iiABSTRACTBehavioral finance has been a popular research subject for a while and therefore the understanding of human behavior when it comes to private financial investments has increased. When comparing human behavior to the financial theories one can conclude thatthe assumption of perfect and efficient markets with fully informed and rational investors is not realistic. This study has therefore looked at the investor rationality when choosing which index fund to invest in. Index funds are to a large degree used asa savings tool for either pensions or other specific purposes. It was therefore interesting to look at the behavior of Swedish investors buyingthe Swedishindex funds available in Sweden with a quantitative analysis of the relationship between flow and other features of index funds. The dependent variable reflecting rationality was the fund flow and the independent variables were return, tracking error, size, fee and risk. No previous studies have been made on the investor rationality regarding index funds in the Swedish market, although similar studies havebeen done on the American S&P 500 investors. 17 index funds were included in this study, which is the whole population of index funds following Swedish indices available in Sweden at the point of time when this study was conducted. From this population funds that had been available for more than 3 years was chosen since we wanted to look at the behavior based on a longer time span than one year. In the end, 17index funds with 51observations was included in the study.Five hypotheses were created and tested of whichtwowereaccepted. From the regression model we found that return and standard deviation (SD) weresignificant andhadpositiverelationshipswith the fund flow. This implies that Swedish investors are rational to some degree but not fully rational since they are not taking any of the other variables into account which a rational investor ought to consider. It is therefore useful information for both investors and fund companies to see which factors weight in the most and how rational the behavior is. Conclusions from this study is that Swedish investors are subject to the index fund rationality paradox to some degree and the rational choice theory applies to some extent. One has to fully consider the outcomes of an action and base the decisionon utility maximization that the outcome will give one. To act fully rational is hard even for the most aware investor and even harder for an ordinary investor with gaps in knowledge and limited resources to information.

Nous étudions le comportement des investisseurs individuels sur les marchés financiers en combinant les réponses au questionnaire MiFID et les données bancaires de plus de 98,000 clients d’une grande banque Européenne. Tout d’abord, nous étudions la participation sur les marchés actions. Nous montrons que la tolérance au risque et les attitudes face aux pertes auto-évaluées des clients sont de forts prédicteurs de l’investissement en actions tout en contrôlant les déterminants classiques. Puis, dans le cadre de la comptabilité mentale, nous créons une typologie d’objectifs mentaux et montrons que les décisions financières des clients sont cohérentes avec leurs objectifs mentaux. Enfin, nous analysons le comportement des investisseurs détenant directement au moins une action étrangère. Nous montrons qu’ils détiennent des portefeuilles d’actions plus diversifiés que les investisseurs domestiques. Ces investisseurs sophistiqués sont plus tolérants au risque, moins sensibles aux pertes et plus instruits en matière financière mais sont sujets au biais national. / We study retail investors’ behavior on financial markets by combining the MiFID questionnaire answers and banking records of more than 98,000 retail clients of a large European retail bank. First, we study stock market participation. We show that retail clients’ self-assessed risk tolerance and attitudes towards losses are strong drivers of stockholding while controlling for classical determinants. Second, under the mental accounting framework, we derive a typology of retail client mental goals and show that retail clients’ actual investment decisions are consistent with their mental goals. Finally, we analyze the behavior of investors directly holding at least one foreign individual stock. We show that they hold more diversified stock portfolios than domestic investors. These sophisticated investors are more risk tolerant, less sensitive to losses and more financially literate but are subject to the home bias.

<p><strong>Syfte:</strong></p><p>Uppsatsens syfte är att testa den svaga formen av marknadseffektivitet på ett antal aktiemarknader genom teknisk analys i form av system med glidande medelvärden.</p><p> </p><p><strong>Teoretisk referensram:</strong></p><p>Den teoretiska referensramen utgår ifrån den effektiva marknadshypotesen men omfattar även behavioral finance och teknisk analys.</p><p> </p><p><strong>Metod:</strong></p><p>Uppsatsen är av förklarande karaktär med ett positivistiskt synsätt och ett deduktivt angreppssätt. Testen av marknadseffektivitetens svaga form sker genom en kvantitativ undersökning av sekundärdata. I undersökningen genomförs en hypotesprövning där en passiv investeringsstrategi jämförs mot aktiva investeringsstrategier byggda på system med glidande medelvärden. Undersökningen omfattar åren 2002-2009 och sker på tio väletablerade aktieindex, i Norden samt på tillväxtmarknader och på etablerade marknader, vilket ger cirka 17 500 observationer.</p><p> </p><p><strong>Empiri:</strong></p><p>Hypotesprövningen visar att system med glidande medelvärden har kunnat skapa riskjusterad överavkastning samt kunnat förutse framtida kursrörelser på aktiemarknaderna i Danmark, Norge, Sverige och Hong Kong. På aktiemarknaderna i Finland, Brasilien, Indien, Japan, Tyskland och USA har systemen som testats inte kunnat skapa riskjusterad överavkastning och/eller inte kunnat förutse framtida kursrörelser.</p><p> </p><p><strong>Analys:</strong></p><p>Det finns flera faktorer som påverkat resultaten i olika grad vilket gör att endast indikationer om grad av effektivitet på de olika marknaderna ges i slutsatsen.</p><p> </p><p><strong>Slutsatser:</strong></p><p>På aktiemarknaderna i Danmark, Norge, Sverige och Hong Kong ges indikationer om att kraven för svag marknadseffektivitet inte uppnåtts. Medan det på aktiemarknaderna i Finland, Brasilien, Indien, Japan, Tyskland och USA ges indikationer om att kraven för svag marknadseffektivitet varit uppfyllda.</p>

