Dany Lang 
18 September 2019Here is the paper by Auvray&Rabinovich.
You can also look at Joel’s presentation of sept-19. Additionnal paper on the (false) financialization of non financial firms.
1 – Econometrics software : R
Download and install R and then Rstudio. Start the tutorial https://www.datacamp.com/courses/free-introduction-to-r. You can read this textbook. In French here.
2 – How to work on the long run
- Write a journal, a log (where the data comes from, what you do to it, issues you face and options you take, ideas, paste your plots)
- save and comment your scripts
- start writing your report from the first day
- start a bibliography
- make a copy of the whole folder on another computer or on the cloud, each time you work on it.
3 – Data : for october 15, 2020
Download the databases. Compustat database and Z1 database. Import them in R.
Run the following commands to download, install and call the tidyverse package:
install.packages("tidyverse")
library(tidyverse)
and start describing and plotting the main variables.
Documentation on Compustat is here
select variables
- description
- operating surplus, net
- consumption of fixed capital, structures, equipment, and intellectual property products, including equity REIT residential structures (NIPA basis)
- dividends paid
- equity and investment fund shares; liability
- gross fixed capital formation with equity REITs (IMA)
Figure 1
- From Z1, calculate
- net payouts = dividends paid – equity and investment fund shares
- investment = gross fixed capital formation
- From compustat, calculate
- Net payouts are prstkc [Purchase of Common and Prefered Stock] +dv [Cash Dividends (Cash Flow)]- sstk [Sale of Common and Preferred Stock]
- investment = capx
For 22 october : Figure 2 (except the means)
- from Z1, define US.economy=net.payouts/(operating.surplus+consumption.of.capital)
- from compustat, define US.listed.firms=net.payouts/(profit-taxes+interest+depreciation)
Figure 5
Download fig5.csv and make the scatter plots.
Here is a script that seems to be working. You can improve it and complete if your want to have figures even more similar to the original paper.
4 – Intermediate report
For November 21th, prepare a 5 to 8 pages report with
- an short explanation of the paper results
- a short presentation of the databases (compustat, Z1 and WIOD)
- choose a variable of one of these databases and study it from a quantification perspective
- replicate figures 1, 2 and 5 (only the scatter plots)
I don’t need your R script.
Here is an example of discussion on the quantification (part V, page 18+)
5 – Company database
Slides on panels and workshop script
Let’s use this framapad to share code and notes.
Try to replicate the baseline financialisation model with the reduced database, with a standard panel model. You should start with a few variables. Here is my script, build from the framapad.
Here is a cleaned database (update Jan 15) for the main models of the paper. Low and hich non-core-non-energy offshoring sectors are defined by comparing ncne with its median. The ncne value is the same for all the compagnies of a given sector.
If you want to get into the sectors, you might use the variable PRIMARY_SIC_CODE and need this function transforming the SIC codes into Fama-French industry categories.
If you want to try to start from the original database, you can find it here.
6 – Panel GMM
Goal of the full report : replicate the baseline model and the “financial and offshoring” model, with the pgmm function. p
If you get there, you can try to split the data according to the size of the firms as in the paper, change something else on your own initiative.
7 – Tutoring schedule
On Nov 26, will meeting at 2PM for a short lecture on sequence analysis, an heterodox way to use data very different from linear model end standard econometrics.
Group appointments on Dec 17 and individual appointments on January 7 and 14.
8 – Recommendations
