Aggregation bias in natural resource price composites: the forestry case

In this paper, we use cointegration tests to evaluate whether two types of aggregation bias are likely to be present in the composite price time series of forestry products. We find that the two types of bias are likely to be present, and furthermore, that forestry price series tend to be difference, rather than trend, stationary processes. These results raise doubts about the information content of composite forestry price time series, and suggest that inferences from such information about timber scarcity trends will likely be spurious. If the time series properties of forestry price series hold for other natural resources, then inferences about resource scarcity trends will also be dubious in these cases. This possibility suggests the need for further research to explore the time series properties of other natural resources.