0
   

How to handle the missing values's issue of newly listed stocks?

 
 
Sima77
 
Reply Sat 1 May, 2021 11:26 am
I am trying to test some asset pricing models on 10 portfolios for the period of 2010-2020. The problem is that three of these portfolios included stocks that are newly listed in 2017 and 2018, so I only have three years of historical data for these stocks. Can anyone help how to solve this problem? I don't want to exclude these stocks; does imputation makes sense for such a long period?
 
Ragman
 
  1  
Reply Sat 1 May, 2021 11:38 am
@Sima77,
Might this be a homework assignment in a college-level economics class? If not, then perhaps pose this question to an economics professor or investment expert. I may be wrong but most likely that answer requires investment expertise far beyond 99.9% of an online forum such as ours.
engineer
 
  3  
Reply Sat 1 May, 2021 11:53 am
@Ragman,
I think it is more of a statistics question. You can't include these stocks. It would distort your analysis to include stocks with only a four or five year history into an analysis of 10 year performance. You could shorten the analysis for all stocks or do two analyses, one short period with all stocks and one long with some excluded.
Sima77
 
  1  
Reply Sat 1 May, 2021 05:49 pm
@engineer,
Thank you for your suggestions. I will consider them, but do you think using interpolation techniques to fill in the missing values is a bad idea in this case?
engineer
 
  3  
Reply Sat 1 May, 2021 07:22 pm
@Sima77,
Yes, very bad idea, especially with stock performance. You would not be interpolating, you would be extrapolating which produces very questionable results
Sima77
 
  1  
Reply Sun 2 May, 2021 09:32 am
@engineer,
Yes extrapolation, it is my bad, Sorry. I really appreciate your answers. I know I am asking a lot, but this is the last one: Actually I have more than 60% of the total data, because I have the historical data of the whole period for the majority of stocks, but only few of these stocks are newly listed and have short period of historical data. My question is if I shorten the period does extrapolation makes sense? lets say instead of 10 years I will do 6 years (so I will extrapolate 2 years) does it make sense? If yes, what is the longest period in which extrapolation is acceptable? Thanks in advance
engineer
 
  3  
Reply Sun 2 May, 2021 01:35 pm
@Sima77,
Honestly, if I saw a stock analysis where data from some stock was extrapolated, even one year, I would consider it invalid. The stock market has been on a tear since 2017 to 2020, but it was pretty flat in 2015 and 2016. If you extrapolated performance back to 2014 using the 2017 to 2020 data you would get a very large error. The purpose of looking at stock performance over time to to see how the portfolio responds in a variety of economic conditions. There should be no expectation that stock performance is consistent which is the assumption that goes into extrapolation. Maybe you could find stocks similar to the ones you want to use but with longer history to act as surrogates in your portfolio.

Look at the DJIA over 10 years.
https://www.5yearcharts.com/wp-content/uploads/2017/04/dow-jones-vs-cpi-10y.png
What line would you draw through this to extrapolate back? You could make this be whatever you wanted just by picking the time range that matched your desires. Nor could you say "I'll use a DJIA correction" because what you are testing is how the portfolios perform compared to each other. Saying they perform like the DJIA assumes the answer you are trying to find.
Sima77
 
  2  
Reply Mon 3 May, 2021 04:21 am
@engineer,
You can't imagine how much I appreciate your replies. Thank you so much.
0 Replies
 
 

Related Topics

Where is the US economy headed? - Discussion by au1929
Debt consolidation - Question by rogers
Shopping Around For Loans - Question by Brandon9000
What is greed? - Discussion by Robert Gentel
bonds series h - Question by allen russell
Naked Short Selling - Question by optimus cubed
HOW TO GET WEALTHY - Discussion by farmerman
 
  1. Forums
  2. » How to handle the missing values's issue of newly listed stocks?
Copyright © 2021 MadLab, LLC :: Terms of Service :: Privacy Policy :: Page generated in 0.03 seconds on 05/18/2021 at 05:05:02