Hi everyone,
I am planning on constructing a Fama french 3 factor model for a period from 1.1.1998-31.12.2015 for a portfolio of about 120 stocks. I have collected the monthly returns for each stock over 36 months since their IPO. The process of doing a Fama french 3 factor model for a single stock is very straight forward as seen in this video: https://www.youtube.com/watch?v=b2bO23z7cwg
However, how should I proceed with a portfolio with returns that all have different starting dates (as each firms have a different IPO date)?
My tough was as follows:
Many papers, such as the one by Levis (The Performance of Private
Equity-Backed IPOs), have used the Fama French 3 factor model but do not explain the mechanics behind the process.Any help is more than appreciated.
Any help is greatly appreciated
-Sebastian
would agree that the portfolio is treated as, and becomes, a single security. you can calculate price and returns data for the portfolio, just as you would for a single security.
easiest is to start from the position of calculating the portfolio's price data
Sebatian, I agree with Brian Barnard. You would average the readings for all available stocks for each month. As new stocks entered or old stocks left, you would obviously include or exclude those in a given month. You would then subtract the risk free rate each month from the average.
I have three published articles using a modified 3 factor model. I use different 3 factors because I was showing other causal inputs. I agree with the video as I used the Excel regression add in as well on these papers. Regards, Sam
Article Are Oil Industry Mergers Becoming Less Profitable?
Article In Asia, Does Airline Service Quality Matter to Investors?
Article With Increased Market Power, Do Aircraft Industry Stock Retu...
Dear Samuel D. Barrows and Brian Barnard thanks a lot for your input, much appreciate it!
Dear Sebastian
Forgive me if you looked at this website but French provides some data (you could use for verification):
http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html
Also, Wharton's wrds database has information on the FF 3 factor model:
https://wrds-www.wharton.upenn.edu/pages/support/research-wrds/research-guides/fama-french-research-portfolios-and-factors/
Dear Mr. Frisch,
I donot know whether Profs. Fama and French will agree with me on this but as we have empricially established equilibrium in an experimental sampling of Emerging Indias Stock Markets by developing the Millenium Mathematics Solutions (Mallick, Hamburger & Mallick (2016, 2017, 2018)) using endogenous portfolios and establishing state-preference approach Capital Assets Pricing Equilibnrium that by systems classification and systems integration of the financial systems data (Mallick (2012, 2014)) that there is a three dimensional factorisation of the four-dimensional market field and hence Econophysically using String Matching Field Theory Mallick et. al. as above. that the trading actions outcome of the Electronic Stock Market System also satisfies the Three factor model with statistically fair degrees of accuracy and quantum mechanicall financial*economic*financial accuracy. But ofcourse the Millenium solutions are necessary for this financial system State Preference Approach Asset Market Equilibrium with Fama-French Three factors satisfying Haag Theory for the protocols and even bypassing with "vector diffeotopy of embedding jump quantum conductance" mechanics (Mallick et. al..
S.K.Mallick
for S.K. Mallick, S.Raychaudhury, S.Mallick and others,
The RHMHM School,
USA, Japan, India
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression