Skip to content

Latest commit

 

History

History
51 lines (35 loc) · 2.69 KB

README.md

File metadata and controls

51 lines (35 loc) · 2.69 KB
Raw

Nifty50 next new highs : Statistical Analysis:

On an average, How may days and ETF long term investor has to wait to for NIFTY50 to achieve next highs?

This notebook is statistical analysis for below questions:

The Questions:

  • Wait Days : If I am a Nifty50 ETF investor, how many days on (average, 90%-tile, worst case) I have to wait till nifty50 makes a new high?
  • How does the Nifty50 chart looks like if we only plot the next highs and remove all jaggedness?
  • How does the distribution of wait days look like?
  • Additional exploratory analysis 1: Possible to check to what kind of distribution given Days data confirms to?
  • Additional exploratory analysis 2: Can you generate and artificial distribution for such data?

Disclaimer:

Standard Disclaimers apply. Also, this is only statistical anaylsis. Past performance may not prodict future perfomance.

1.1 Data Collator:

This section collates the individual csv files into a single file in order to facilitate the next data processing and analysis.

  • Note: We don't have to re-export data from Nifty website. Just fetch it from my previous github repo

1.2 Data derivation:

Derive a new column which is date difference. This date difference will tell you

  • number of days a long term investor have to wait till the next New High of Nifty50

1.3 Algorithm:

  • Wait Days : If I am a Nifty50 ETF investor, how many days on (average, 90%-tile, worst case) I have to wait till nifty50 makes a new high?

  • How does the Nifty50 chart looks like if we only plot the next highs and remove all jaggedness? : Plot using plotly.

  • How does the distribution of wait days look like? Plot sorted chart. Plot histogram.

1.4 Data Visulization:

image image image

1.5 Conclusion:

  • Thus there are 447 closing high values out of 6399 close values which is about 7%.
  • Average wait is about 20 sessions. Which is somewhat wrong.
  • Max wait we have seen is 1406 Day which is about 46 month and that happened between 11-02-2000 to 18-12-2003 (Dot com bubble)
  • Followed by 1094 days wait from 08-01-2008 to 05-11-2010 (Financial crisis) and then
  • 1032 days wait from 05-11-2010 to 03-11-2013
  • 90th percentile is ~18, 95th is ~48, 99th is ~529, 99.99 percentile is ~1267 (3 years, 6 months)
  • So it make sense to hold the position for about 529 days (about 18 months)