analyze-13f-filings-and-holdings

name	analyze_13f_filings_and_holdings
description	Guidelines for analyzing 13F filings, finding fund accession numbers, calculating AUM, and counting exact stock holdings by utilizing existing scripts and correctly excluding options.

1. Utilize Existing Scripts

Before writing custom pandas code from scratch, explore your environment:

• Always check the scripts/ directory first by running ls scripts/.
• There are likely pre-existing Python scripts built specifically for these tasks (e.g., fuzzy searching fund names, analyzing a single fund, and getting total stock holdings/AUM).
• Run these scripts with the -h or --help flag (e.g., python scripts/analyze_fund.py --help) to understand how to leverage them. This saves time and prevents reinventing the wheel.

2. Fuzzy Searching for Accession Numbers

To find a fund (e.g., "Renaissance Technologies" or "Berkshire Hathaway"):

• Search the COVERPAGE data in the appropriate quarter's directory (/root/2025-q2/ or /root/2025-q3/).
• If a fuzzy search script exists in scripts/, use it to extract the correct accession_number. Note that accession numbers change between quarters, so you must find the specific accession number for each respective quarter.

3. Counting Stocks vs. Total Holdings

When a question specifically asks for the number of stocks held by a fund:

• CRITICAL LOGIC: Do not simply count all rows in the holding table.
• SEC 13F filings include options derivatives like PUT and CALL. These must be excluded to get the correct stock count.
• If using an existing script, check if it has a flag to filter stocks or exclude options.
• If writing custom Pandas code, load the holdings file (Information Table) and filter out the options using the putCall column (or equivalent):

import pandas as pd

# Load the holdings/infotable
df = pd.read_csv("path_to_infotable.csv")

# Filter out Put/Call options to only count true stock holdings
# Options are usually labeled 'PUT' or 'CALL' in the 'putCall' column
if 'putCall' in df.columns:
    stocks_only = df[df['putCall'].isna() | (df['putCall'] == '')]
else:
    stocks_only = df

stock_count = len(stocks_only)
print(f"Number of stocks: {stock_count}")

4. Calculating AUM

AUM (Assets Under Management) is typically the sum of the value column across all holdings for a specific accession_number. Ensure you use the provided scripts in scripts/ to calculate this accurately, or sum the value column in the Information Table.