This study analyzes the Nordic pine (Pinus sylvestris L.) sawlog markets in four main regions in Finland by using monthly real stumpage prices over the period January 1995 to June 2005. The special emphasis is on the short-run forecasting of different time-series models up to April 2006. As a benchmark case, we compare the models performance in terms of root mean square forecasting errors (RMSE) of standard autoregressive moving average (ARIMA) and vector autoregressive (VAR) models to those of Harvey's (1989) structural time series model (STSM), which, unlike the standard methods, decomposes the time series into unobservable components, such as deterministic and stochastic trend and seasonal and cyclical behaviour. The results indicate that, in most cases, the STSM together with Kalman filter estimation outperform ARIMA and VAR estimation. With hindsight, stumpage markets experienced a price decrease during July–December 2005 and a turning point up in early 2006 that none of these models were able to accurately predict. Based on these results, it seems to be that, in real-life forecasting situations, it is quite difficult to get precise estimates for the stumpage prices solely using the time-series approach, irrespective of how flexible the models may be with respect to structural changes.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.