Research Project Submitted to Departamento de Economia e Relações Internacionais (UFRGS) and to The Alfred P. Sloan School of Management to apply to a Visiting Scholar Position with Professor Rodrigo Verdi.

1. The Research Problem and its Importanceion

We propose a method for estimating the impact of analysts' forecasts on stock returns based upon Lucas (1978) The research consists of two parts. The first one is to be done with US data and the second one to be done with Brazilian data in order to account for information asymmetry in a \citet*{Ma2016} fashion.  This project is the plan of the of first part of the research where we will collect price target data from analysts of Brendan E. Cryan & Co., Citadel Securities LLC, GTS Securities LLC, IMC Financial Markets, and Virtu Financial Capital Markets LLC, which are market makers in New York Stock Exchange (NYSE) and test if there are correlation and causality in the sense of \citet{Granger_1969} between those series in order to evaluate the impact of market makers analysts' reports.  In addition we will apply a methodology which aims to make it feasible to repeat that procedure taking into account not only market makers but all brokerage companies analysts' reports which operate in NYSE in accordance with Financial Industry Regulatory Authority (FINRA) database and Investment Adviser Public Disclosure website.  Nevertheless, even after we collect the data the big number of those firms and reports might make it hard to develop to right application to treat those data and therefore that might be postponed to the second part of the project when we will be working with Brazilian data.
From a theoretical viewpoint, in equilibrium agents' expectations collapse into actual prices as clearly posed in \citet{Lucas_1978} and \citet{Breeden_1979}.  They say nothing though about the role of accounting numbers in the formation of those expectations.  Actually, they are compatible with \citet*{Fama_1970}  semi-strong market efficiency hypothesis where all public information is somehow already into market prices, which doesn't leave room for any kind of forecast based on accounting information.  On the other side, the empirical literature derived from \citet*{FAMA_1992} and  \citet*{Fama_1993} finds price effects of accounting indices in expectation of returns while  \citet*{OHLSON_1995} then develops a model where accounting data matters in a, as the author says, "neoclassical framework" (p. 662), which means in his terms that "value equals the present value of expected dividends" (p. 662).  Indeed, in Lucas setting, that implies marginal rate of inter-temporal substitution is constant for all agents, which allows to use macroeconomic data to evaluate stock prices.  Besides, this approach allows to rationalize stock analysts forecasts, in particular, when it comes to fundamentalist analysts.
Therefore this research aims to contribute to fill this gap,  to a better understanding of the role of accounting information in stock returns, and also to find the channel through those data may flow to impact on market returns.
As posed earlier the sample we need to collect to perform our test is very big.  Therefore, although the subject is akin to economics, finance and accounting, the methods needed to deal with such a big sample will be computational and statistical intense.  Consequently, I need to be hosted by a US university nearby New York area in case I shall need to go to NYSE to do field research.  Also, that university must have a strong tradition not only in Economics, Finance and Accounting but also an emphasis in data science. Anyway, the best choice is to do the research with one of the authors of the survey paper I mentioned to have someone to talk about it.  Luckily, The Alfred P. Sloan School of Management at Massachusetts Institute of Technology fulfills all those requests and therefore is the perfect place to apply to do this research.
The organization of the project is as follows.  The second section presents the research problem and its importance in terms of the theoretical and empirical asset pricing literature.   In the third section I present the model I hypothesize to be the main one in a first attempt to describe the impact I hope analysts' forecasts should have on stock returns, which is \citet{Lucas_1978}. The fourth section shows the methods and data I plan to use in order to achieve the main objective. Section five presents the expected results and the next one concludes the project.

The main hypothesis

The research consists of two parts. The first one is to be done with US data and the second one to be done with Brazilian data in order to also account for information asymmetry in a \citet*{Ma2016} fashion.  This project is about the of first part of the research.