Do El Niño and La Niña years move commodity prices?

Across 38 benchmark prices and 64 years (1961 to 2024), the median real price change was −3.7% in El Niño years and +3.5% in La Niña years. That gap is larger than chance usually produces (p = 0.006), but it is nearly as large for energy, metals and other non-crops (−3.3% against +2.8%) as for crops (−4.8% against +5.0%), so the data do not point to weather as the cause.

Method: each calendar year is labelled from NOAA's Oceanic Niño Index. Calendar-year mean of the twelve three-month ONI values: +0.5 or higher El Niño, -0.5 or lower La Niña, otherwise neutral. Only complete years. That gives 14 El Niño, 15 La Niña and 35 neutral years. For every price, the real change is the year's mean price, adjusted by US inflation, against the year before. The comparison is the median change in El Niño years minus the median in La Niña years. Chance benchmark: the labels were shuffled across the years 2,000 times (keeping how many of each there are) and the same gap recomputed. Shuffling years, not individual prices, matters because every price shares the same years and the prices are not independent of each other. This is a simple annual label, not NOAA's list of episodes, which can straddle two calendar years.

Median real price change by phase

Figures in brackets are the number of price-years. The "following year" rows ask whether last year's phase shows up in this year's price change, since a crop's harvest and the price reaction can fall in different calendar years.

What is driving it

Of the 8 years with the largest median real price rises across all prices, 6 were La Niña years (2021 La Niña, 1974 La Niña, 2010 neutral, 1973 La Niña, 2008 La Niña, 2011 La Niña, 2004 neutral, 2007 La Niña). Of the 8 largest falls, 2 were La Niña, 3 El Niño and 3 neutral (2012 neutral, 1985 La Niña, 1975 La Niña, 2015 El Niño, 2023 El Niño, 1981 neutral, 1982 El Niño, 2009 neutral). A handful of broad price booms that happened to fall in La Niña years (the table below lists every year) account for a large share of the gap, and they raised energy and metals as much as crops. That is why the finding is better read as "La Niña years have coincided with commodity booms in this sample" than as "La Niña raises prices".

One price at a time

For each price with at least 30 years of data and five El Niño and five La Niña years, the same test was run on that price alone. 10 of 38 are below 0.05, against about 1.9 expected if there were no link at all. The ones below 0.05 are not only crops: crude oil, phosphate rock, coal, urea, potash, phosphate fertilizers are not, which again points at the shared cycle. Click a heading to sort.

Every year

None of this says El Niño and La Niña never affect crops: individual harvests in particular regions clearly respond to them, as the El Niño and La Niña explainer describes. It says that in annual, world-benchmark price changes since 1961, that effect cannot be separated from the broader price cycle with this data. Correlation is not causation, and with only 14 El Niño and 15 La Niña years a few unusual years carry a lot of weight.

Prices: World Bank Pink Sheet primary benchmark series, annual mean of twelve months, deflated by the US consumer price index (FRED CPIAUCSL). Climate: NOAA Climate Prediction Center, Oceanic Nino Index (public domain). Years with a full set of twelve monthly prices only. Chance benchmark: 2,000 label shuffles, seeded, so the page is the same on every build.