Regional variation in radiata pine sawlog prices in New Zealand

There is considerable regional variation-in market characteristics and log quality for radiate pine sawlogs in New Zealand, which should be reflected in a corresponding variation in sawlog prices between regions. This thesis presents a study of log price variation for three broad categories of domestic sawlogs (P, S and L grades) in four wood supply regions in New Zealand (Central North Island, Nelson/Marlborough, Canterbury and Otago/Southland). A survey of log producers was used to gather data on log price and log quality data. The survey data covered the years 1997-1999. Additional information on regional sawlog market characteristics and firm characteristics was obtained from published sources, as well as directly from the firms covered in the survey. Differences between sawlog prices for the three sawlog grades were tested using ANOVA. Sawlog prices in the four regions were found to be significantly different. This result was consistent over the period 1997-99, and across all three sawlog grades. A hedonic price model was used to predict log prices that could be compared to actual prices. Sawlog physical attributes were used to calculate a single sawlog residual value, using programme SAWMOD and a spreadsheet developed by INTERFACE. The relationship between sawlog prices and physical log attributes was tested for significance by a regression that used residual value to predict saw log prices. The regressions were highly significant for both pruned and unpruned sawlogs For unpruned sawlogs, the sawlog residual value explained most of the variation in sawlog prices in the survey data. However, regional and firm factors also had a significant effect on prices. A regression using the four wood supply regions in the study, the size of the firm, the market knowledge of the firm, and whether the firm sold wood on the open market as additional explanatory variables, had an R-squared value of 96.1%. The estimated coefficients for the dummy variables denoting regions, and firms' market knowledge were statistically significant. Further research arising from this study could be undertaken into the degree to which firms are using log scanning and database technology to capture data on sawlog physical attributes. In addition, this study found no clear evidence of differences in the competitiveness of regional sawlog markets. A closer investigation of buyer and seller behaviour at the firm level, as well as the potential for outside firms to enter a regional sawlog market, may shed more light on this issue. Finally, the methods used by forestry consultants and forestry companies to estimate sawlog prices for valuation are usually not disclosed. Research into price estimation methods used by forestry firms may reveal the extent to which sawlog price estimates take account of regional and inter-firm differences in sawlog and market attributes.