A Count Model Analysis of Traded Contracts in the JEPX Forward Market

SUMMARY It is desirable that the number of forward contracts traded on the Japan Electric Power Exchange increase. However, few studies have clarified what factors have contributed to impacting the number of forward contracts traded. In this study, the authors analyzed the number of forward contracts using four count regression models. The negative binomial regression model and the zero‐inflated models were better able to express the expected counts, by incorporating the overdispersion and excess zeros present in the observed data. Among others, the spot market can carry positive influences on the expected counts, by about 12% for a 1 yen/kWh increase in price, and by about 27% for a 0.1%‐point increase in volumes. The zero‐inflated models revealed that as many as three‐fourth of the entire forward products have a high probability of zero counts, while the remaining one‐fourth may see an increased number of counts as the spot market price and/or the spot volume becomes higher.