Source

I came across a time series analysis of German Market in the light of federal elections and decided to try the same for BSE SENSEX(Bombay Stock Exchange Sensitive Index) which is a stock market index of 30 companies listed at the Dalal Street with Indian Lok Sabha Elections.

Getting Data

I downloaded the dataset from BSEIndia.com. I believe this should also be available on Quandl.

Loading Sensex

## 
## Attaching package: 'zoo'
## The following objects are masked from 'package:base':
## 
##     as.Date, as.Date.numeric

This reads the dataset from the csv file which seems to have a problem with the format. I had to add a column name, Datum. This caused a shift of columns as well, to tweak that I shifted the values from Low to Close.

Lok Sabha Election Dates

The election date is the date of counting of votes. If it isn’t available, I’ll pick the dates when the next Lok Sabha. There’s usually a gap of couple of days when the last Lok Sabha is dissolved and new Lok Sabha sworns in.

Dates not included because not interesting,

Computing Rolling Variances

Plot it!

I see two things here,

  1. Sensex has become very volatile over time.
  2. There is a slight hint that market is effected by the results, however it can only be seen in 2004, 2009 and 2014 elections.

Zoom In

Sensex volatality 100 days before and after the election.

I see a sign that markets have started becoming nervous around the elections and that volatality has increased.