Fertilizer Price Volatility and its Impacts on Producers
AgMRC Renewable Energy Newsletter
October 2009
Mike Boland
Director
Arthur Capper Cooperative Center
Kansas State University
mboland@ksu.edu
In the past few years there has been a great amount of volatility in agricultural commodity prices. This has caused a great deal of stress within the agricultural input industry. A great deal has been said about upheavals in the ethanol industry, but less has been written about the need for financing requirements caused by the frequent margin calls and need for capital. The fertilizer industry holds no exception and had historically high prices in the early fall of 2008 followed by a large decline almost immediately afterwards. This article looks at the components of the fertilizer industry, its major players and impacts to producers.
Phosphates are an important fertilizer input used in production agriculture. These include monoammonium phosphate (MAP) and diammonium phosphate (DAP) products produced by Mosaic and PCS. Almost 35% of this is exported from the U.S. into India and Latin America. The U.S. imports phosphate rock from Morocco to use with existing supplies in Florida to create these products. Potash is another form of phosphorus and the two large firms in that industry are PCS and Mosaic. The U.S. imports almost 80% of its potash from Canada. Phosphate plants are built to utilize imported and domestic supplies of phosphate rocks, ammonia and sulfur (Florida and Texas coast).
Nitrogen includes ammonia, urea and urea and ammonium nitrate solution (UAN). Ammonia is manufactured by Mosaic, PCS, Agrium, CF, Terra and Koch. Almost 50 percent is imported through Trinidad and Tobago in the Caribbean which are close to a supply of natural gas. Urea is manufactured by CF, Agrium, Terra, Mosaic and PCS and 70 percent is imported from Canada, Eastern Europe, China and Venezuela. Finally, UAN is manufactured by Terra, CF, PCS and Koch and almost 30 percent is imported from Canada, Russia and Eastern Europe. Nitrogen plants are built near natural gas supplies (near Gulf ports) and transported to consumption points (corn, wheat and other crops). The price of energy is a major source of price volatility.
The key inputs in fertilizer manufacturing are globally traded products and are affected by transportation costs, exchange rates, policy decisions and other uncontrollable factors. They are also increasingly becoming concentrated. Agricultural retailers have little or no negotiating ability with wholesalers who are pushing risk back onto the retailer. Vertically integrated operations have advantages in transfer pricing.
The number of firms continues to decline nationally according to USDA. USDA data suggests that market share of cooperatives is increasing in retail agribusiness industry (agronomy, energy, feed, grain) for farmer-owned cooperatives (CHS, MFA, TN Farmers, Growmark). Non co-ops include Agrium (Crop Production Services-east coast, Eastern Corn Belt, Southwestern Kansas), Simplot (Upper Midwest and Pacific NW), Helena Chemical (national), and Wilbur Ellis (Western US, ND, MI, TX). CPS is a product of three mergers.
Because of recent price volatility, some large producers are trying to go direct to the wholesaler. However, barriers to entry exist in services (cost of application equipment, labor, cost of compliance), facilities (regulations, volume) and learning curve (not an industry with significant entry and exit patterns). New entrants in retail agronomy operations are not likely to surface except through existing firms who are increasing their footprint and trying to obtain economies of size.
Purdue University data suggest that there are two typical distribution channels to producers. The first is a wholesaler-retailer-producer channel that is not as transaction oriented, but more attuned to timeliness of service and individual solutions. The second is a wholesaler-distributer-producer channel where producers require simpler solutions and are more transaction-oriented, which puts more pressure on margins. This channel includes distributers who do not provide services, but simply aggregate large volumes of inputs in a warehouse or similar facility. A wholesaler-producer channel exists for some large producers.
Regardless, fertilizer application is intensely seasonal with short periods in the fall and spring and producer decision-making is often close to planting time. Retailers must buy six to nine months in advance with no opportunity to hedge. Producers desire a local solution which is often decided upon in a short window of time. Retailers must anticipate demand six to nine months ahead.
Price discovery occurs in this industry with a transaction occurring six to nine months prior to a producer actually buying and using the product. The retail fertilizer dealer has some negotiating ability (based on volume and product formulation) and has more than one buyer to select from. Various price sheets exist from industry associations but no information on volumes is given for those prices. Exchange rate risk has increased in the last five years due to weakening of the U.S. dollar, which made imports more expensive and increased volatility in energy prices. There is no way for a retail dealer to hedge the transaction. The lender requires collateral for the purchase and when inventory values change, the “margin” requirements change.
Figure 1 shows the changes in fertilizer in the past five years. For managing risk in this industry a retailer would like to find a way to hedge fertilizer but no futures market exists. Forward contracting at some price is difficult because the key input in manufacturing is extremely volatile due to factors beyond the control of the wholesaler and unlikely to devise a “fair” contract and significant premium for basis risk would be included. Consequently, many firms take a position and then aggressively try to sell fertilizer as quickly as possible using prepays.
Three models were used to analyze the risk in this industry. The first was a hedging weekly model (not available) using weekly wholesale fertilizer prices and CBOT (corn) and KBT (wheat) futures. The second model is a base model of buying fertilizer during planting season and selling grain at harvest which is widely used by many producers as it provides flexibility and easily understood. The third is the dollar cost averaging model where a producer pays for 1/12 of the fertilizer used and sells 1/12 of the grain each month. This is a proxy to hedging weekly.
The dollar cost averaging model starts with a producer determining an amount of grain and fertilizer needed for that grain that you want in that program. This is placed into a “margin” account which is needed by the retailer because the fertilizer is collateralized with their lender and the collateral is lost once it is applied by the producer. Each month the producer pays 1/12 of the fertilizer costs and receives 1/12 of the grain revenue. The price of fertilizer that day is multiplied by 1/12 of the fertilizer volume applied and the price of grain that day multiplied by 1/12 of the expected grain volume. An example is provided for a Southwest Kansas irrigated corn operation. Three agronomy options were considered: custom dry (urea, MAP), custom liquid (UAN, MAP) and NH3 + dry spread (anhydrous ammonia, MAP). An expected return per acre above each fertilizer costs for each agronomy option was calculated for the 2003 to 2008 time period using nearby corn price, Garden City cash price, an estimated yield of 200 bushels per acre, and Southern Plains fertilizer prices for ammonia, UAN, urea, and MAP (no potash was recommended in these three agronomy options) adjusted for delivery into Garden City. Government program payments were not included. The dollar cost averaging model assumes you start in first week in March and proceed over the year while the base model assumes you buy fertilizer in first week in March and sell grain in mid-October. There is little difference in return above fertilizer costs between the three agronomy options and only the NH3 + dry spread option is presented in Figure 2.
As you would suspect, the weekly hedging model and dollar cost averaging model are closer to one another. The traditional model has a lower mean and less risk as measured by the CV. None of this is surprising. The dollar cost averaging model is patterned after the weekly hedging model. There is no statistical difference among the three models. The mean deviations are greater because of the great increase in volatility especially during periods outside the March and October time periods. The weekly hedging model has the most variability and most profit.
The dollar cost averaging mimics some of the patterns one sees in the weekly hedging model.
