baskets: ETF Holdings Downloader, Parser and Disaggregator

Introduction

baskets is a project that provides Python library code that, given a list of Exchange Trade Funds and quantities, can

1. automatically download the compositions of those ETFs (i.e., the list of holdings) from the issuer's web pages and maintain a lightweight database of those downloads at multiple dates, and

2. collect the list of holdings and from the fractional constituents of each ETF reconstruct the dollar amount exposure to each constituent stock.

If you invest mostly with ETFs / baskets of stocks, it essentially answers the question: "What's my exposure to stock X?"

Motivation

You may be surprised that none of the brokers currently offer this service, and that automatically scraping the list of holdings from the issuers web pages is made difficult on purpose (although always available, as retail investors may want this information). However, download formats that would be convenient to pull from a computer program in a common format are typically only offered to financial institutions, or for an expensive fee (e.g., see services like etf.com, etfdb.com, etc.). I preferred to cook my own, since I had to write code on top anyway.

Usage

Two Python scripts are provided for the two aforementioned tasks: basksets-updatedb and baskets-portfolio.

Input Format

The tools require you to provide your portfolio in order to provide it with a list of ETFs and their quantities. The input format is a CSV file with the following fields:

• ticker: The ETF symbol, in capital letters. Individual stock symbols are supported as well (if no issuer is supplied, it is assumed that the position is for a single stock; see `issuer field below).

• account: The account which holds the ETF. This is used for detailed reporting whereby we list the presence of a particular stock in various ETFs and accounts. You can put any value here, it's a free-form string.

• issuer: The financial company which issues the ETF, that is, which allows traders to create or redeem it. Examples include iShares_, Vanguard_, PowerShares, SPDRs, and more <http://etfdb.com/issuers/>_. See the baskets/issuers directory for supported sources.

• price: The current price of the ETF. This project does not implement automatically price fetching, so you must provide the price of each instrument (manually, or perhaps via a script).

• quantity: number of shares of that ETF or stock which you hold. This is used along with the price to compute the dollar amount.

.. _iShares: http://www.ishares.com .. _Vanguard: https://investor.vanguard.com/etf/list .. _PowerShares: http://www.invescopowershares.com .. _SPDRs: https://us.spdrs.com

Here's an example input file::

ticker,account,issuer,price,quantity VFINX,Vanguard:401k,Vanguard,104.32,648.2332 VHT,Schwab,Vanguard,169.87,230 VNQ,Schwab,Vanguard,82.17,400 VTI,Schwab,Vanguard,145.78,1235

Installation

You need a recent Python-3 (at the time of writing this I've been developing using Python-3.7).

You will need to install the following Python libraries:

• requests
• xlrd
• openpyxl
• selenium
• networkx
• numpy
• pandas
• pytest

You can install them like this:

python3 -m pip install -r requirements.txt

Install this package:

python3 -m pip install .

Furthermore, you will need a WebDriver implementation for Selenium to control to fetch the holdings pages (from baskets-updatedb). You need the Chrome WebDriver, the others aren't guaranteed to work (I've witnessed the Firefox WebDriver fail on some pages, for example). Git it here:

https://sites.google.com/a/chromium.org/chromedriver/

Untar the binary and install it under /usr/loca/bin/chromedriver. Alternatively, baskets-updatedb accepts the --driver-exec= option where you can provide the location of your WebDriver executable.