After a small test run in 2017, Chile launched its green bonds in 2019, but has seen a decrease in green bonds issuances since.
There is a correlation between green bond issuances and environmental protection expenditure.
There is no correlation between green bond issuances and renewable energy prouction, but there may be insufficient data.
Green bonds are issued debt with the obligation of being used for climate or environmental purposes (1). Green bonds, first issued in 2007, are seen as revolutionary because they allow bond purchasers to have a choice on where their money goes for the first time. In Latin America, Chile was the first country to issue green bonds, as its political and geographic circumstances made it ripe with opportunities for climate and environmentally friendly projects. Since then, Chile’s green bond issuance and framework have been widely praised (2). In spite of this, green bond critics argue that green bonds do not add additionality or impact to a countries climate projects, as the money that they are bringing in from green bonds would have been used on climate projects anyways if regular bonds were issued instead (3). This research looks at the available data and determines if green bonds have any impact on the additionality of renewable energy and environmental protection expenditure in Chile.
Chile’s ramped up green bond issuances with its new green bond framework 2019 and subsequently saw a deceleration in green bond issuances in following years. Figure 1 shows that after a smaller $500 milllion USD green bond issuance in 2017, Chile increased green bond issuances to $5 billion USD in 2019. Since then, green bonds issued in 2020 and 2021 fell to $3 billion USD and issuances in 2022 plumetted to under $1 billion USD. The decrease in green bond issuances in spite of international approval could be due to a number of different reasons. On one hand, Chile could be responding to additionality criticisms of green bonds. It has since issued the more impact focused sustainability-linked bonds, which could reflect a shift in the purpose in its bonds and a signal to lenders that it is serious about its climate and sustainability goals (4). On the other hand Chile could just be scaling back its issued debt in general. In recent years, Chile’s fiscal deficit has risen, so issuing less bonds in general could be its way of balancing its budget (5).
Code
#| error: false#| output: false#| warning: false#| include: false#| message: falselibrary(tidyverse) library(janitor)library(here)library(readxl)library(tidyr)library(dplyr)library(rnaturalearth)library(countrycode)library(wbstats)library(ggplot2)library(gganimate)#install.packages('gganimate')library(gapminder)#install.packages('gifski')library(gifski)#install.packages("ggthemes")library(ggthemes)#install.packages("plotly")library(plotly)renewable_energy <-read_csv("00_data_raw/Renewable Energy by year and country.csv")renewable_energy <- renewable_energy[,!names(renewable_energy) %in%c("INDICATOR", "SUBJECT", "MEASURE", "FREQUENCY", "...8")]renewable_energy <- renewable_energy %>%rename("year"="TIME", "energy_ktoe"="Value","iso3c"="LOCATION")#Create iso3c index function.iso3c_to_country_name <-function(iso3c) { iso3c |>countrycode(origin ="iso3c", destination ="country.name")}country_name_regex_to_iso3c <-function(country_name) { country_name |>countrycode(origin ="country.name", destination ="iso3c",origin_regex =TRUE)}renewable_energy <- renewable_energy %>%mutate(country_name =iso3c_to_country_name(iso3c)) renewable_energy <- renewable_energy %>%na.omit()Environmental_Protection_Expenditures <-read_csv("00_data_raw/Environmental_Protection_Expenditures.csv")Environmental_Protection_Expenditures <- Environmental_Protection_Expenditures[,!names(Environmental_Protection_Expenditures) %in%c("ISO2", "Source", "CTS_Code", "CTS_Name", "CTS_Full_Descriptor")]Environmental_Protection_Expenditures <- Environmental_Protection_Expenditures %>%pivot_longer("F1995":"F2021",names_to ="year",values_to ="percent_of_gdp" )Environmental_Protection_Expenditures <- Environmental_Protection_Expenditures %>%na.omit()Environmental_Protection_Expenditures <-subset(Environmental_Protection_Expenditures, Unit!