“Food Supply Chains Are Buckling as World Runs Short of Workers”
Bloomberg, September 2, 2021
“Ruined Brazil Harvest Sparks Food Inflation Everywhere”
Bloomberg, September 28, 2021
First appeared in our Q3 2021 commentary
The world is slowly slipping into an agricultural crisis. Demand for grain remains strong, while certain supply trends that have been in place for 40 years are about to reverse. Reflecting these trends, grain prices have surged over the last 12 months. Although copper captured investors’ attention this spring as it made a new all-time high, corn and soy came within 5% of their 2012 drought-related highs as well.
Between 1980 and 2000, global grain consumption grew by 1.3% per year. Since then, driven by emerging market demand, grain consumption growth has ratcheted up to 2.3% annually – an acceleration of 75%.
As an emerging market gets richer, demand for most commodities rises rapidly, be it oil, natural gas, grain, etc. No market is more prone to this phenomenon than agriculture. The desire to consume more protein and less starch rises rapidly as a country progresses from poor to middle income to wealthy.
In our next quarterly commentary, we will discuss the relationship between historical protein consumption and wealth going back 100 years – including in Japan through the 1960s when per capita GDP was less than $500.
Also in our next quarterly commentary, we will discuss various weather issues affecting the 2022-2023 crop year. Although most journalists and investment professionals believe global warming has negatively impacted global growing conditions, our analysis suggests the opposite. We believe that over the last 40 years, warming trends have produced consistently better and better growing conditions as seasons have lengthened and crop-killing frosts have diminished. Now that we are likely entering a global cooling phase due to decreased sunspot activity, growing conditions may become more challenged. These conditions may already be at work.
Note: If you are not already signed up to be notified when we release our quarterly commentary, please register here.
Grain prices were weak in Q3, driven mostly by technical trading and less by any significant change to the underlying fundamentals. Corn fell 25% during the quarter while soybeans fell 13% and wheat advanced 8%.
Spot corn peaked at $7.75 per bushel back in May 2021, only 5% below its all-time high, set during a severe North American drought in 2012.
In the May 2021 World Agricultural Supply Demand Estimate (WASDE) report, the United States Department of Agriculture (USDA) projected corn 2021 ending stocks (inventories just before the current harvest begins) of only 1.1 bn bushels. Although not quite a record low, inventories were projected to have fallen low enough to put severe upward pressure on price. As summer progressed, corn prices eased as the new corn crop began to be harvested in July and inventories began to rebuild. The same phenomenon was at work in a very tight soybean market. Soybean prices also peaked in May and began to ease as summer progressed and the new harvest began in September.
Although grain prices pulled back during Q3, 2022 is set for another spike higher. In our last commentary we wrote about our recent development of an artificial neural network to help predict corn yields (with a soybean neural network to follow).
Estimating crop yields is both critical in predicting grain production and fiendishly difficult to get correct. Given our success in building a neural network to deal with difficult questions in shale productivity, we decided to try the same approach to corn yields.
We set out to develop a neural network that would explicitly incorporate the vagaries of weather. This model would start out making a wide prediction that would narrow over time as actual weather was updated throughout the growing season. Our estimate for crop yields would evolve in real time.
In our last quarterly commentary, we outlined our initial prediction for 2021 corn yields, but warned that further updating was needed as more weather data came in.
We wanted to make sure our models were not “over-fitting” the data -- coming up with spurious relationships instead of finding true causal links.
Our first attempt at predicting the 2021 US crop yield turned out to be quite interesting. Because of near drought conditions across large swaths of the corn belt, we strongly believed the USDA’s 2021 record yield estimate of 179.5 bushels per acre would disappoint. Based upon crop conditions that existed in July, our neural network projected corn yields would fall between 169 (the 2019 season) and 172 (the 2020 season) bushels per acre – well below the USDA estimates.
No sooner than we had published our letter, the USDA slashed their yield estimate to 174.6 in their August WASDE report causing prices to rally. The USDA was forced to recognize the very challenging growing conditions in Iowa, Minnesota and South Dakota. As a result, the USDA lowered the 2021 corn harvest estimate by almost 425 mm bushels.
Offsetting this revision, the USDA also reduced domestic corn usage and export demand by almost 200 mm bushels. By the end of August, the USDA forecast the 2021-2022 corn ending stocks would only reach 1.24 bn bushels -- only marginally higher than last year’s extremely low levels.
In their last two monthly reports (September and October), the USDA has increased yield assumption higher back to 176.5m, which has raised the 2021 crop size up by 270 mm bushels and brought back the 2021-2022 ending stock estimate to 1.5 bn bushels.
Is this 176.5 bushels per acre estimate anywhere near a realistic corn yield for the 2021 corn crop? Because of the extreme dry conditions that existed in large swaths of the northern and western corn belts, we believe this year’s estimate of 176.5, which we should point out ties the record corn yield of both 2018 and 2019, is too high. Turning to our neural network, it too is telling us that the USDA is again vastly overestimating yields. Based on crop growing conditions up to mid-October, our neural networks say that we should expect corn yields to come in between 172 and 174 bushel per acre, below the USDA estimate. If corn yields were to fall to these levels, this would drop the 2021-2022 corn ending stock to dangerously low levels. Any weather related problems that emerge in next year’s harvest would send corn prices skyrocketing.
In soybeans, the situation is not as tight, but similar to corn: any disappointment in the 2022 crop could put extreme upward pressure on price. The USDA is expecting soybean yields to reach a near record 51.5 bushels per acre. Even with near record yields, soybean ending stocks are only expected to rise from this year’s 185 mm bushels (extremely low) to 320 mm bushels (near average) next year.
Most of the soybean regions experienced much better growing conditions than corn, so we do not expect a large yield disappointment. That is ultimately a good thing since any disappointment would throw the soybean market into turmoil given extremely low carry- over inventories from last year. Once our soybean neural network is complete, we will be able to make much more accurate predictions for soy as well as corn yields.
Strong demand, low inventories and the rising probability of disruptive weather patterns all leave agriculture markets primed for a sharp rally.
The rapidly evolving energy crisis has spilled over into global fertilizer markets, with bullish consequences. We have maintained significant exposure to fertilizer producers, and will continue to do so. Fertilizers have been the best performing commodities over the last nine months. Urea (solid nitrogen) doubled, while phosphate is up 70% and potash (the fertilizer every analyst loves to hate) is up 160%.
Fertilizer production (especially nitrogen and phosphate) is very energy intensive. China has already restricted production of both fertilizers to help conserve energy and has announced export restrictions. The ramifications are huge: China produces half of global urea and nearly 60% of global phosphate-based fertilizers making it the world’s largest exporter.
In the near term, these restrictions will put huge upward pressure on fertilizer prices. Over the medium term, less available fertilizer will negatively impact crop yields leading to higher grain prices as well.
Over the last 30 years, global crop yields have surged, driven in part by much greater availability of fertilizers. If fertilizer supply falls and prices rise, there will be a clear negative impact on crop yields as farmers are forced to reduce applications.
The global agricultural crisis has now entered its first phase. We continue to recommend investors maintain significant exposure to agricultural-related equities, with a special emphasis on fertilizer producers.
Intrigued? To read more on this subject, we encourage you to download the full commentary, available below.