Our previous blog on the uniqueness of individual timber markets and wood baskets generated questions and follow-ups on the tyranny of regional averages and the challenge of localizing price forecasts. In our world, there is no such thing as an “average timber market.” Which, by extension, means there is no such thing as a “regional” timber price. Who buys or sells timberland, manages timber sales for forest owners or procures logs and stumpage for mills based on regional averages?
In our quarterly price forecasts for the US South, Pacific Northwest and North, we don’t forecast regional prices. Rather, we project prices by state or defined market. For example, for the South, we forecast prices individually for eleven states by product, and our Southern forecast is a volume weighted average of the eleven states. To us, the Southern price or forecast has little meaning; it is simply an “output” and consequence of doing all of the local work in the individual states. When we back-test and evaluate our models each quarter, we go state-by-state and product-by-product. So if we are “on track” with 8 states and “off track” with 3 (so to speak), we are more interested in those individual stories.
To me, this simply aligns with how things work in the world. When I buy gas for my car, I don’t buy based on the Southwide average for gasoline prices. I compare the local price of the QuikTrip near the house to the RaceTrac near the office. And local can mean different things. Mint Mentos at the airport can cost $2.99 per roll but the drugstore two miles away charges $0.99. While the average price is $1.99, who pays that, and how? (And how “local” is another store anyway after you come through airport security and your car sits in long-term parking?)
How do local timber markets and wood baskets perform relative to each other given whatever else is happening in the world? Where should you allocate capital or move harvest or divest? We have found that local markets have wide ranges of price-to-demand elasticities and mill risk profiles, beyond those established in regional or national analyses. Differences across markets are statistically significant and provide a rigorous basis to adjust market-specific discount rates, stumpage price forecasts and expected rates of recovery.
If you are in the timber or wood-using sectors, you have to look at the world this way. Deciding where and how to invest in properties or mills is totally different than deciding whether or not to get in or stay in a sector. In this way, forecasting applications serve two totally different masters, and we should be clear on this point. Our team is, for the most part, “bottom up” in our work and thinking. This is how we spend our time.