Abstract
The Japanese Pollutant Release and Transfer Register system could help investors identify a firm′s potential chemical risk related to toxic releases and transfers. In this case, the stock market may value the risk negatively. The conclusions of previous studies to explore how market values releases and transfers of toxic chemicals are inconclusive. These inclusive conclusions are likely to result from inappropriate methods. Some previous studies do not consider the omitted variable bias on the chemical risk variable. The other studies use the chemical risk variable including both releases and transfer, although the market is likely to value the releases and the transfers differently. This article addresses these shortcomings of the previous studies using firm level data: we use two risk variables, which are made from releases and transfers respectively, in the model and the instrumental variable for both releases and transfers. Our finding is that market values releases negatively but not transfer. We also find that the chemical risk variable is not significant, when we use the sum of releases and transfers of chemical substances following the previous study and thus the analysis, which does not distinguish risk of releases from that of transfers is likely to mislead the conclusion.
