With the entry into the Single Intraday Coupling (SIDC) platform in September 2021, the Italian intraday market has moved from seven implied auctions to a hybrid model consisting of continuous trading interspersed with three implied auctions. This report aims to analyze the possible economic-financial modeling techniques applicable to continuously traded electricity markets, with the goal of building a model capable of replicating their performance as faithfully as possible.

Initially, a comparison between the two different types of intraday markets (before and after joining the SIDC) is presented. Analyses are concerned with the volumes traded in the different market sessions, the trend and volatility of bid prices (also as a function of the Day-Ahead Market price), and the frequency of submission of such orders during the continuous trading session. Interest is focused, among other aspects, on orders submitted by two different types of generating plants (thermoelectric and wind).

Next, an overview of mathematical models capable of describing continuous trading markets is presented. In particular, optimization, equilibrium, and simulative models are presented, with emphasis on the various game theory techniques that describe the strategies that can be adopted by traders to optimize their profits or reach equilibrium in a cooperative manner.

Of these models, further insight is given into the simulative ones, which are considered to be the best ones to apply to continuously traded intraday markets. We then expose the most widely used algorithms in the literature to simulate traders’ strategies aimed at maximizing their profits based on information coming from the market itself, from external events, or from historical data available to each trader.

Particular attention is paid to Vytellingum’s adaptive-aggressiveness model, capable of describing an agent’s willingness to be active in the market based on the information it receives during the trading session.

Finally, two innovative simulation models of continuous-trading electric market are presented. The first, starting from the adaptive-aggressiveness model, optimizes traders’ strategies through a genetic algorithm that iteratively modifies the main parameter of Vytellingum’s model, optimizing total welfare and the amount of energy traded.

The second, on the other hand, is a model that can replicate the most characteristic aspects of the market with continuous trading, such as its efficiency, price movements driven by factors inside and outside the market, and the frequency with which traders submit orders. The two models are then applied to test cases, the results of which are presented.