Finance

Generating model documentation, sourcing, and comments

Use AI to draft sourcing notes and explanatory comments inside a financial model—summarizing where figures came from and why certain modeling decisions were made—to improve transparency and reviewability.

Why the human is still essential here

A human must verify the sourcing narrative and rationale, apply professional judgment on modeling choices, and ensure documentation is accurate and appropriate for stakeholders.

How people use this

Cell-level comments for key drivers

Generate concise cell comments that document assumption logic (e.g., growth, margins, working capital days) and cite the source for each key input for reviewer sign-off.

Claude / Microsoft Copilot

Model methodology and assumptions memo draft

Draft a short methodology/assumptions write-up from the workbook and source documents to accelerate VP/associate review and client-ready deliverables.

ChatGPT / Claude

Tie-out checklist and QA variance notes

Produce a review checklist (balance sheet balances, circularity handling, cash sweep behavior) plus a variance log template to support systematic model QA.

Microsoft Copilot / ChatGPT

Community stories (1)

LinkedIn

Investment Banking: Ranking the Best AI Tools for Financial Modeling (2026) 🤯

Investment Banking: Ranking the Best AI Tools for Financial Modeling (2026) 🤯

We extensively tested OpenAI's ChatGPT, Anthropic's Claude, Microsoft Copilot Agent Mode, and Shortcut on a real three-statement Excel model using investment banking standards.


Here’s what actually works, and what doesn’t.


Quick Answer: The Rankings 👇


Best Overall: Shortcut


Close Second: Claude (Opus 4.6)


Third Place: Microsoft Copilot (GPT-5)


Distant Fourth: ChatGPT (GPT-5.2)


The Testing:


Criteria 1: How It Feels Working With The Tool


Understanding the Assignment:


Winners: Claude and Shortcut


Claude and Shortcut asked thoughtful clarifying questions after receiving the prompt, about:


• Forecast preferences

• Revenue segmentation

• Share repurchases

• Layout decisions

• Schedule structure


That behavior closely resembles what you’d want from a good junior analyst.


Copilot and ChatGPT asked none.


Speed:


Winner: Shortcut


Shortcut and Claude completed the setup in ~15 minutes vs ~25 minutes for Copilot.


ChatGPT took close to an hour.


A decent analyst would have taken 1-2 hours to complete the assignment.


So all were faster than a human analyst at initial setup.


Criteria 2: Data Extraction, Formatting and Best Practices


Formatting:


Winner: Shortcut


Shortcut and Claude produced the most “investment-bank-like” outputs.


Shortcut was more consistent with input coloring and structure.


Claude missed several formatting conventions.


Copilot ignored IB formatting entirely.


Accuracy:


Winner: Copilot


Copilot won, but this was disappointing across the board.


Shortcut and Claude hallucinated significant portions of historical data.


In both cases, the errors were subtle enough to be dangerous, with slightly incorrect line items all adding up to correct subtotals.


Shortcut’s second attempt returned almost no mistakes.


Claude continued to generate bad data.


Fixing this would require careful cell-by-cell auditing that takes longer than just inputting the numbers yourself.


As a rule, analysts should not rely on these agents to find data and should instead upload PDFs and spreadsheets for the agents to work with.


Copilot and ChatGPT were more accurate.


ChatGPT’s presentation was the least polished, but its historical balance sheet was easiest to audit.


Had Shortcut and Claude used correct data, they would have won as they were attempting a more analytically rigorous presentation.


Shortcut was also going into the footnotes to break out certain items when appropriate.


Sourcing and Commenting:


Winner: Claude


Claude provided the best explanations of where data came from and why certain modeling decisions were made.


Copilot added no comments.


ChatGPT added too many.


Shortcut did some, but less consistently.


Claude was also the only tool to backsolve EBITDA correctly.


The Bottom Line:


Shortcut and Claude significantly outperform Copilot and ChatGPT.


But right now, even the best tool still underperforms a Junior Analyst.

MF
Matan FeldmanCEO & Founder, Wall Street Prep
Feb 23, 2026