Use Claude to Prepare a Quarterly Portfolio Review Report

What This Does

With Claude Pro, you can upload a portfolio data export (CSV or pasted table) and get a structured quarterly review report, summarizing risk rating migration, covenant compliance trends, watch list changes, and concentration exposures. Converts raw data into an executive-ready narrative.

Before You Start

• You have a Claude.ai Pro subscription
• You have a portfolio summary with key metrics per credit (CSV or spreadsheet export)
• You know what your bank's quarterly review report needs to cover

Steps

1. Prepare your data

Export from your HRIS, loan system, or Excel tracking sheet. At minimum: borrower name, loan type, outstanding balance, risk rating, DSCR, covenant status.

2. Open Claude Pro and upload the file

Use the paperclip icon to upload a CSV or paste the table directly.

3. Prompt for analysis

4. Ask for the narrative sections separately

The Prompts

Portfolio analysis setup:

I'm a commercial credit analyst preparing a quarterly portfolio review. I've uploaded [or: here is] a table of my loan portfolio with the following columns: [list columns].

Please analyze this data and provide:
1. Portfolio summary: total credits, total exposure ($), breakdown by loan type
2. Risk rating distribution: how many credits in each rating category
3. Covenant compliance summary: how many pass, watch, breach
4. Average DSCR across the portfolio; flag any credits below 1.10x
5. Any concentration risks you observe (borrower, industry, loan type)

Risk rating migration summary:

Compare the current risk ratings in column [X] to the prior quarter ratings in column [Y]. 
List: how many credits upgraded, downgraded, or stayed the same. Identify the 3 largest downgrades by exposure amount.

Executive narrative:

Based on this portfolio analysis, write a 3-4 paragraph executive summary suitable for presenting to the senior credit committee. Highlight: overall portfolio health, key risks to monitor, any material changes from last quarter, and recommended actions. Use formal, direct language.

Tips

• Before uploading any data, confirm your bank's data governance policy on using external AI tools with borrower information. You may need to anonymize names first.
• For anonymized data, use borrower codes: "Borrower A, $2.3M, DSCR 1.08x, Special Mention"
• Save Claude's output in a Word doc. Use it as the first draft of your quarterly review package.

Always have senior credit officer review before distributing portfolio review materials.