This thesis tries to find long run steady-state relations between wood pellets prices and other market relevant time series for the heating sector. Thirteen years of data are examined and processed for further use in vector autoregressive models. Basic theoretical results and advanced econometrical methods such as Johansen-s cointegration test and the varimax transformation by Kaiser are presented and applied to develop various models to describe and investigate the data. These models are compared to each other and eventually used to perform forecasts that are being discussed. The computations are performed by the statistical programming language R with a mix of precast packages and self-developed routines.