<p>Bakgrund: När ingen kan förklara vad som händer på börsen har känslorna tagit över, börsens egen psykologi. Mycket kan förklaras logiskt med hjälp av så kallade fundamentala faktorer. Dessa faktorer kan inte förklara alla kraftiga svängningar som sker på börsen, men kanske kan psykologiska faktorer bidra till att förklara. </p><p>Syfte: Syftet med denna uppsats är att undersöka om psykologiska faktorer påverkar priset av en aktie och identifiera vilka dessa faktorer i så fall huvudsakligen är. Avgränsningar: Uppsatsen är avgränsad till att enbart behandla noterade aktier som handlas på en börs öppen för allmänheten. Dessutom är det en teoretisk uppsats, vilket innebär att vi inte genomfört några egna empiriska undersökningar. Genomförande: Vi har valt att göra en teoretisk uppsats. </p><p>Resultat: Vi har funnit att psykologiska faktorer påverkar en aktievärdering. Dessa faktorer kan vara flockbeteende, förlustaversion, anchoring samt över- och underreaktion på ny information.</p>

<p>Background: An evaluation of analysts´ forecasting ability is interesting since their estimates constitute an important part in stock valuation and investment decisions. The recent years´ development in the stock market has lead to criticism of analysts’ deficient forecasts. </p><p>Purpose: The purpose of this thesis is to evaluate analysts´ forecasting ability concerning companies quoted at Stockholmsbörsen between 1987 and 2002. We also intend to discuss possible explanations for analysts’ behavior in case of deficient accuracy. </p><p>Method: Regression analysis is used to compare consensus estimates of earnings per share to actual earnings per share. We attempt to investigate the existence of a relation between forecasting ability and forecast horizon, the volatility at Stockholmsbörsen and the industry in which the firm operates. Behavioral finance and economic incentives is used to discuss the most convincing explanations to analysts´ behavior in cases of deficient accuracy. </p><p>Result: The study indicates over optimistic forecasts and overreaction to earnings information. Analysts tend to give more accurate forecasts closer to earnings announcement. We believe that herding, economic incentives and the fact that analysts get information from the company explains a significant part of analysts’ behavior. In addition, the study shows a possible relation between more accurate forecasts and lower volatility. Concerning industries we find stronger overreaction in healthcare and heavy industry. The study shows the most exceptional optimism in consumer goods/services and IT/telecom.</p>

Bakgrund: När ingen kan förklara vad som händer på börsen har känslorna tagit över, börsens egen psykologi. Mycket kan förklaras logiskt med hjälp av så kallade fundamentala faktorer. Dessa faktorer kan inte förklara alla kraftiga svängningar som sker på börsen, men kanske kan psykologiska faktorer bidra till att förklara. Syfte: Syftet med denna uppsats är att undersöka om psykologiska faktorer påverkar priset av en aktie och identifiera vilka dessa faktorer i så fall huvudsakligen är. Avgränsningar: Uppsatsen är avgränsad till att enbart behandla noterade aktier som handlas på en börs öppen för allmänheten. Dessutom är det en teoretisk uppsats, vilket innebär att vi inte genomfört några egna empiriska undersökningar. Genomförande: Vi har valt att göra en teoretisk uppsats. Resultat: Vi har funnit att psykologiska faktorer påverkar en aktievärdering. Dessa faktorer kan vara flockbeteende, förlustaversion, anchoring samt över- och underreaktion på ny information.

Background: An evaluation of analysts´ forecasting ability is interesting since their estimates constitute an important part in stock valuation and investment decisions. The recent years´ development in the stock market has lead to criticism of analysts’ deficient forecasts. Purpose: The purpose of this thesis is to evaluate analysts´ forecasting ability concerning companies quoted at Stockholmsbörsen between 1987 and 2002. We also intend to discuss possible explanations for analysts’ behavior in case of deficient accuracy. Method: Regression analysis is used to compare consensus estimates of earnings per share to actual earnings per share. We attempt to investigate the existence of a relation between forecasting ability and forecast horizon, the volatility at Stockholmsbörsen and the industry in which the firm operates. Behavioral finance and economic incentives is used to discuss the most convincing explanations to analysts´ behavior in cases of deficient accuracy. Result: The study indicates over optimistic forecasts and overreaction to earnings information. Analysts tend to give more accurate forecasts closer to earnings announcement. We believe that herding, economic incentives and the fact that analysts get information from the company explains a significant part of analysts’ behavior. In addition, the study shows a possible relation between more accurate forecasts and lower volatility. Concerning industries we find stronger overreaction in healthcare and heavy industry. The study shows the most exceptional optimism in consumer goods/services and IT/telecom.

INFORMATION IN THE FINANCIAL NEWS: EFFECT OF MARKET COMMENTARY ON STOCK MARKET PERFORMANCE GIRAY, Aynur MBA, Department of Business Administration Supervisor: Dr. Seza Danisoglu September 2012, 72 pages This paper studies the effect of investment sentiment on asset prices. A sentiment proxy is calculated by performing content analysis on the Wall Street Journal&lsquo / s

INFORMATION IN THE FINANCIAL NEWS: EFFECT OF MARKET COMMENTARY ON STOCK MARKET PERFORMANCE GIRAY, Aynur MBA, Department of Business Administration Supervisor: Dr. Seza Danisoglu September 2012, 73 pages This paper studies the effect of investment sentiment on asset prices. A sentiment proxy is calculated by performing content analysis on the Wall Street Journal&lsquo / s

INFORMATION IN THE FINANCIAL NEWS: EFFECT OF MARKET COMMENTARY ON STOCK MARKET PERFORMANCE GIRAY, Aynur MBA, Department of Business Administration Supervisor: Dr. Seza Danisoglu September 2012, 73 pages This paper studies the effect of investment sentiment on asset prices. A sentiment proxy is calculated by performing content analysis on the Wall Street Journal&lsquo / s