Market structure in U.S. southern pine roundwood
Time series of commodity prices from multiple locations can behave as if responding to forces of spatial arbitrage. cvcn while such prices may instead be responding similarly to common factors aside from spatial arbitrage. Hence, while the Law of One Price may hold as a statistical concept, its acceptance is not sufficient to conclude market integration. We tested the factors hypothesized as linked to integration of forest products markets by applying a combination of bivariate and multivariate techniques. Bivariate cointegrntion tests were conducted for price pairs among 21 price regions and were done for both delivered southern pine sawlogs and delivered southern pine pulpwood logs. Multivariate meta-analytic regressions of cointcgration test results on hypothesized explanatory factors were run for pulpwood and sawlog markets separately. Cointegration test results offer limited support for the Law of One Price in the South for both products. Results of the meta-analytic regressions show that a proxy for the cost of product transfer between regions is statistically significant and negatively related to the probability that two local market prices arc cointegrated for only sawlogs. For pulpwood, the proxy was not significant. The results of the bivariate cointegration tests and the multivariatc meta-analyses were used to delineate apparently spatially segmented sub-markets for both products. The maps show overlapping geographical segments, resulting from both spatial arbitrage and possible output dominance for certain firms in those sub-markets.The southern pine sawlog market can be divided into four or five sub-markets, distributed north to south and east to west.The southern pine pulpwood log market can bc drawn into three, largely separate sub-markets: a coastal zone that stretches from Texas to Virginia, and two distinct interior zones.