Forecast Combinations for Electricity Prices: The case of the Italian Market

In Italy, the electricity market was established as a result of the European Union's directive to create an internal energy market that is governed by a set of rules and regulations that ensure fairness and transparency of transactions, allowing for a competitive and efficient market. Previously, the electricity market was characterised by a monopoly that had control over nearly all aspects of transactions and price formation. Over the years, the formation of the electricity market in Italy has led to a significant increase in competition between suppliers, resulting in improved prices and services for consumers. The marketplace is composed of three main segments: the Day-Ahead Market, the Intra-Day Market, and the Dispatching Services Market. The Day-Ahead Market (Spot Electricity Market) and Intra-Day Market (Balancing Market), also known as Energy Markets, are managed by Gestore dei Mercati Energetici (GME), whereas the Dispatching Services Market is managed by Terna. As electricity can’t be stored, a wide range of different contracts is necessary on the market to keep the balance between supply and demand. In such a complex and dense environment, electricity price forecasts are of great use to market participants and electricity market monitoring bodies, since they provide crucial information for the management activities of electricity generation, transmission, and distribution. For market operators, price forecasts help in planning electricity supply and choosing trading strategies. For example, short-term price forecasts allow traders to decide whether to buy or sell electricity in the market to maximize profits, while medium and long-term price forecasts help the operators to plan future investments in new generation capacity and infrastructure, ensuring a long-term stable and reliable supply of electricity. Moreover, electricity price forecast plays a determining role while looking at the risk management side of market participants; in fact, it is extremely useful to reduce exposure to volatility, allocate resources efficiently and make informed investment decisions. Single forecasts in the context of electricity prices can sometimes be inaccurate, leading to potential financial losses and risks for market participants. Therefore combining different forecasts coming from different models can result helpful as it can allow us to obtain a better and more accurate forecast. In the context of electricity prices, the forecast combination can be useful, for instance, to mitigate uncertainty in energy markets and to reduce the risk associated with investment decisions. In addition, one model could outperform others while looking at a specific hourly timeframe but not for others. It is important to consider that to achieve a substantial improvement in forecast combination, it is not only the quality of the single forecasts that is important but also the estimation of the weights attributed to each of them.