Forecasting The Stock Market

This was a project I did for an economics class called Probability Models for Economic Decision Making. The purpose of this project was to determine how forecasting the stock market influences investors’ decisions on whether or not to invest in a certain stock. We used the Consumer Price Index to determine how public attitude and confidence weighs into forecasting decisions made in the stock market.

Abstract of Project

In 1936, Keynes likened a beauty contest popular in England at the time to the stock market. In these contests, participants were told to select the 6 most beautiful faces from a group of 100 and when the voting was over, whoever selected the most correct guesses would win a prize. Because of this, contestants weren’t necessarily picking who they felt the most beautiful, but making assumptions about how other people would rate the models.
 A similar experiment was developed in which a group of people are asked to predict a value equal to two thirds of the average  number from 1 - 100 of what all other contestants will choose. More clearly, all participants are asked to choose a number between 0 and 100. However, their objective is to select the number that wil be some percentage (p)*mean of all choices submitted. This is appropriately called the “P-beauty contest.” Assuming p = 2/3, there are different guesses that represent different levels of thinking. A “first-level” thinker would recognize the mean to be 50, and thus would choose the value of 2/3*50 = 33. This guess would only be logical if only one person in the group and the rest of the participants chose numbers at random. However, it would be more logical to assume that other participants would recognize this idea, and if all thinkers were first-level, the new mean would equal 33. Therefore, someone who was thinking on the “second-level” would guess a value of 2/3*33 = 22. If these levels of thinking were then followed they would eventually reach a Nash Equilibrium at 0. However, because not everyone recognizes this, and not everyone is on the same level of thinking, the winning value is often somewhere close to 25***********.
While extremely simple, this experiment reflects many key ideas about investing.  It is often not about the value of the product that shareholders depend on, but more on their expectations of how other investors will behave in the situation.
Stock market crashes continue to remain a critical subject of analysis for many economists today.  It has been largely recognized today that there is a significant psychological contribution to the swings of the stock market.  Keynes was one of the first economists to discuss this idea of collective crowd behavior, emphasizing the strength of many interacting agents rather as more influential than the values derived from analysis of present conditions and future prospects of firms.  Keynes termed this the “non-fundamental” factor, suggesting that in order for investors to form their demand for an asset, they not only forecast the future payoffs but also try to guess other market participants’ forecasts and others’ forecasts of others’ forecasts (CITE). 

Purpose: The question that this research paper attempts to answer is how the P-Beauty contest relates to the stock market. To do this we look at the Consumer Price Index in order to determine how public attitude and confidence weighs into forecasting decisions made in the stock market. The Consumer Price Index is important in our analysis because it represents changes in the price level of goods and services purchased by households. The CPI is also a strong measure of inflation periods so it can be compared to how well consumers and investors believe the economy is performing. Specifically, we looked at the stock of Apple in order to simplify the analysis.