Testing Timber Market Linkages with a STAR Model with Housing Start-Controlled Transitions

Market integration has been widely studied in the past to gain a better understanding of how different markets for the same good interact with one another. Traditional cointegration tests have been used to perform these analyses; however, these tests can only reveal the presence of stable or average long-term relationships. To evaluate market dynamics over time, we modify a time-varying smooth transition autoregressive (TV-STAR) model to examine pine stumpage markets in the U.S. South. The proposed model incorporates an economic indicator and allows us to evaluate market integration as it changes throughout a specified time period. The potential significance of this work is twofold. First, market integration can be observed as a function of an economic indicator, and secondly, the proposed model will allow us to draw inference about market integration at specific time points.