Research Review | 18 July 2024 | Artificial Intelligence and Finance

The Finance AI Challenge: An Evaluation of the Top Six Free Web-based AI Models
David Krause (Marquette University)
June 2024
This article evaluates six free web-based AI models-ChatGPT, Gemini, Copilot, Claude, Perplexity, and Meta AI-in their performance on finance-related tasks. Utilizing a structured approach, we assessed the models’ abilities to handle factual, conceptual, and computational queries, as well as their proficiency in Python coding through a financial case study. Our findings indicate that ChatGPT, Copilot, and Perplexity consistently excelled, particularly in delivering accurate, comprehensive, and well-structured responses. However, challenges such as maintaining context, ensuring factual accuracy, and mitigating biases persist. The study underscores the need for future research to enhance domain adaptation, explainability, and ethical considerations to ensure reliable and responsible use of AI models in finance.

A First Look at Financial Data Analysis Using ChatGPT-4o
May 2024
Zifeng Feng (University of Texas at El Paso), et al.
May 2024
OpenAI’s new flagship model, ChatGPT-4o (with its “o”standing for “omni”), released on May 13, 2024, offers enhanced natural language understanding and more coherent responses. In this paper, we investigate ChatGPT-4o’s capabilities in financial data analysis, including zero-shot prompting, time series analysis, risk and return analysis, and ARMA-GARCH estimation. We find that ChatGPT-4o’s performance is generally comparable to traditional statistical software like Stata, though some errors and discrepancies arise due to differences in implementation. Despite these issues, our findings indicate that ChatGPT-4o has significant potential for real-world financial analysis. Integrating ChatGPT-4o into financial research and practice may lead to more efficient data processing, improved analytical capabilities, and better-informed investment decisions.

Advancing Portfolio Construction and Optimization: AI’s Role in Boosting Returns, Lowering Risks, and Streamlining Efficiency
Michael Schopf (Schopf Meta Consult)
February 2024
This paper is a practical guide on how Artificial Intelligence (AI) and Machine Learning (ML) can support professional investors in portfolio construction and optimisation and identifies three methods for seamlessly integrating ML-based portfolio construction into an existing investment process. It provides a compelling comparative analysis of traditional techniques and modern ML-based approaches to portfolio construction and optimisation. The paper illustrates how, unlike traditional tools such as mean-variance optimisation and the capital asset pricing model, ML methods adapt dynamically to market changes, acting like a navigation system or GPS in the ever-evolving financial terrain. This adaptability enables ML based portfolios to outperform traditional methods through better predictive analytics, automated rebalancing and risk management, leading to more efficient, scalable and customised portfolio solutions. The paper argues that integrating ML into portfolio construction is not just an upgrade, but a significant innovation in asset management. This offers precision and efficiency beyond the capabilities of traditional methods, thereby increasing portfolio returns, reducing risk, and improving efficiency.

Can Large Language Models Beat Wall Street? Unveiling the Potential of AI in Stock Selection
George Fatouros (Alpha Tensor Technologies), et al.
January 2024
This paper introduces MarketSenseAI, an innovative framework leveraging GPT-4’s advanced reasoning for selecting stocks in financial markets. By integrating Chain of Thought and In-Context Learning, MarketSenseAI analyzes diverse data sources, including market trends, news, fundamentals, and macroeconomic factors, to emulate expert investment decision-making. The development, implementation, and validation of the framework are elaborately discussed, underscoring its capability to generate actionable and interpretable investment signals. A notable feature of this work is employing GPT-4 both as a predictive mechanism and signal evaluator, revealing the significant impact of the AI-generated explanations on signal accuracy, reliability and acceptance. Through empirical testing on the competitive S&P 100 stocks over a 15-month period, MarketSenseAI demonstrated exceptional performance, delivering excess alpha of 10% to 30% and achieving a cumulative return of up to 72% over the period, while maintaining a risk profile comparable to the broader market. Our findings highlight the transformative potential of Large Language Models in financial decision-making, marking a significant leap in integrating generative AI into financial analytics and investment strategies.

Integrating Generative AI into Financial Market Prediction for Improved Decision Making
Chang Che (George Washington University), et al.
April 2024
This study provides an in-depth analysis of the model architecture and key technologies of generative artificial intelligence, combined with specific application cases, and uses conditional generative adversarial networks ( cGAN ) and time series analysis methods to simulate and predict dynamic changes in financial markets. The research results show that the cGAN model can effectively capture the complexity of financial market data, and the deviation between the prediction results and the actual market performance is minimal, showing a high degree of accuracy. Through investment return analysis, the application value of model predictions in actual investment strategies is confirmed, providing investors with new ways to improve the decision-making process. In addition, the evaluation of model stability and reliability also shows that although there are still challenges in responding to market emergencies, overall, GAI technology has shown great potential and application value in the field of financial market prediction. The conclusion points out that integrating generative artificial intelligence into financial market forecasts can not only improve the accuracy of forecasts, but also provide powerful data support for financial decisions, helping investors make more informed decisions in a complex and ever-changing market environment. 

AI at the Frontier of Economic Research
Oliver Giesecke (Stanford University-Hoover Institution)
February 2024
The use of AI gets increasingly incorporated into economic research. Some of the applications include sentiment analysis, classification, clustering, or function approximation for the development of flexible models. Another promising development is the use of AI for the selective extraction of data that is subsequently analyzed by the conventional economic toolkit. Latter is enabled by the recent advances in image and language transformer models. This approach has the potential to significantly expand the possibilities for economic research as demonstrated by an example that extracts debt maturity data from thousands of PDF documents.

Learn To Use R For Portfolio Analysis
Quantitative Investment Portfolio Analytics In R:
An Introduction To R For Modeling Portfolio Risk and Return
By James Picerno