How to Download End-of-Day Stock Prices for the National Stock Exchange of India using Python?

How to Download End-of-Day Stock Prices for the National Stock Exchange of India using Python?

Welcome Readers 🤩 to the first article in the series "NSE EOD Stock Prices".

This article will cover the best way to download the End-Of-Day stock prices for every stock listed on the National Stock Exchange of India using Python without any sweat.

So without any delay, let's get started!! Because I am excited 🤭

We will be covering the following points in this article:

  1. What is the NSE Bhavcopy File?

  2. How to download the NSE Bhavcopy File using Python?

By the way, if you prefer watching videos rather than reading an article, we have covered this on our Youtube channel as well. Don't forget to subscribe :)

So let's jump in 🤹‍♂️

Update: On 5th March 2021, I had to make major changes to the article, as NSE stopped supporting full_bhavcopy_save for all years before 2020, if you are interested in reading the previous article, just see the Github change history.

What is the NSE Bhavcopy File?

Bhavcopy's literal translation in Hindi would be "stock-price sheet", and it fits in with the name because the Bhavcopy File will have all the end-of-day details of the securities trading on the exchange. You can download a sample Bhavcopy File by clicking here.

Consider Bhavcopy as a high-level snapshot of all trading activity for the day, which everyone in the industry can use. The following details are available in this File:

(Scroll below to see sample image to make sense of below points)

  • **SYMBOL: **The latest official symbol of the traded security.

  • **SERIES: **The latest Series the traded security is on (E.g., "EQ" for Equity, "BE" for Compulsory Delivery), TickerTape has a good blog on this, you can read more about it here. Frankly, for this project, we will only be dealing with the ["EQ", "BE", "SM"] series.

  • **OPEN: **The opening price of the security at 09:15 AM IST.

  • **HIGH: **The highest price of the security during the trading day.

  • **LOW: **The lowest price of the security during the trading day.

  • **CLOSE: **The closing price of the security at 15:30 IST. The closing price is the weighted average price of the last 30 mins of trading.

  • **LAST: **The actual last traded price of the security, which can differ from CLOSE_PRICE

  • **TOTTRDQTY: **The total units of security traded.

  • **TOTTRDVAL: **The total value of all the units of the security traded in Rupees.

  • **TIMESTAMP: **This will be the same date for which you are accessing the File.

  • **TOTALTRADES: **The total number of trades for that security for the day. (E.g., one trade might have 100 units of BUY and another might have 20 units of BUY, this will be considered as 2 total trades)

  • **ISIN: **The globally recognized identifier of the ticker.

I am sure you would now understand the importance of this File; this is the golden source of data used by most market participants whose trading strategies depend on EOD data. The preliminary data is released every day at ** 15:40 IST** but the majority of data becomes available at ** 16:00 IST**

Below is the snapshot of the sample file which was given in the link above.

image.png

How to download the NSE Bhavcopy File using Python?

Here comes the secret sauce 🥣, How do you download this File daily using Python, and how do you do it over the last 5 Years by running your code in a loop?

To be honest, it is straightforward and tempting to create your own program using the requests library of Python, which can do this task for you every day after 16:00 IST to collect this data. But why reinvent the wheel when there are already several options out there?

Introducing jugaad-data library by jugaad-py

This library is the easiest to use with its simple to understand syntax; it also supports the new NSE website. It can download bhavcopies for Stocks, Futures & Options, Indexes, Index Futures, and Options with a single line of code. It also supports caching which is cool. If that wasn't enough, it also supports fetching the live prices of particular security from the NSE website, but I will caution you to use that, given we don't want to make too many requests to NSE who can block our IP address. The detailed documentation of this API can be found here.

I wanted to let the readers know that this is not an official API by NSE, and I am in no way associated with this API.

