In this thesis we investigate whether the inclusion of an exogenous variable in the form of patentapproval can improve the ARIMA model's predictions for the pharmaceutical companyAstrazeneca. A point of departure for the study is the questioning of the efficient markethypothesis. When comparing data on patent date approval with stock exchange data for threepharmaceutical companies, it could be observed that share prices increased on the date ofapproval in 65 percent of the cases. This observed correlation combined with the fact that severalpapers have established that the stock market may not be efficient make it interesting to studywhether the value of a patent has been included in the stock price prior to approval date.To investigate this, an ARIMA and an ARIMAX model was estimated. The exogenous variable,which controls for patent approvals, was created by retrieving data from the EPO's databasePATSTAT. The retrieved data was then formatted into a dummy variable. The purpose ofincluding an exogenous variable is to investigate whether the market reacts to patent information.If the addition of the exogenous variable proves significant, the result is in conflict with theefficient market hypothesis.During the model selection, it was found that an ARIMA (4,1,2) was the superior model. Themodel was then compared with the corresponding ARIMAX model. When comparing themodels, it was found that the predictions of the ARIMAX model follow the observed datasomewhat better, but a t-test concluded that the improvement was not statistically significant.This implies that the value of the patent has already been included in stock prices prior to patentapproval and indicates that the price increase is random. This results thus lends support for theefficient market hypothesis. To investigate this further, the stock market data was compared witha random walk and by conducting a t-test it could be concluded that it was not possible to rejectthe hypothesis that share prices follow a random walk, thus the result further supports theefficient market hypothesis.
Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:liu-180927 |
Date | January 2021 |
Creators | Hill Anderberg, Camilla, Gustafson, Alice |
Publisher | Linköpings universitet, Nationalekonomi, Linköpings universitet, Filosofiska fakulteten |
Source Sets | DiVA Archive at Upsalla University |
Language | Swedish |
Detected Language | English |
Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
Format | application/pdf |
Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0017 seconds