This research assessed market efficiency, volatility behaviour, asset pricing, and oil price risk exposure of the oil and gas companies quoted on the London Stock Exchange with the aim of providing fresh evidence on the pricing dynamics in this sector. In market efficiency analysis, efficient market hypothesis (EMH) and random walk hypothesis were tested using a mix of statistical tools such as Autocorrelation Function, Ljung-Box Q-Statistics, Runs Test, Variance Ratio Test, and BDS test for independence. To confirm the results from these parametric and non-parametric tools, technical trading and filter rules, and moving average based rules were also employed to assess the possibility of making abnormal profit from the stocks under study. In seasonality analysis, stock returns were tested for the day-of-the-week and month-of-the-year effects. Volatility processes, estimation, and forecasting were undertaken using both asymmetric and symmetric volatility models such as GARCH (1,1) and Threshold ARCH or TARCH (1,1,1) to investigate the volatility behaviour of stock returns. To determine the effect of an exogenous variable on volatility, Brent crude oil price was used in the models formulated as a variance regressor for the assessment of its impact on volatility. The models were then used to forecast the price volatility taking note of the forecasting errors for the determination of the most effective forecasting model. International oil price risk exposure of the oil and gas sector was measured using a multi-factor asset pricing model similar to that developed by Fama and French (1993). Factors used in the asset pricing model are assessed for statistical significance and relevance in the pricing of oil and gas stocks. Data used in the study were mainly the adjusted daily closing prices of oil and gas companies quoted on the exchange. Five indices of FTSE All Share, FTSE 100, FTSE UK Oil and Gas, FTSE UK Oil and Gas Producers, and FTSE AIM SS Oil and Gas were also included in the analysis. Our findings suggest that technical trading rules cannot be used to gain abnormal returns, which could be regarded as a sign for weak form market efficiency. The results from seasonality analysis have not shown any day-of-the-week or monthly effect in stock returns. The pattern of stock returns’ volatility can be estimated and forecasted, although the relationship between risk and return cannot be generalised. On a similar note, the relationship between volatility attributes and the efficient market hypothesis cannot be clearly established. However, we have established that volatility modelling can significantly measure the quantum of risk in the oil and gas sector. Market risk, oil price risk, size and book-to-market related factors in asset pricing models were found to be relevant in the determination of asset prices of the oil and gas companies.

This thesis is a case study in collaboration with the company Getswish AB. GetswishAB provides the mobile application and payment service Swish with the purpose ofdelivering smooth money transfers for individuals and companies in Sweden. About80 percent of the Swedish population are connected to Swish, and the majority seethe service as an apparent part of everyday life. This work studies a small part of alltransactions that take place daily between individuals and companies. Specifically, thispaper examines which factors affect the Swish transaction amount (TA) to companieswithin five different industries. The five industries studied are: Sports, leisure,and entertainment activities; Restaurant, catering, and bar activities; Retail trade,except for motor vehicles and motorcycles; Trade and repair of motor vehicles andmotorcycles; and Telecommunications. In combination with descriptive analysis andseasonality studies, a multiple linear regression model is used to evaluate patternsin the amount transferred to companies within the various industries. The responsevariable is the daily aggregated TA and the seven responding regressors examined are:i) The number of employees of the company, ii) The revenue of the company, iii) Thedate for registration to Swish service for companies, iv) The age of the customers, v) Thegender of the customers, vi) The number of transactions, and vii) The transaction date.The estimated parameters for each regressor are studied to evaluate correlations withthe TA. This thesis states that it is possible to construct a model from the regressorsanalyzed, which can predict the amount with an explanation degree of above 85% forfour of the five industries. The model constructed for the motor vehicle industry nevergives satisfactory results and must be further investigated to conclude. / Detta examensarbete är en fallstudie i samarbete med företaget GetSwish AB.GetSwish AB tillhandahåller mobilapplikationen och betaltjänsten Swish, vars syfteär att leverera smidig pengaöverföring för privatpersoner och företag i Sverige. Idagär cirka 80 procent av Sveriges befolkning anslutna till Swish och majoriteten sertjänsten som en självklar del av vardagen. Detta arbete kommer dock endast fokuserapå en liten del av alla transaktioner som dagligen sker mellan privatpersoner ochföretag. Specifikt undersöker denna rapport vilka faktorer som påverkar Swishstransaktionsbelopp till företag inom fem olika branscher. De fem branschernasom studeras är: Sport-, fritids- och nöjesverksamhet; Restaurang-, catering ochbarverksamhet; Detaljhandel utom med motorfordon och motorcyklar; Handelsamt reparation av motorfordon och motorcyklar; och Telekommunikation. Ikombination med en deskriptiv analys och säsongsstudier skapades en multipel linjärregressionsmodell för att utvärdera mönster i transaktionsbeloppet från kund tillföretag inom de olika branscherna. Responsvariablen är det dagliga aggregeradebeloppet och de förklarande variablerna som undersöktes var: antalet anställda,omsättning, datum för registrering till Swish för företag, kundernas ålder och könsamt antal transaktioner och transaktionsdatum. De skattade parametrarna förvarje regressor studerades för att utvärdera magnitud samt positiva eller negativakorrelationer med beloppet. Denna rapport visar att det är möjligt att konstrueraen modell från de analyserade regressorerna som kan förutsäga beloppet med enförklaringsgrad på över 85% för fyra av de fem branscherna och kan användas föratt förutspå beloppen på de dagliga transaktionerna. Modellen som konstruerats förfordonsindustrin gav aldrig tillfredsställande resultat och bör undersökas vidare innanslutsatser dras.