Can Searching in Google Predict a Profit in a Stock?

Hepran, Agus Putra Marunduri (2021) Can Searching in Google Predict a Profit in a Stock? Diploma thesis, Universitas Andalas.

Preview	Text (Cover and Abstract) abstract.pdf - Published Version Download (131kB) | Preview
Preview	Text (Chapter 1 Introduction) chapter 1 introduction.pdf - Published Version Download (164kB) | Preview
Preview	Text (Chapter 5 Summary and Conclusion) chapter 5 summary and conclusion.pdf - Published Version Download (274kB) | Preview
Preview	Text (Bibliography) bibliography.pdf - Published Version Download (120kB) | Preview
	Text (Fulltext thesis) fulltext thesis.pdf - Published Version Restricted to Repository staff only Download (3MB)

Abstract

The objective of this research is to verify that a google trend of a company can not be used to predict a stock return of that company. The grand theory of this research is random walk theory that claims none company stock is predictable. To reach that objective, this research employed an Autoregressive Integrated Moving Average (ARIMA) model from week 44th of 2014 to week 52nd of 2019 (270 weeks) of companies that listed in LQ45 index. The operational research objectives of this research were to found the correlation between the trends of stock return and google trend, to calculated the significance of the combination of stock return and google trend data, and last to generated a forecast of stock return change in the next 8 weeks. The result of this research was random walk theory held true to LQ45 index. The majority of companies showed that google trend has no significant effect to stock return of a company. Based on that result, the recommendation of this research is do not rely on google trend to make profit in a stock. Instead of trying to generate profit in the short time by using google trend, rely on the long term growth of the invested company. It is because a company tends to grow up in the long term.

Actions (login required)

View Item