FINANCE AND ARTIFICIAL INTELLIGENCE
Author Sushant Shashikant Borge
Advantages and weaknesses of classic and modern aids techniques
LONGBING CAO (2023) state that AI in finance has been a continued substantial research direction over decades with increasing cross-disciplinary interactions and fusion between AI, data science, machine learning, finance, and economics. This trend has been further enhanced in recent years with the fast development of new generation AI and data science and their applications to broad-based financial applications. This review presents a comprehensive and dense overview of the advantages and weaknesses of classic and modern AIDS techniques in finance. They further review and comment on the data-driven methods in financial applications. The review also discusses the open issues and future opportunities of new-generation AIDS in finance and their synergy. This review significantly leverages many related reviews where only specific AI methods or financial problems are the focus.
AI APPROACH TO MEASURING FINANCIAL RISK
YU, L. et al (2023) they proposed and developed an AI based measure for systemic risk in financial markets: the FRM. The FRM is a measure for systemic risk based on the penalty term λ of the linear quantile lasso regression, which is defined as the average of the λ series over the 100 largest US publicly-traded financial institutions. The implementation is carried out by using parallel computing. The risk levels are classified by five levels. The empirical result shows that our FRM can be a good indicator for trends in systemic risk. Compared with other systemic risk measures, such as VIX, SRISK, Google Trends with the keyword “financial crisis”, they find that the FRM and VIX, FRM and SRISK, FRM and GT mutually Granger cause one another, which means that FRM is a good measure of systemic risk for the US financial market. All the codes of FRM are published on www. quantlet.de with keyword FRM. The R package Risk Analytics (Borke, 2017b) is another tool with the purpose of integrating and facilitating the research, calculation and analysis methods around the FRM project (Borke, 2017a). The up-to-date FRM can be found in http://frm.wiwi.hu-berlin.de.
The Robots are Coming … But Aren’t Here Yet
BAKARICH, K. M. and O’BRIEN, P. E. (2021) they survey public accounting professionals to gauge the extent to which AI technologies, specifically RPA and ML, are currently being utilized by the profession, as well as perceptions about the impact and receptiveness to the technology. While much has been written about the potential for AI to vastly change the profession, this study aims to capture the opinions of a variety of public accounting professionals on the current and future state of adoption. Ninety participants representing various firms, service lines, and positions responded. The results of quantitative and qualitative responses indicate that both RPA and ML are currently not being used extensively by public accountants nor by clients of public accounting firms, and firms are conducting some, but not extensive training on these technologies for employees. However, respondents strongly indicated in both numeric and written responses that AI will significantly impact their daily responsibilities in five years and that employees in the profession are very receptive to these changes. These results indicate that while large-scale AI adoption has not yet come to public accounting, substantial changes are on the horizon.
Fields theory of finance
ČIŽINSKÁ, R et al (2016) say that the idea of a “fields theory of finance” that builds upon vector fields, information theory and complex networks theory. It is compatible with both mainstream finance and artificial intelligence tools. The FieldsRank approach can be applied to the brand or the business as a whole, even in conditions of uncertainty without access to private information. We validated the model using real data from the microinvesting industry, a branch of Fintech. Bankers have developed interest in recent months, not only because of competitive pressures, but because they have started partnering with and developing those businesses themselves.
Adapting Expert System Technology to Financial Management
HOLSAPPLE C. W et al (1988) say that the potential impact of expert systems on financial management is certainly significant. The successful use of this technology lies in a thorough understanding of what it is and the creation of more powerful development tools. For the present there are both limitations and opportunities . For financial applications that are less sophisticated or require a computation speed that cannot be delivered by a human (e.g., spotting an arbitrage opportunity in real time), developers will find the current technology can provide feasible systems. However, current technology is inadequate for applications requiring insight, creativity, and intuition. Today, the technology can be usefully employed to build financial decision support systems that exhibit artificial intelligence. Yet financial applications of expert systems are still in an infant stage. As hardware and software technology advances, particularly in the environment direction, more applications will become feasible. Until then, expert systems can play a modest supportive role in financial domains by easing the burden from human experts.
The link between statistical learning theory and finance
MAASOUMI, E. and MEDEIROS, M.( 2010) say that forecasting a macroeconomic or financial variable is challenging in an environment with many potential predictors whose predictive ability can vary over time. In the first four articles in this compendium, this problem is tackled with techniques which have their roots in the Statistical Learning Theory, such as factor models, bagging, and forecast combination.
