A quantitative analysis of weather effects on traded volume in the Swiss energy spot market

This paper quantitatively examines the effect of weather uncertainty on the traded volume of energy spot prices. Traded volume is driven by agents' anticipation of numerous factors of limited predictability, such as renewable energy in-feed or volatile fuel prices. Such stochasticity leads to volume time-series with noise like characteristics. The goal of our study is to investigate if there are structural artifacts in the time-series attributable to weather conditions. Employing a set of simple assumptions concerning the market behavior and the financial products available, we arrive at qualitatively sensible results from a more robust theoretical perspective. Our model is trained on historical data from the Swiss market, which depends on a highly predictable renewable energy, namely hydroelectric power. A filtering process is applied to remove deterministic elements of the volume time-series such as patterns due to dates or seasons. An analysis of weather impact on this time series proves to be consistent with an intuitive analysis of the data for the months of winter. An estimation of the relative volume shift during warm and cold days in winter is provided. Results prove to be consistent with the features of heating demand, particularly for spring and autumn, when the heating status (on/off) is highly sensitive to weather conditions. Numerical evidence also leads to the conclusion that there is no clear impact of extreme temperature events on the traded volume time-series in Summer months.