run2_13f-stock-search

Overview

When searching for funds heavily invested in a single stock, be aware that CUSIP codes might be entered in varying cases (e.g. 69608A108 vs 69608a108), so normalizations like .str.upper() are critical.

Usage Patterns

Finding Top Funds by Value

This strategy avoids issues with case sensitivity and missing commas when parsing TITLEOFCLASS.

import pandas as pd

cusip = "69608A108"
quarter = "2025-q3"

infotable = pd.read_csv(f"/root/{quarter}/INFOTABLE.tsv", sep="\t", dtype=str)
coverpage = pd.read_csv(f"/root/{quarter}/COVERPAGE.tsv", sep="\t", dtype=str)

# Filter by CUSIP defensively
infotable["CUSIP_UPPER"] = infotable["CUSIP"].str.upper()
stock_holders = infotable[infotable["CUSIP_UPPER"] == cusip].copy()

# Sum the VALUE by ACCESSION_NUMBER
stock_holders["VALUE"] = stock_holders["VALUE"].astype(float)
fund_sums = stock_holders.groupby("ACCESSION_NUMBER")["VALUE"].sum().reset_index()

# Sort to get top 3
top_funds = fund_sums.sort_values(by="VALUE", ascending=False).head(3)

# Merge back to COVERPAGE to obtain FILINGMANAGER_NAME
merged = pd.merge(top_funds, coverpage[["ACCESSION_NUMBER", "FILINGMANAGER_NAME"]], on="ACCESSION_NUMBER", how="left")

print(merged[["FILINGMANAGER_NAME", "VALUE"]])