="Domestic Currency")Environmental_Protection_Expenditures <- Environmental_Protection_Expenditures[,!names(Environmental_Protection_Expenditures) %in%c("Unit")]Environmental_Protection_Expenditures <- Environmental_Protection_Expenditures |>pivot_wider(names_from ="Indicator",values_from ="percent_of_gdp")Environmental_Protection_Expenditures <- Environmental_Protection_Expenditures |>rowwise() |>mutate(total_percent_gdp =sum(`Expenditure on biodiversity & landscape protection`, `Expenditure on environment protection`, `Expenditure on environmental protection n.e.c.`, `Expenditure on environmental protection R&D`, `Expenditure on pollution abatement`, `Expenditure on waste management`, `Expenditure on waste water management`, na.rm =TRUE)) Environmental_Protection_Expenditures <- Environmental_Protection_Expenditures |>mutate(iso3c =country_name_regex_to_iso3c(Country)) |>mutate(country_name =iso3c_to_country_name(iso3c))Environmental_Protection_Expenditures <- Environmental_Protection_Expenditures[,!names(Environmental_Protection_Expenditures) %in%c("Country", "ISO3")]Environmental_Protection_Expenditures$year<-gsub("F","",as.character(Environmental_Protection_Expenditures$year))Environmental_Protection_Expenditures$year <-as.numeric(as.character(Environmental_Protection_Expenditures$year)) depedent_variables <-full_join(renewable_energy, Environmental_Protection_Expenditures, by =c("year"="year", "country_name"="country_name", "iso3c"="iso3c"))green_bonds <-read_excel("00_data_raw/Green Bonds Database.xlsx", sheet ="Sheet2")green_bonds <- green_bonds %>%pivot_longer("France":"Singapore",names_to ="country_name",values_to ="issuance_billion_usd" )country_name_regex_to_iso3c <-function(country_name) { country_name |>countrycode(origin ="country.name", destination ="iso3c",origin_regex =TRUE)}green_bonds <- green_bonds %>%mutate(iso3c =country_name_regex_to_iso3c(country_name)) |>mutate(country_name =iso3c_to_country_name(iso3c)) clean_dataset <-full_join(depedent_variables, green_bonds, by =c("year"="year", "country_name"="country_name", "iso3c"="iso3c"))dataset_countries <-subset(clean_dataset, country_name %in%c("France", "Germany", "Italy", "Netherlands", "Spain", "United Kingdom", "United States", "Brazil", "Chile", "China", "India", "Mexico", "Turkey", "United Arab Emirates", "Hong Kong SAR China", "Singapore"))Chile <-subset(dataset_countries, country_name %in%c("Chile"))Chile %>%filter(year >2012) %>%ggplot(aes(x = year, y = issuance_billion_usd)) +geom_line(color ="green", na.rm =TRUE) +labs(title ="Figure 1: Green Bond Issuance in Chile by Year (2013-2022)",subtitle =str_wrap("Chile issued a significant amount of green bonds starting in 2019.", width =80),x ="Year",y ="Bond Issuance (billions of USD)", caption ="Source: Sustainable Debt Monitor Database | Latest Data: 2022 | Calculations by Ryan Showman" ) +scale_y_continuous(labels = scales::label_dollar(suffix ="B"), expand =c(0,0)) +scale_x_continuous(breaks =~axisTicks(., log =FALSE)) +theme_minimal()
Environmental protection expenditure in Chile has consistently risen since 2019. As shown in Figure 2, Chile increased in expenditure on environemental, biodiversity, and landscape protection in 2019. While, expenditure on environmental protection fell in subsequent years, expenditure growths in biodiversity and landscape protection more than made up for it and resulted in the largest environmental protection expenditure to date in 2020. Interestingly, these increases in environmental protection expenditure do align with the beginning of Chile issuing green bonds on a larger scale. This could provide evidence that when it comes to environmental protection expenditure, Chile’s green bond issuances did in fact provide additionality. However, this could be correlation rather than causation. The increase in both green bond issuances and environmental protection expenditure could just mean that Chile was changing its national priorities to greater value the environment and climate at that time. In spite of this, the correlation between the bond issuances and environmental expenditure does contribute interesting data to the debate on green bond additionality.
