New Caledonia’s nickel exports by @ellis2013nz | R bloggers

Just a quick blog post that basically lists some thoughts and resources on New Caledonian nickel exports.

This post was motivated by my (still incomplete) reading of this book from 2017 Large-scale mines and politics at local level – between New Caledonia and Papua New Guineaedited by Colin Filer and Pierre-Yves Le Meur. It’s fascinating, although clearly a bit dated now.

At the time that book was written, the portrayal of nickel mining in New Caledonia was more optimistic, both in purely economic terms and as a tool for Kanak reconciliation and independence, than it would be today in 2025. Without going into all the details, several mines have been mothballed or closed in recent years. An already shaky economic trend was exacerbated by civil unrest in New Caledonia in 2024. I wanted to quickly check: “Is nickel being exported at all? How much is it worth?” – and so quickly looked at the data.

Exports by value

My main data source for this is the ISEE of New Caledonia (Institute for Statistics and Economic Studies New Caledonia), which maintains a page about nickel in economics. This looks like a page from 2019, but the data downloads in Excel format at the bottom are updated monthly and are current through October 2025.

New Caledonia exports both ore and processed metal – there’s quite a story behind that – and the export markets differ slightly. For example, Australia imported ore, but the Queensland processing plant that took it closed. Here are the data on ore exports, by value and broken down by destination:

…and here are the comparable data on processed metal:

The following graph shows them both, now ignoring the destination:

None of these graphs have been adjusted for inflation. Beyond that, of course, what we see from all this in recent years is quite clear: a big increase in value exports in 2021 and 2022, a decline in 2023 and a collapse in 2024 and 2025.

Here is the code to download this data from the ISEE website and draw those graphs:

library(tidyverse)
library(readxl)
library(fredr)

# Nickel sector info from ISEE available at
# https://www.isee.nc/economie-entreprises/entreprises-secteurs-d-activites/secteur-du-nickel

download.file("https://www.isee.nc/component/phocadownload/category/147-consultez-les-donnees-historiques?download=676:les-exportations-de-nickel",
              destfile = "nc-nickel-exports.xls", mode = "wb")

#--------------------minerai---------------------

ore <- read_excel("nc-nickel-exports.xls", sheet = "Minerai mensuel", range = "A29:HG32",
                      col_names = c("destination", 
                                     as.character(seq.Date(from = "2008-01-15", to = "2025-10-20", by = "month")))) |> 
  gather(date, value, -destination) |> 
  mutate(value = as.numeric(value),
         date = as.Date(date),
         datem = format(date, "%b %Y"),
         datem = fct_reorder(datem, date),
         type = "Ore")

labs <- levels(ore$datem)
labs[!grepl("Oct", labs)] <- ""

ore |> 
  ggplot(aes(x = datem, y = value, fill = destination)) +
          geom_col() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1),
        panel.grid.minor.x = element_blank(),
       panel.grid.major.x = element_blank()) +
  scale_x_discrete(labels = labs) +
  scale_y_continuous(label = comma) +
  labs(x = "", y = "XPF (m)", fill = "Destination:",
       title = "Nickel ore exports from New Caledonia, 2008-2025")

#--------------------metallurgie------------------

metal <- read_excel("nc-nickel-exports.xls", sheet = "Métallurgie mensuel", range = "A134:HG138",
                      col_names = c("destination", 
                                     as.character(seq.Date(from = "2008-01-15", to = "2025-10-20", by = "month")))) |> 
  gather(date, value, -destination) |> 
  mutate(value = as.numeric(value),
         date = as.Date(date),
         datem = format(date, "%b %Y"),
         datem = fct_reorder(datem, date),
         type = "Processed metal")

metal |> 
  ggplot(aes(x = datem, y = value, fill = destination)) +
          geom_col() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1),
        panel.grid.minor.x = element_blank(),
       panel.grid.major.x = element_blank()) +
  scale_x_discrete(labels = labs) +
  scale_y_continuous(label = comma) +
  labs(x = "", y = "XPF (m)", fill = "Destination:",
       title = "Nickel metal exports from New Caledonia, 2008-2025")

       
#----------------------combined------------------

both <- rbind(metal, ore) |> 
  group_by(datem, type) |> 
  summarise(value = sum(value, na.rm = TRUE))

both |> 
  ggplot(aes(x = datem, y = value, fill = type)) +
          geom_col() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1),
        panel.grid.minor.x = element_blank(),
       panel.grid.major.x = element_blank()) +
  scale_x_discrete(labels = labs) +
  scale_y_continuous(label = comma) +
  labs(x = "", y = "XPF (m)", fill = "Type of export:",
       title = "Nickel exports from New Caledonia 2008-2025")

Seasonality

I was interested in some of the variability in these patterns and converted the total value of these exports into a single time series. I broke this down into trend, seasonal and random components and was interested to see that there is indeed a noticeable seasonal component:

The low months in the seasonal pattern are usually January, February and sometimes April. This is probably due to the heavier rainfall in those months, which has made production and transport somewhat more difficult. But this is just a guess, I’m open to hearing about what else could be causing this seasonality.

Here is the code for the time series decomposition:

nickel_ts <- both |> 
  group_by(datem) |> 
  summarise(value = sum(value)) |> 
  pull(value) |> 
  ts(frequency = 12, start = c(2008, 10))


par(bty = "l")
plot(stl(nickel_ts, s.window = 7), col = "steelblue",
      main = "Nickel exports (ore + metals, in millions of XPF) from New Caledonia")

Causes

Why have New Caledonia’s nickel exports collapsed so sharply? I’m not an expert in this area, but I think this is largely driven by global nickel prices; with the secondary factor being difficulties specific to New Caledonia (local politics such as those that are the subject of the book I started this post with; security, especially since the civil unrest of 2024, but actually dating back earlier; and high labor costs).

I won’t go into the New Caledonia-specific reasons here. But here are world prices, showing a clear increase from 2016 to a peak in 2022 and a rapid collapse between 2023 and 2025:

The structural price increase was related to the demand for nickel in batteries, including in electric vehicles. The 2022 spike came from a real supply crunch when Russia invaded Ukraine, and fears that Russian nickel would disappear from the market. Prices then fell as Indonesian supply picked up and the rise of nickel-free batteries for electric vehicles.

When prices fall, the value of exports naturally falls, even if volumes remain the same. But if the price falls due to more competition, the volume also falls. And the impact can be strong enough to close entire mines (as has been the case in New Caledonia).

Code for world nickel price data from the US Federal Reserve (FRED) data system:

#------------------nickel prices---
fredr_set_key(Sys.getenv("FRED_API_KEY")) # assumes was previously set with Sys.setenv(FRED_API_KEY = "XXXXXX")


nickel_prices <- fredr(
  series_id = "PNICKUSDM",
  observation_start = as.Date("2008-01-01"),
  observation_end   = Sys.Date()   # or a chosen end date
)

nickel_prices |> 
  ggplot(aes(x = date, y = value)) +
  geom_line(colour = "brown") +
  scale_y_continuous(label = dollar) +
  labs(x = "", y = "US dollars (m)",
        title = "World nickel prices",
        subtitle = "(Not adjusted for inflation)")

That’s all for today. There is a lot more to this nickel story and how crucial it is for New Caledonia, but I don’t have time to go into it today and I’m probably not the right person to do it anyway.

Happy new year!

Related