Forecasting Industrial Wood Pellet Prices

FutureMetrics has developed a comprehensive model for estimating the cost to produce and ship industrial wood pellets to a foreign port.
By William Strauss | September 24, 2018

FutureMetrics has developed a comprehensive model for estimating the cost to produce and ship industrial wood pellets to a foreign port. One of the significant drivers of the cost estimate is the cost of the diesel fuel used in the harvest and transport of forest products and the transport of pellets to the shipping terminal. Using forecasts for petroleum prices and known or expected values for other components of the supply chain for pellets, the model can be used to estimate the future costs to produce and deliver pellets.

Delivered Pellet Historical Prices
FutureMetrics has estimated the historical delivered (CIF) price of wood pellets from international trade data. Data on the value and quantity of imported pellets gives insight into the prices that buyers are paying for the fuel. Because most of the international trade of pellets is under long-term contracts, those prices reflect a competitive, long-run, market-clearing value.

The spot price is influenced by short-term supply and demand imbalances.  If the market is in a state of excess supply, prices fall. If the market in a state of excess demand, prices rise. Long-term offtake agreement prices are based on mutually sustainable values for the pellet fuel. Those prices are unaffected by short-term supply and demand imbalances.

FutureMetrics has estimated CIF prices for several common destinations, including ARA and Japan. The estimates for CIF U.K. are in Figure 1.  Note that the weighted average considers the market share of the exporting countries. The U.S., Canada and the Balkan states dominate the share of pellets imported into the U.K. The most recent data for June shows that the U.S. has 58.4 percent of the U.K. market, with Canada having 19.6 percent, the Balkan states about 19.8 percent, with the small remainder from Russia. 

Since the trade data includes pellets imported at spot prices, lower-volume exporters such as Russia, which has traded almost entirely on spot, will exhibit higher price volatility. The three-month moving average and the trend smooth out most of that volatility.  Over the past several years, based in this analysis, the average long-term offtake price delivered to the U.K. has been between $180 and $190 per metric ton (MT). 

The Model
The goal of the model is to replicate the estimated delivered price of pellets, and then, based on commonly forecast macroeconomic parameters, forecast future prices.

The analysis assumes that that market is not in a condition of excess supply or demand.  In the short-run, there may be a supply and demand imbalance that would impact spot prices. However, the intrinsic costs of producing and delivering pellets, including typical profit margins, will set the market prices for long-run supply contracts.

The methodology involves developing independent submodels for each of the main components of pellet costs, which are wood costs delivered to the pellet mill, pellet mill conversion costs (excluding wood costs) plus margin, inland transportation from the mill to the port and port storage and loading costs, and shipping.

The wood cost model begins with stumpage cost and then accounts for equipment, labor and trucking for harvest and delivery. Within the model are inputs for distance traveled on four classes of roads, from in-woods to highway. The use and cost of diesel fuel is a large component of the total cost of delivered wood. If the pellet mill uses only sawmill residuals, the model accounts for the fact that the sawmill has absorbed most of the harvest and roundwood delivery costs.

Conversion costs are the costs to take incoming fiber and converting it into wood pellets.  The margin is based on historic values for EBITDA per MT.

For inland transportation from the mill to the port, costs per MT-kilometer are calculated based on several inputs. Storage and loading costs are based on typical rates and the amortized cost of the storage and ship loading infrastructure.

To estimate shipping costs, from the Argus Biomass Market Report, FutureMetrics used over 200 weeks of pellet freight rates from five different routes and two vessel sizes, and performed a regression analysis based on oil prices, distances traveled, vessel sizes, transit and loading/unloading time, and a few other parameters. The regression output provides a robust set of coefficients for estimating long-term shipping costs. 

FutureMetrics has added an adjustment to the per-MT shipping rates based on the Jan. 1, 2020, implementation of the International Maritime Organization sulfur cap on emissions from shipping.  The adjustment is based on distance traveled, and thus the increased cost per MT increases with longer distances.

There are several other inputs including general cost inflation rates.
FutureMetrics has estimated the potential variability of the critical inputs to the model and has developed probability distributions for those inputs. That allows for a series of Monte Carlo simulations. The simulations yield ranges for the price forecasts that are shown in Figure 2, which uses several stylized assumptions. The inputs for each pellet mill will be different and unique.

Results of the Analysis
Each quarter’s estimated CIF price is based on a probability distribution produced by the simulation. For example, for quarter four 2018 pellets shipped from the U.S. Southeast to Immingham, U.K., on a 45,000-MT vessel, the distribution of possible CIF prices from 10,000 interactions of the simulation is shown in Figure 3. 

Transport distances for the pellet feedstock have a strong impact on the delivered cost of wood to a pellet mill that uses roundwood or in-woods chips. The delivered cost of wood is the most influential factor in the total cost of producing pellets. The green bar gives some indication of the ability of the producer to buffer external cost changes with a lower or higher operating cash flow margin. The blue bar shows how if the producer can improve plant efficiency (for example, more consistent operation and output) then they can lower the price, but maintain their margin. 

Unless oil prices do not follow the upward trajectory forecast by the EIA and general cost inflation rates are below the assumption in this example, or producers can improve conversion costs or accept a lower EBITDA per MT, the delivered cost of industrial wood pellets is likely to be above $250 per MT by 2030.

Author: William Strauss
President, FutureMetrics