Informs Knowledge Cafe' (Spring 2021)

Tutorial-Data Collection

The Informs Student Chapter, University of Iowa holds an informal gathering known as Knowledge Café in which people share their knowledge and expertise. Anyone interested in learning from other researchers is welcomed. More about the University of Iowa Student INFORMS can be found here!
As a member of INFORMS Student Chapter, I hold a workshop to teach PhD and Master students (Business Analytics and Finance) how to setup the Application Programming Interface (API) for collecting data from Blomberg, Eikon, and WRDS. We mainly focus on Wharton Research Data Services (WRDS), Bloomberg terminal, and Refinitiv Eikon using Python API.

1. WRDS

Setup

Step 1

The installation is fairly easy. You need to create account or register here. For more information, you may ask Kim Bloedel, the head of Pomerantz Business Library.

Step 2

Install the WRDS Python Module by running the following command in your Anaconda Terminal.

 pip install wrds

Data Collection from WRDS

import the WRDS Python Module

import wrds

Connect to WRDS

db = wrds.Connection(wrds_username='your-username')

Find your library and tables. They can be found in the "Variable Description" of interested libraries, see the following picture:

Or, another method for discovering what tables are included inside a libraries, you can run the db.list_tables(library="intersted library").
For example, in the follwing picture, I am going to find out what tables are included in the library of 'comp':

The following command describes the variables of table 'wrds_seg_customer'

Now, let's say we are going to collect 'company name', 'gvkey' and 'sales to cutomers' from the table 'wrds_seg_customer' and the library 'comp'

Sometimes, it is just helpful to collect the required data. Let's say, we have a database taht shows companies and whether thay have twitter account or not.

Now, we are going to collect their 'sales', 'company name', and 'cusip' of these companies.

First, we need to create a dictionary of tickers with the following format:

Once, we create that dictionary, we submit the job to wrds for colleccting those variables only for these companies and only from January of 2010 till January of 2020. See the following picture:

Once the data is collected, we may again merge two dataframes:

So, this process can be applied to any other databases, the only thing you need to change is your wrds library and table.

2. Bloobmerg

Two python APIs are available:

2.1. Installation

2.1.1. pdblp

I personally do not recommend this API since I did not find it very user-friendly compared to the Xbbg (next one that I am using). But you can find installation and documentation in the following links:

Installation

Documentation

2.1.2. xbbg

You can find the installation and documetation

here

. Since the installation is not very straightforward compared to wrds, we will do it here again.

Step 1: Install xbbg

Run this code in your Anaconda terminalpip install xbbg

Step 2: Install blpapi

Run this code in your Anaconda terminalpconda install -c conda-forge blpapi

Step 3: Replace blpapi3_32.dll,etc

To do so, you need to login (this is only for the first time) and type 'WAPI' in Bloomberg terminal, as shown in the following pictures:

Next, download the file. I am using Windows, so I download the file which is compatible with windows.

Once you downloaded the file, you need to copy blpapi3_32.dll and blpapi3_64.dll under bin folder to Bloomberg BLPAPI_ROOT folder, normally blp/DAPI.

Step 4: Upload them in your Jupyter

Done!

2.2. Data Collection from Bloomberg by using xbbg

Step 1: Find the field

Let's say I want to find the revenue of Ford motors Co. from 2019-01-01 till 2020-12-30. So, I first type "revenue" in bloomberg terminal for finding the field asscoaied with the revenue. see the following picture

Collecting Historical data:

After finding the field, I use the syntax of blp.bdh('tickers', field,override,start_date='YYYY-MM-DD', end_date='YYYY-MM-DD') .
'tickers': This is where you put tickers. Notice that tickers format in WRDS are differnt than Bloomberg. For example, the ticker of Ford motors Co. in Bloomberg is 'F US Equity' but it 'F' in WRDS.
See the following figure which shows revenue of Ford motors Co. from 2019-01-01 till 2020-12-30

Collecting Non-historical Data:

For non-historical data, we use blp.bdp(tickers,fields)

Likewise, we can send a list of tickers for collecting historical or non-historical data:

3. EIKON

Eikon is a set of software products provided by Refinitiv. At the Uiowa we have one MAchine in Fin Lab have access to EIKON. API installation is fairly easy, you can find more information here!.