***Before we get into coding, if you don't have jugaad-data installed on your machine, just type in pip install jugaad-data, and that should do the job. ***

image.png

Step 1: Testing how the output looks like

#Making all necessary imports for the code
from datetime import date
from jugaad_data.nse import bhavcopy_save

#Saving the Bhavcopy file for 01-01-2021
bhavcopy_save(date(2021,1,1), "path/to/folder")

#Tip: You can use "." instead of "path/to/folder"
#If you want to save your File where your code file is. 
#If unsure where your code file is, import os and run os.getcwd()

Can you believe it? That was it! Yes, for real, it's that sample to save the bhavcopy for any particular date; once you run the above, the cm01Jan2021bhav.csv should be saved in the path you defined.

image.png

Things to Note:

  • There are two functions in the jugaad-data library, one is full_bhavcopy_save and the other is bhavcopy_save, we are using the latter one as it even when it has less information because NSE is slowly removing the full_bhavcopy data from its website if you try to run full_bhavcopy_save for any dates before 2020, it throws up an error.

  • If you run the function bhavcopy_save for any date where the stock market is closed, it will throw up an error.

  • If you run the function for a future date, it will give you an ugly-looking error like this because it cannot find data on the NSE website for that day. This will also happen if you try to get this File before 16:00 IST for the day.

image.png

Step 2: How to download the Bhavcopy File for a range of dates?

This sounds simple to do, and you must have thought to do this in a for loop in Python, but there are a few complexities here, we need to account for "Weekends" and "Stock Exchange Holidays" in our code. Otherwise, we might get that ugly-looking error I showed above.

Don't worry, pandas has a special function, bdate_range, which will make our lives easier. Let's try and generate a bdate_range between 01-12-2020 to 31-12-2020, which has a couple of holidays.

import pandas as pd
from jugaad_data.holidays import holidays

date_range = pd.bdate_range(start='12/01/2020', end = '12/31/2020', 
                         freq='C', holidays = holidays(2020,12))

#bdate = business days (weekends excluded by default)
# start and end dates in "MM-DD-YYYY" format
# holidays() function in (year,month) format
#freq = 'C' is for custom

print(date_range)

image.png

As you can see in the screenshot above, all weekends have been excluded, and public holidays like Christmas have also been excluded.

Let's go ahead and now download this data for the whole of 2020 in one go; you can download going back as much as you want, but I wouldn't advise you to do that in one go as too many requests to NSE servers from your IP address may get you blocked.

date_range_2020 = pd.bdate_range(start='01/01/2020', end = '12/31/2021', 
                    freq='C', holidays = holidays(2020))

dates_2020 = [x.date() for x in date_range_2020]
#This is known as a list comprehension; let's print it out to see what happens

print(dates_2020)

image.png

You can see the difference between the two dates, the first one is a Pandas DatetimeIndex and the second one is a datetime.date, we have to use the latter because that's supported by jugaad-data, look at the very first example where we give it date(2021,1,1)

from random import randint
import time

for dates in dates_2020:
     try:
          bhavcopy_save(dates, "path/to/folder")
          time.sleep(randint(1,4)) #adding random delay of 1-4 seconds
     except (ConnectionError, ReadTimeoutError) as e:
          time.sleep(10) #stop program for 10 seconds and try again.
          try:
               bhavcopy_save(dates, "path/to/folder")
               time.sleep(randint(1,4))
          except (ConnectionError, ReadTimeoutError) as e:
               print(f'{dates}: File not Found')

image.png

And that's it, it will take a while to download the whole year's data in one go given the random delay of 1-4 seconds, but the process should be smooth, and you will see files appearing in the path you have defined.

You must have noticed the try and except method in the code; that's called exception handling; Make sure to Google more about it later.

You can also access the Github link here to access the whole code in one single file.

That's it for today, folks; I hope you enjoyed this article and our approach to downloading this data programmatically. In the next article, we will discuss more about what to do with so many data files and how to store them in time-series format. Remember to subscribe to my newsletter 📬 to get regular updates.

💡
Please note we haven't made any new posts since Nov 2021 on this blog, you are free to subscribe to the mailing list, however, you will be auto-added to the new blog's (thealtinvestor.in) mailing list as well.

If you need any help with the above or need some more information, you can ping me on Twitter or Linkedin.

If you like it up till now, consider buying me a coffee ☕ by clicking here or the button below.

Did you find this article valuable?

Support Trade With Python by becoming a sponsor. Any amount is appreciated!