Incerteza, sentimento e precificação de ativos: uma análise a partir de uma proposta de quantificação para “noise”

Abstract

The relation between the ambiguity of the macroeconomic environment and the investor sentiment has not yet been sufficiently explored by the financial literature. Thus a possible joint effect of ambiguity and sentiment on stock returns is not known. There are times when there is a lot of unpredictability about the economic outlook and distinguishing news from noise becomes a more difficult task for the investor when deciding their exposure to risk. At such times, the use of heuristics and biases in decision-making intensify. Faced with this, this thesis contributes to a better understanding of the information structure and the decisions of the investors developing a measure for noise. Measure hitherto lacking in the international and national literature, given the difficulty of identifying and measuring it. Thus, a new approach was proposed, in which eight ambiguous measures of the macroeconomic environment were taken as proxies for the "predictability" of the economy and "cognitive state of uncertainty" of the agents, orthogonalized to a measure of fundamental uncertainty. So, noise was quantified from a linear combination of latent common factors. Data were used for Brazil, in monthly frequency, for the period from 2002 to 2017. The evidence indicated that NOISE captures the noise trading behavior and the fly to quality behavior for actions whose salient characteristics were pointed out in the literature as more susceptible to sentiment. It has been shown that NOISE can be used as an alternative proxy for investor sentiment, but more sensitive to higher-order shocks in the distribution of the benchmark measure for sentiment (IBW), such as risk perception errors and fear. It was verified that pricing models can gain explanatory capacity and accuracy by incorporating NOISE as a risk factor. The results suggest new researchs and potential utility for the market (tool for investment decisions and risk management), policy makers (economic indicator of noise in expectations) and regulatory authorities (monitoring of price informaticity).