Artificial intelligence (AI) seemly is very helpful in science and medicine and elsewhere. At the time of writing the book, your authors were skeptical about the ability of Artificial Intelligence to have the “good thinking” required for value investing, especially to “value by the book.” That skepticism was expressed in chapter 1. Let’s say we stand at a point of opportunity and uncertainty.
But let’s monitor development and see how AI progresses. We’ll introduce AI techniques on the website chapters when they are relevant to the material there. If you see other applications to fundamental analysis, please pass them along. As you read, think of two things. First, there is the all important question of the performance of AI: How well does it work? What is the best model for a specific task, if any? Second, how do you efficiently integrate AI into your analysis workflow?
“This thing is not intelligence. It has no understanding of what it is saying.”
Usama Fayyad, Institute for Experimental AI at Northeastern University, as told to WSJ. WSJ, January 24, 2024, p. R4.
AI and Accounting
Accountants were the first to use spreadsheets, invented in 1979 and marketed as VisiCalc. That was not just efficiency for a hum-drum task; there is something about rows and columns that imbeds accounting thinking. A spreadsheet by itself is no good unless you get the numbers in the right cells and accounting is a discipline that dictates the cells (called accounts) in an organized way such that the accounts as a whole convey a message. Value investing by the book does the same accounting, and spreadsheets lend themselves to developing an investing tool as many students have found. Is AI up to the task? AI will be more intelligent if it is governed by an accounting discipline that organizes the inputs. Good thinking for AI?
AI and Finance
The following paper shows how AI (as of 2024) can be helpful in finance with innovations in research methods linked to AI tools.
Eisfeldt, A. and G. Schubert. 2024. AI and Finance. At https://ssrn.com/abstract=4988553.
AI and Investing
A paper concludes that predicting the stock market with Machine Learning is difficult, as with most other quantitative techniques. The conclusions reflect the remarks of panelists on a March 28, 2023, panel titled “Advances of ML Approaches for Financial Decision Making & Time Series Analysis” at the 2022 Applied Machine Learning Days (AMLD) organized by the Swiss Federal Institute of Technology, now published in The Journal of Financial Data Science, Spring 2023 issue.
That is of 2023. Watch for any further reports.
Here is a paper that compares AI to analysts and how they complement each other:
Cao, S., W. Jiang, J. Wang, and B. Wang. 2024. From Man vs. Machine to Man + Machine: The Art and AI of Stock Analyses. Journal of Financial Economics 160, article 103910.
AI and Economic Indicators
The following paper reports that managerial expectations cleaned by generative AI from conference call transcripts predict GDP, production, and employment, and better than survey forecasts.
Jha, M., J. Qian, M. Weber, and B. Yang. 2024. Harnessing Generative AI for Economic Insights. At https://www.ssrn.com/abstract=4976759.
AI Performance Evaluation
FinanceBench is a suite for evaluating LLMs on answering financial questions. See https://arxiv.org/abs/2311.11944. A paper at that site tested 16 state of the art AI models in answering questions on financial topics as of 2023. The performance has improved as of 2025 with DeepSeek and ChatGPT-03 (we are told).