI may also support stress-testing investment strategies in the future by creating new ‘synthetic’ market data and market conditions for testing purposes.
On the customer-facing side, we can expect better virtual financial assistants and robo-advisors that offer more personalised banking and wealth management solutions.
An error in Generative AI models where chatbots produce inaccurate or fabricated information. Hallucinations can lead to instances like citing non-existent books or incorrectly stating that elephants lay the largest eggs among mammals.
AI hallucinations are a challenge for financial firms that use Generative AI for customer service and other customer-facing services. It is unlikely to impact investment strategies directly.
Human in the Loop
A combined system of humans and AI, where the human plays a significant role, such as by training or tuning the AI. This collaboration leverages the strengths of both human and artificial intelligence.
By keeping human oversight at critical points in the process either as a monitor or as a gatekeeper, AI users can manage risk or build confidence in the technology during early-stage deployments.
One example of a successful use of human in the loop is the case of a widespread healthcare algorithm that a Stanford study discovered was making biased healthcare decisions for the roughly 200 million people that it had been applied to. By introducing a human in the loop they were able to significantly reduce the bias.
Large Language Model (LLM)
A model trained on vast textual data for language tasks. LLMs, part of the broader natural language processing field, can generate human-like text.
For examples of the impact of LLMs on finance, see Generative AI.
A branch of AI where algorithms learn patterns from a set of training data and apply those patterns to new data. It has promising applications across nearly every industry.
In the investment sector, machine learning technology is playing a pivotal role in guiding investment choices by pinpointing risks using past data and probability analysis. Additionally, it aids in evaluating the potential distribution of returns and therefore helping to develop investment strategies. The ability of an AI system to re-write elements of its own code as it learns new patterns is arguably the most powerful component of AI.
An AI Model is a program that’s been trained to recognise patterns in data. Models could be trained to predict the weather, translate languages, identify images, and identify human voice requests among other tasks.
In the investment sector, AI models are playing an ever increasing role in the investment process. Once trained, an AI model can be deployed to do anything from picking stocks to actually executing trades. While the success of any given AI model depends on how well it was designed and trained, they can be extremely useful in technical trading strategies or as tools to assist investors in building or rebalancing their portfolio.
Natural Language Processing (NLP)
An AI field focused on analysing and generating human language. NLP algorithms detect linguistic patterns, powering tools like chatbots and automatic translators.
NLP is used in finance for applications such as automated customer service via telephone. It is also applied widely in investment management for deciphering earnings calls and reports or performing sentiment analysis.
AI systems inspired by human brains, consisting of interconnected computational units. They excel in complex tasks like face and voice recognition as well as other pattern recognition applications.
For examples of impact in finance see Machine Learning.
Scams - Phishing
The fraudulent practice of sending emails or other messages purporting to be from reputable sources in order to induce individuals to send money or reveal private information, such as passwords and credit card numbers.
Americans lost over $10 billion to digital fraud in 2022. A big part of that stems from the increasingly sophisticated scams that cybercriminals can accomplish using AI. One example is the grandparent scam, in which a scammer uses voice cloning AI to impersonate a child or grandchild, usually calling for bail money or cash to cover an emergency expense. Scammers are even able to impersonate people via video calls by using AI to create deepfake voice and video.
We hope you found this article interesting as well as useful. Our aim is to provide succinct clarity in the area of AI in Investment Management. Please send us any feedback, or if you would like to see a specific subject covered in future, you can contact us at firstname.lastname@example.org.
Investors should consider the investment objectives, risks, charges and expenses carefully before investing. For a prospectus or summary prospectus with this and other information about the Fund, please call (800) 617-0004 or visit our website at dipetf.com. Read the prospectus or summary prospectus carefully before investing.
The Fund is distributed by Quasar Distributors, LLC. Exchange Traded Concepts, LLC (the “Adviser”) serves as the Fund’s investment adviser. Kaiju ETF Advisors (the “Sub-Adviser”) serves as the Fund’s investment sub-adviser.
Investing involves risk, including loss of principal. The Fund is subject to numerous risks including but not limited to: Equity Risk, Large Cap Risk, Management Risk, and Trading Risk. The Fund is actively managed and may not meet its investment objective based on the Sub-Adviser’s success or failure to implement investment strategies for the Fund. The Fund’s principal investment strategies are dependent on the Sub-Adviser’s understanding of artificial intelligence. The Fund relies heavily on a proprietary artificial intelligence selection model as well as data and information supplied by third parties that are utilised by such a model. Specifically, the Fund relies on the Kaiju Algorithm to implement its principal investment strategies. To the extent the model does not perform as designed or as intended, the Fund’s strategy may not be successfully implemented and the Fund may lose value. A “value” style of investing could produce poor performance results relative to other funds, even in a rising market, if the methodology used by the Fund to determine a company’s “value” or prospects for exceeding earnings expectations or market conditions is wrong. In addition, “value stocks” can continue to be undervalued by the market for long periods of time. The Fund is expected to actively and frequently trade securities or other instruments in its portfolio to carry out its investment strategies. A high portfolio turnover rate increases transaction costs, which may increase the Fund’s expenses. Frequent trading may also cause adverse tax consequences for investors in the Fund due to an increase in short-term capital gains. The fund is new, with a limited operating history.
This information should not be relied upon as research, investment advice, or a recommendation regarding any product, strategy, or security in particular. This material is strictly for educational or informational purposes, is subject to change, and is not intended to be a forecast of future events or a guarantee of future results. The information and opinions contained in this material are derived from proprietary and nonproprietary sources deemed by Kaiju to be reliable, are not necessarily all-inclusive and are not guaranteed as to accuracy. Reliance upon information in this material is at the sole discretion of the viewer.