Estimating Processed Commodity Prices from Raw Material Inputs
Summary
The document discusses estimating downstream commodity prices, such as flour from wheat, when direct price series are unavailable. A conversion factor from raw material to finished product is not enough to capture the finished price. One response describes industry use of regression, which requires historical observations of the product price and relevant inputs. Another frames the processed good’s price as reflecting raw materials, labor, energy, and other costs, along with a margin shaped by finished-product demand.
The examples point to important limits: input costs alone may not explain changing margins, and access to historical prices can be difficult because market surveys are commercially valuable. Regression can relate finished prices to input costs, but the document provides no fitted model, data, or performance evidence. For markets where supply and demand materially affect margins, a useful estimate may require modeling those conditions as well as production costs.
Key ideas
- A raw material price and conversion factor do not fully determine a processed product’s market price.
- Regression on historical finished-product prices and input costs is one possible estimation method.
- Processed prices can include labor, energy, and other production costs.
- Supply and demand can change the production margin and limit cost-only estimates.
- Historical downstream price data may be difficult to obtain.
Tags
Full text
# How to derive the prices of downstream products from raw commodity prices? # How to derive the prices of downstream products from raw commodity prices? I am looking for a simple way to estimate price time series of downstream products based on price of the main "raw" commodity. For example, would like to estimate a price for wheat flour based on available wheat price data and wheat per unit weight flour conversion factor. Some cursory googling seems to suggest that there is no standard way to do this. How would a quant go about such a task? Incidentally, if anyone knows of a source for readily available flour price data, please let me know. ## Answer by SRKX (score 1) https://quant.stackexchange.com/a/32296 This is a typical issue in physical commodities trading and I don't think you'll find free sources for such prices because it's precisely some firm's business to talk to market participant and perform surveys in order to come up with an approximate. As suggested by noob2, what I've seen in the industry is usually a regression, but you need at least historical data to perform this. ## Answer by ZRH (score 1) https://quant.stackexchange.com/a/43870 In my mind, there is no obvious way how to do this. The price of the processed good will be be AT LEAST the cost of the raw material plus the cost of labour for processing plus the cost of other goods that are inputs to the process (energy etc.). A linear regression of the price of the finished good with respect to prices of the inputs will allow you to get rid of labour cost dependence and other price dependencies. However on top of this you expect to see a markup that is driven by demand for the finished good. As a simple example take a power plant: The price of power will be the cost of fuels plus the cost of labour plus a production margin. The latter will fluctuate over time as a function of supply/demand In other words, in many cases you will not gain much insight without supply/demand modelling
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.