Co-opting Artificial Intelligence as an Opportunity for Financial Service Professionals.
WHEELER, D. W. (2020) say that while some see the replacement of humans, others see AI as a tremendous tool to streamline advisors activities to focus on more high-impact interactions by off-loading mundane repetitive service tasks and number-crunching to the AI-powered virtual assistants. Financial professionals will be freed to spend more meaningful time acting as the interpreter and primary communicator of the data output from AI based virtual assistants. Extinction of financial advisors is not soon likely as organizations learn to leverage AI as a tool, providing engaged advisors with efficient and effective tools for analytics, automation, and operational support. With this in mind, financial service professionals are wise to embrace these technological developments and adapt to the potential while filling the voids that still require social skills, emotional intelligence, creativity, communication, and understanding of human nuances. Understanding human nuances enables professionals to hear a client’s stated goals and use emotional intelligence to engage in a deeper dialog to uncover additional client needs, goals, and desires. Emotional intelligence has always been important but will have increased significance as a differentiator between humans and AI-powered virtual advisors.
Artificial Neural Network Assistant (ANNA) for Continuous Auditing and Monitoring of Financial Data
KOSKIVAARA, E. and BACK, B. (2007) state that ANNA was developed in a step-by-step manner involving a continual dialogue with a CPA-auditor, internal auditors, and the management of the organization (Koskivaara 2000). The idea of ANNA is to take advantage of the data that already exists in the companies and use it in a monitoring and controlling context. A well-known fact is that companies collect monthly reports to support their operations. Therefore, the core of ANNA is built on the monthly data. As mentioned in the introduction, ANNA produces expectations both with the simple quantitative techniques and with the artificial neural network (ANN) technology.
Artificial Neural Networks to Pick Stocks
KRYZANOWSKI. L et al (1993) state that as a pattern classifier, the BM appears to have potential for stock picking based on a modest number of learning examples. Superior precision based on greater learning is presumably achievable if a greater number of learning cases are used. However, the reported results may be unique to the limited sample of smaller companies and the relatively short time period (three years) studied here. The literature suggests that smaller companies are more likely to experience delayed stock responses to new information. Whether an actively managed portfolio using a BM can outperform a passive portfolio remains a question for future testing. Future research should explore the eflrectiveness of using a BM for stock picking for a greater number of examples covering a longer time frame and a wider range of firm sizes. This should be done for various future time horizons using additional features such as the number of shares outstanding, market capitalization and analysts’ earnings expectations.
Notes on an Expert System for Capital Budgeting
MYERS, S. C.( 1988) says that most of the Analysis block of the system is designed to help the user get the cash flow forecasts right. Finance people exhort operating managers to make consistent, unbiased forecasts. When managers fail to do so. it is easy to assume that they do not hear, understand, or agree. However, poor forecasts often reflect the problems managers have in translating their business knowledge into long-run numerical projections. The chemical company management’s grasp of their competitors” future behavior may have been more sophisticated than the microeconomic model underlying of competitors’ impact. But the capital budgeting procedures of the company did not tap the managers’ knowledge and sophistication. The capital budgeting forms did not ask for numerical input about the competition, only for I inc-by-l ine forecasts of sales, cost of goods sold, inventory, taxes, etc., probably for a much longer future period than the managers normally thought about. Of course, financial staff checked for internal consistency and for realism line by line. But without the system no one was prompted to model the project explicitly from the competitors” point of view.
Conclusion
AI is increasing day by day and now it has also entered in finance . It has some good and bad effects.AI calculate the financial risk and also gives solution .Many professionals are using AI specially RPA and ML. Ninety participants representing various firms, service lines, and positions responded. Field theory of finance is build on victor field, information theory and complex network theory. Field theory of finance is suitable for both main stream finance and AI tools . Bankers have develop interest because they have started partnering . To manage finance people they adopt expert systems technology. Financial professionals will be free to spend more meaningful time acting as the interpreter and primary communicator of the data output from AI based virtual assistants. ANNA Controlling and monitoring the context and it also helps us to select stock. Many finance people use AI for capital budgeting .They share their idea on AI then AI give them budget to establish their business.
