Strategie di portafoglio sul mercato azionario basate su notizie pubbliche

Public news is a relevant source of information describing events occurring in the real world. Despite information leakages and fake news readers are often willing to pay to gather information conveyed through the news. In this work, we investigated the effects of micro news sentiment on the equity market. Therefore, according to several market theories of Behavioral Finance, we want to understand if public news is interesting for the equity market investor. The Adaptive Market Hypothesis is one of these theories and states that also if markets are efficient in general, inefficiency may affect them in particular periods. These inefficiencies may then leave open profit opportunities for the informed investor. For this reason, public news could have an important impact on investor's choices. In the second chapter we present the Efficient Market Hypothesis, Behavioral Finance, and the Adaptive Market Hypothesis. After the basic theoretical introduction, we report a review of the available literature regarding studies on the effect of the news on markets. The reported papers cover many relevant aspects regarding how to measure the news impact on the market returns. The main factors extracted from the news are: the news sentiment, the cumulated sentiment, the news relevance and novelty, and the news topics and categories. The third chapter tries to describe the events happening in the market that could be conveyed through the news. At first, the events characterizing the value generation process and the highly complex firms' interconnection are described. The description then moves on to events concerning the firms' evaluation process, which is split into two sections: the former about fundamental indicators evaluation and the latter about perspectival views and ratings. In the last section of the chapter events regarding the ownership structure of a firm are reported and a possible path of transmission from the news to the effects on the market is outlined. The fourth chapter describes the market data and news sentiment database and reports statistics about the considered stocks. The description of the data also reports statistics about category groups, a high-level grouping of news events in a hierarchical taxonomy of financial-related news events. The fifth chapter presents news-based strategies for intraday open to close trading. The strategies are based on a naive beauty context model that takes into account only indicators generated by firm-specific public news sentiment or volume subdivided by category group. Three main patterns emerged from the analysis and characterize different category groups: sell on news volumes, buy according to news sentiment, and sell according to news sentiment. The analysis has also shown that many portfolio strategies based on category groups are characterized by a reversal effect overnight, while few others by a continuation trend. The sixth chapter presents a more complex model that tries to enhance baseline models with the use of public news. The baseline models rely on three different portfolio optimization criteria: Sharpe Ratio, Second-order Stochastic Dominance, and Scaled Second-order Stochastic Dominance. The Second-order Stochastic Dominance optimization criteria are based on enhanced indexation, where the optimal portfolio is supposed to be the best portfolio dominating the reference market index. The aim of the models is to understand which category groups bring useful information for portfolio optimization. The discussion of the results is divided into two parts, the former reports profitability and turnover results for the three baseline strategies, while the latter reports the results for the different category groups and cumulation periods. In the last part of the chapter a series of possible further researches are proposed.