Summary of the Key Points
This article discusses a long-standing phenomenon in the financial industry where “smart people do stupid work”: graduates from prestigious universities spend countless hours each day on repetitive tasks such as data collection, cleaning, and organizing tables. However, AI tools similar to Codex in the programming industry (referred to as “Agents” in this article) are now entering the financial sector, taking over these basic chores. As a result, the focus of financial professionals has shifted from working late into the night to creating drafts to reviewing, analyzing, and forming meaningful opinions. The competitive landscape has also changed from who can complete these tasks faster to who can make more valuable decisions based on the AI-generated information.
1. The Daily Routine in Finance: Smart People Doing Time-Consuming Tasks
Many seemingly sophisticated jobs in finance are actually physically demanding. For example:
- Creating DD decks (due diligence presentation documents): When a boss requests a company analysis, you need to search for information on various platforms, parse data from financial reports, and compile it into a PPT—this is nothing more than transferring information.
- Quantitative research: Although it involves using mathematical models to identify market patterns, researchers spend most of their time collecting data, cleaning fields, writing code, and testing parameters, leaving very little time for strategizing.
- DCF valuation (estimating a company’s future value): Manually creating a complete model can take hours just for tasks like importing data from financial platforms and preparing the balance sheet.
These tasks require no special technical skills but are time-consuming. In the past, people accepted this as part of the job, and even working late was seen as a competitive advantage.
2. The Arrival of AI Agents: Taking Over the “Stupid” Work
Modern AI is no longer just a simple chatbot; it can perform complex tasks on its own, like a personal assistant:
- In quantitative research: AI can collect data and validate patterns (such as price-volume reversals or volatility), providing insights on which factors are ineffective and what adjustments should be made next. Researchers no longer need to repeat the coding process; they just need to evaluate the AI’s recommendations.
- DCF valuation: AI can automatically import data, prepare financial statements, and set assumptions, significantly speeding up this task.
- Daily tasks: For DD decks, AI can retrieve company information, organize summaries, and generate preliminary PPTs. When creating comparative tables, it can gather data from multiple companies and standardize the format.
In short, AI has taken over the tasks that were previously time-consuming but not particularly valuable.
3. The Change in Work: From Drafting to Analyzing and Guiding
Previously, financial professionals competed on how quickly they could complete drafts (e.g., by working late to finish a DD deck or producing a comparative table). Now, AI can generate preliminary drafts quickly, and the focus shifts to:
- Reviewing: Checking the accuracy of data sources and the consistency of measurement methods.
- Adjusting assumptions: Revising unrealistic projections (e.g., adjusting a company’s growth rate forecast).
- Refining: Transforming AI-generated content into compelling analyses (e.g., identifying key risks from financial data or selecting appropriate valuation indicators).
Senior analysts no longer need to start from scratch; they can build on the AI’s foundation to develop their investment strategies. The value lies in using the drafts as a starting point for more meaningful analysis.
4. Learning from the Programming Industry: How Codex Changed Programmers
The article uses the example of Codex in the programming industry to illustrate this transformation:
- Initially, programmers saw it as a tool that automatically completed minor tasks. But eventually, it helped them focus on more complex issues by handling routine tasks like document research and code writing.
- The result was a higher threshold for proficiency, as AI replaced time-consuming tasks, forcing everyone to develop more critical skills.
5. The New Competitive Landscape
The real competition now lies in what happens after the initial drafts are prepared:
- Identifying anomalies: Detecting unusual patterns in financial data (e.g., sudden revenue growth without a clear explanation).
- Choosing the right metrics: Deciding whether to use EV/EBITDA or P/E for valuation.
- Communicating effectively: Presenting complex information in a way that is easy for managers and clients to understand.
Those who adapt to these changes quickly will demonstrate their core capabilities. After all, while AI can handle basic tasks, the ability to analyze, make decisions, and communicate effectively remains the key to success in finance.
In summary: AI is not replacing financial professionals; it is helping them shift from mundane tasks to more valuable activities, transforming the competition from physical effort to intellectual challenge.