Rows: 1523 Columns: 36
-- Column specification --------------------------------------------------------
Delimiter: ","
chr (9): Country, ISO2, ISO3, Indicator, Source, CTS_Code, CTS_Name, CTS_Fu...
dbl (27): F1995, F1996, F1997, F1998, F1999, F2000, F2001, F2002, F2003, F20...
i Use `spec()` to retrieve the full column specification for this data.
i Specify the column types or set `show_col_types = FALSE` to quiet this message.
Code
Chile2 <- Chile2[,!names(Chile2) %in%c("ISO2", "Source", "CTS_Code", "CTS_Name", "CTS_Full_Descriptor")]Chile2 <- Chile2 %>%filter(Country =="Chile")Chile2 <- Chile2 %>%pivot_longer("F1995":"F2021",names_to ="year",values_to ="percent_of_gdp" )Chile2 <- Chile2 %>%na.omit()Chile2 <-subset(Chile2, Unit!="Domestic Currency")Chile2 <- Chile2 |>mutate(iso3c =country_name_regex_to_iso3c(Country)) |>mutate(country_name =iso3c_to_country_name(iso3c))Chile2 <- Chile2[,!names(Chile2) %in%c("Country", "ISO3")]Chile2$year<-gsub("F","",as.character(Chile2$year))Chile2$year <-as.numeric(as.character(Chile2$year)) Chile2_graph <- Chile2 %>%filter(Indicator %in%c("Expenditure on biodiversity & landscape protection", "Expenditure on environment protection", "Expenditure on environmental protection n.e.c.", "Expenditure on pollution abatement")) %>%filter(year >2012) %>%ggplot(aes(x = year, y = percent_of_gdp/100, fill = Indicator)) +geom_area() +labs(title ="Figure 2: Environmental Protection Expenditure\nin Chile by Year (2013-2020)",subtitle =str_wrap("Environmental protection expenditure saw a rise from 2019 anwards.", width =70),x ="Year",y ="Environmental Protection Expenditure\nPercent of GDP", caption ="Source: IMF | Latest Data: 2020 | Calculations by Ryan Showman" ) +scale_y_continuous(labels = scales::percent) +theme_classic()Chile2_graph <-ggplotly(Chile2_graph, tooltip="text")Chile2_graph
Renewable energy produced in Chile has gradually increased in recent year but sharply decreased after 2020. Chile’s sunny north and windy south has provided it with a geography strongly favoring renewable energy production. This can be reflected in Figure 3, demonstrating an increase in renewable energy used in Chile over time. However, renewable energy production peaked in 2013 and dramatically decreased after 2020. The renewable energy peak in 2013 and decline after 2020 demonstrate that green bonds do not seem to be adding significant additionality to renewable energy in Chile. However, there are three problems with this conclusion. First, it takes a time for renewable energy to be built so any additionality added by green bonds may not yet appear. Second, Chile is has historically gotten much of it renewable energy from hydro power, so any changes in renewable energy trends could just reflect changes in water availability for hydropower (6.). Finally, these changes could also reflect changes in energy demand. Fall of renewable energy production after 2020 could reflect a decrease in energy demand during the demand. Until there is more data, any insight on the additionality of green bonds to renewable energy production is inconclusive.
Code
#| error: false#| warning: falseChile %>%filter(year >1999) %>%ggplot(aes(x = year, y = energy_ktoe)) +geom_line(color ="green3", na.rm =TRUE) +labs(title ="Figure 3: New Renewable Energy Produced in Chile per Year (2000-2021)",subtitle =str_wrap("Renewable energy used in Chile sharply decreased after 2020.", width =80),x ="Year",y ="Renewable Energy Produced (kiloton of oil equivalent).", caption ="Source: OECD | Latest Data: 2021 | Calculations by Ryan Showman" ) +scale_x_continuous(breaks =~axisTicks(., log =FALSE)) +scale_y_continuous(expand =c(0, 0), limits =c(0, 12500)) +theme_minimal()
The data analysis provides mixed answers as to whether or not Chile’s green bond issuances included additionality. While there is some evidence that green bond issuances resulted in an increase in environmental protection expenditure, there is no evidence that green bond issuances had an impacted on produced renewable energy. More and up-to-date data will be needed to provide a more econometrically rigorous analysis on Chilean green bonds and additionallity.
