rm(list=ls())
setwd("C:/Users/harari/Desktop/Dropbox/enseignement/M2 EPOG econometrics/replication")
library(tidyverse)

#Compustat ------

compustat_original <- read_delim("compustat-1.csv",";", col_names = T,
                        locale = locale(decimal_mark = ","))

compustat<-compustat_original %>% #select(-X11) %>% 
  mutate(invest_netpayout_ratio=capx/(prstkc+dv-sstk),name="compustat")

head(compustat)


ggplot(compustat)+
  geom_line(aes(x=year,y=invest_netpayout_ratio),color="grey")

# Z1 ------
Z1_original <- read_delim("Z1.csv", ";", col_names = FALSE, skip = 5)
Z1_names <- read_delim("Z1.csv", ";", col_names = T) %>% slice(0)
names(Z1_original)<-names(Z1_names)
rm(Z1_names)

Z1<-Z1_original %>% select(year=1,operating.surplus=9, consumption.of.capital=3,
                  dividends.paid=18,equity.liability=67,
                  fixed.capital.formation=29) %>% 
        filter(year>1945) %>%
        mutate(invest_netpayout_ratio=fixed.capital.formation/
                 (dividends.paid-equity.liability),name="Z1")

Z1<-Z1 %>% select(year,invest_netpayout_ratio,name)
compustat<-compustat %>% select(year,invest_netpayout_ratio,name)
merged_db<-bind_rows(compustat,Z1)
tail(merged_db)


ggplot(merged_db,aes(x=year,y=invest_netpayout_ratio))+
ggtitle("Investment as a ratio of net financial payouts for the US economy and US-listed firms, and
offshoring intensity, 1946–2016")+
  geom_line(aes(colour=factor(name)))+
coord_cartesian(xlim = c(1945, 2016),ylim=c(0, 10))+
  scale_x_continuous(breaks = seq(1945, 2016, by = 10))+
  theme(plot.title=element_text(hjust=0.5),
        legend.position="right",
        panel.background =element_rect(fill="white"))


#+scale_y_continuous(sec.axis = sec_axis(~ . + 10)) for a secondary axis


# Fig 2 ----------
compustat.2<-compustat %>% filter(year>1970) %>% 
  rename(profit=pi,taxes=txt,depreciation=dp,interest=Xint) %>%
  mutate(net.payouts=prstkc+dv-sstk,
    US.listed.firms=net.payouts/(profit-taxes+interest+depreciation),
         decade=floor(year/10)*10,
          decade2=case_when(year>=1970&year<1980~1970,
                            year>=1980&year<1990~1980,
                            year>=1990&year<2000~1990,
                            year>=2000&year<2010~2000,
                            year>=2010&year<2020~2010)) %>% 
  group_by(decade) %>% mutate(average=mean(US.listed.firms)) %>%
  ungroup()


Z1.2<-Z1 %>% filter(year>1970) %>%
  mutate(net.payouts=dividends.paid-equity.liability,
         US.economy=net.payouts/
           (operating.surplus+consumption.of.capital))

ggplot(data=Z1.2,aes(x=year,y=US.economy,colour="US economy"))+geom_line()+
  geom_line(data=compustat.2,aes(x=year,y=US.listed.firms,colour="US listed firms"))+
  geom_point(data=compustat.2,aes(x=year,y=average, colour="Decade average"))+
  ylab(label="")+ scale_colour_manual(limits=c("US economy","US listed firms","Decade average"),values=c("red","blue","blue"))

#fig 5 ------
fig5 <- read.csv("./fig5.csv", sep=";")
head(fig5)

install.packages("gridExtra")
library(gridExtra)
names(fig5)<-c("isic","Industry","ncne_offshoring_mean","ptir_mean","ptir_p50","ptir_p75")
p1<-ggplot(data=fig5,aes(x=ptir_p50,y=ncne_offshoring_mean))+geom_point()+geom_smooth(method="lm")
p2<-ggplot(data=fig5,aes(x=ptir_p75,y=ncne_offshoring_mean))+geom_point()+geom_smooth(method="lm")
grid.arrange(p1,p2,ncol=2,nrow=1)