Student: Diego Silva de Salles
Title: Detecção e análise multi-scale de eventos originados por fatores externos de incerteza em séries financeiras
Advisors: Eduardo Ogasawara e Eduardo Bezerra
Committee: Eduardo Ogasawara (advisor), Eduardo Bezerra (co-advisor), Rafaelli Coutinho (Cefet/RJ), Carlos Eduardo Mello (UNIRIO)
Day/Time: December 22, 2022 / 9a.m.
Abstract: Different external factors reported in the media can impact a financial time series. Such factors, such as government transitions, economic crises, or corruption scandals, can be related to events that increase uncertainty in the time series. In particular, these external factors can increase the risk perceived in a financial time series through events such as anomalies or change points. A study based on the different characteristics that make up an event can determine predictions, in addition to helping to minimize the risk in investments. The influence of these factors can have different cycles of fluctuations, affecting a time series over months or years. Hence, discovering these events in the financial time series is a challenging task. This paper presents Multi-Scale Event Detect (MSED), a technique for detecting events in non-stationary and nonlinear time series. Added to this, this work makes an associative study of the events found by the detection methods in the components of the Intrinsic Mode Function (IMF) with the external factors labels obtained through the Economic Policy Uncertainty (EPU). The objective is to identify which type of event is reflected by a given external uncertainty factor in a financial series, using this approach it is possible to determine the most predominant nature of the events based on the uncertainty variations presented in the series of (EPU). This information makes it possible to determine a set of time series where the influence of uncertainty generates acceptable events for a given investment profile.Thus, mitigating the risk to which it is intended to be exposed.