Discussion Papers in Statistics and Econometrics No. 8/10
University of Cologne, Seminar of Economic and Social Statistics, Cologne
We elaborate economic explanations for the time-varying risk of month, quarter and year base load electricity forward contracts traded on the Nord Pool Energy Exchange from January 2006 to March 2010. Daily risk quantities are generated by decomposing realized volatility in its continuous and discontinuous jump component. First, we analyze the relation between volatility and trading activity. Coherent with existing studies we find that the driving factor of the relation between continuous variation and trading activity is the number of trades. New insights are obtained by considering the relation between jump factor and trading activity. Our results indicate that the number of trades and absolute order imbalance, which can be explicitly measured in our dataset, are positively related to the jump factor, a result in line with theoretical models. Second, we study unscheduled news announcements causing high volatilities. For this, a unique dataset of urgent market messages (UMMs), published by the Nord Pool Energy Exchange, is created. We extract relevant unscheduled UMMs, here failures, from both transmission system operators (TSOs) and market participants (MPs), and measure their impact over varying event windows. We find that certain unscheduled TSO/MP-UMMs have a significant impact on continuous variation, especially when they are published close to maturity, their content refers to a rare and extreme event or the contract is a month forward. The analysis also provides economic evidence for the occurrence of price jumps.
Electricity Forward Contract High Frequency Data Realized Volatility Price Jump Trading Activity Urgent Market Message