REFERENCES
BAKARICH, K. M.and O’BRIEN, P. E. (2021)The Robots are Coming … But Aren’t Here Yet: The Use of Artificial Intelligence Technologies in the Public Accounting Profession. Journal of Emerging Technologies in Accounting, [s. l.], v. 18, n. 1, p. 27–43, 2021. DOI 10.2308/JETA-19-11-20-47. Disponível em: https://research.ebsco.com/linkprocessor/plink?id=53ad478c-086c-308b-b49d-9696e28e5c6f. Acesso em: 24 fev. 2024.
ČIŽINSKÁ, R.; KRABEC, T.; VENEGAS, P. FieldsRank: (2016) The Network Value of the Firm. International Advances in Economic Research, [s. l.], v. 22, n. 4, p. 461–463, 2016. DOI 10.1007/s11294-016-9604-x. Disponível em: https://research.ebsco.com/linkprocessor/plink?id=0f3e70c1-c951-306f-afa6-6c5be0f8f772. Acesso em: 24 fev. 2024.
HOLSAPPLE, C. W.; KAR YAN TAM; WHINSTON, A. B. , (1988) Adapting Expert System Technology to Financial Management. FM: The Journal of the Financial Management Association [s. l.], v. 17, n. 3, p. 12–22, 1988. DOI 10.2307/3666068. Disponível em: https://research.ebsco.com/linkprocessor/plink?id=e211ad5c-5f5f-34eb-80c9-942f33f4ac53. Acesso em: 24 fev. 2024.
KOSKIVAARA, E. and BACK, B. (2007) Artificial Neural Network Assistant (ANNA) for Continuous Auditing and Monitoring of Financial Data. Journal of Emerging Technologies in Accounting, [s. l.], v. 4, p. 29–45, 2007. DOI 10.2308/jeta.2007.4.1.29. Disponível em: https://research.ebsco.com/linkprocessor/plink?id=5db89ff2-888a-36d0-a961-268097a5ead8. Acesso em: 24 fev. 2024.
KRYZANOWSKI, L.; GALLER, M.; WRIGHT, D. W.( 1993) Using Artificial Neural Networks to Pick Stocks. Financial Analysts Journal, [s. l.], v. 49, n. 4, p. 21, 1993. DOI 10.2469/faj.v49.n4.21. Disponível em: https://research.ebsco.com/linkprocessor/plink?id=63b5a6ca-215b-35e7-9708-a5d53ffad5fa. Acesso em: 24 fev. 2024.
LONGBING CAO(2023). AI in Finance: Challenges, Techniques, and Opportunities. ACM Computing Surveys, [s. l.], v. 55, n. 3, p. 1–38, 2023. DOI 10.1145/3502289. Disponível em: https://research.ebsco.com/linkprocessor/plink?id=569c1193-ec50-3ce6-b6e6-c07435993edb. Acesso em: 24 fev. 2024.
MAASOUMI, E. and MEDEIROS, M.( 2010)The Link Between Statistical Learning Theory and Econometrics: Applications in Economics, Finance, and Marketing. Econometric Reviews, [s. l.], v. 29, n. 5/6, p. 470–475, 2010. DOI 10.1080/07474938.2010.481544. Disponível em: https://research.ebsco.com/linkprocessor/plink?id=2f25f2d0-1b51-393c-a953-26c82b480017. Acesso em: 24 fev. 2024.
MYERS, S. C.( 1988) Notes on an Expert System for Capital Budgeting . FM: The Journal of the Financial Management Association, [s. l.], v. 17, n. 3, p. 23–31, 1988. DOI 10.2307/3666069. Disponível em: https://research.ebsco.com/linkprocessor/plink?id=b724628f-a766-33d5-9d06-5ac9542a01f4. Acesso em: 24 fev. 2024.
WHEELER, D. W. (2020)Co-opting Artificial Intelligence as an Opportunity for Financial Service Professionals. Journal of Financial Service Professionals, [s. l.], v. 74, n. 1, p. 66–72, 2020. Disponível em: https://research.ebsco.com/linkprocessor/plink?id=5923d78d-6425-3be4-b93c-59f7d6c9e016. Acesso em: 24 fev. 2024.
YU, L. et al. An Ai Approach to Measuring Financial Risk. Singapore Economic Review, [s. l.], v. 68, n. 5, p. 1529–1549, 2023. DOI 10.1142/S0217590819500668. Disponível em: https://research.ebsco.com/linkprocessor/plink?id=4dd297ad-23b6-3b39-905c-5212f29da019. Acesso em: 24 fev. 2024.