.. raw:: html

    <style>
    .my-box {
    border: 1px solid #ddd;
    border-radius: 6px;
    padding: 10px;
    margin-bottom: 10px
    }
    </style>

.. _research:

Research
********

My research so far has been conducted within the environment of the `Federal University of Rio Grande do Sul <http://www.ufrgs.br/ufrgs/inicial>`_ (UFRGS),
where I completed my :ref:`Undergraduate Thesis <undergraduate-thesis>` in Economics and my :ref:`Master's Thesis <masters-thesis>` in the
Professional Master's in Economics program.

I contributed, through the `Data Science and Computational Economics and Finance Unit <https://www.ufrgs.br/ecompfin/>`_ (e-CompFin), to the research project of computational finance called `Quant-FAME: Quantitative Financial, Accounting, Management and Economics <http://dgp.cnpq.br/dgp/espelhogrupo/37785>`_.
I employed methods of **econometric modeling** and **quantitative finance** to develop data science applications aimed at answering research questions through **empirical hypothesis testing**. These applications were developed using the Python programming language and automated the entire process of data collection, transformation, processing, and visualization.

The entire research was conducted under the paradigm of `Open Science <https://en.wikipedia.org/wiki/Open_science>`_, which aims to make scientific research accessible to everyone, including both the academic community and individuals outside it.

Below, I present the abstracts of the articles I produced during my academic journey at UFRGS.

.. _masters-thesis:

Master's Thesis: Politics and Informational Efficiency: Brazilian Term Structure
================================================================================

.. raw:: html

    <embed>
    <div class="my-box">
    <h4>
    Resumo
    </h4>
    <div class="my-box">
    <p align="justify">
    Economic policy, as it affects the state of the economy, has a major impact on financial returns. The continuation or change in economic policy, however, does not occur without context, and is always a result of the continuous day-to-day political process. In this context, the problem we address in this research is whether daily politics affects the dynamics of the term structure of interest rates. We use web-scraping to collect political news from major Brazilian news portals. We consider the term structure of the one day inter-bank deposits rate, which is closely related to Brazilian monetary policy, and we estimate the term structure with data on the actual interest rate and its future contracts. We finally test whether the daily volume of political news causes volatility in the term structure. We find that politics does affects the term structure of interest rates. For specific maturities and with limited lags, our results show significant causality between political news and the term structure. The major implications of our findings relate to the management of risk in investment portfolios. Day-to-day politics is found to be a valuable source of information for the prediction of term structure risk, and thus the data on events of the continuous political process may be added to the set of information used by investors in the portfolio management process.
    </p>
    </div>
    </div>
    </embed>

:download:`Politics and Informational Efficiency: Brazilian Term Structure <../../assets/BernardoPaulsen-dissertacao.pdf>`

.. _undergraduate-thesis:

Undergraduate Thesis: Some Evidence on Political Information and Exchange Coupon in Brazil
==========================================================================================

.. raw:: html

    <embed>
    <div class="my-box">
    <h4>
    Abstract
    </h4>
    <div class="my-box">
    <p align="justify">
    We investigate whether political news affect the exchange coupon in Brazil, on a period
    ranging from November 24, 2016, until April 16, 2019. Our sample of news was collected
    via web scrapping, which we applied on a Brazilian news portal. We used two measures
    for the exchange coupon, and we utilized a GARCH model to estimate conditional volatil-
    ity, which we filtered with both parametric and non-parametric approach. The results from
    the parametric analysis indicate that the exchange coupon was affected by the corruption
    scandal of President Temer (May 2017), the announcement of Jair Bolsonaro as candi-
    date for the republic’s presidency (July, 2018), the first round of the presidential elections
    (October, 2018), and the announcement of a new Pension Reform (January, 2019). The
    results from the non-parametric analysis indicate that, besides the events above, the ex-
    change coupon was also affected by news related to the Labour Reform (March 2017) and
    the impeachment requests of President Temer following the corruption scandal (August
    and October 2017).
    </p>
    </div>
    </div>
    </embed>

:download:`Some Evidence on Political Information and Exchange Coupon in Brazil <../../assets/BernardoPaulsen-TCC.pdf>`