The equity premium puzzle and myopic loss aversion in Europe: A comparative empirical study of the equity premium and non-standard investor preferences in Scandinavia and selected European countries

Anders Frank Christensen

Student thesis: Master thesis

Abstract

The relation between asset returns is at the heart of economic and financial theory, with the central premise being that market prices, and returns, are derived from the interaction between rational and utility maximizing agents. In 1985, Mehra & Prescott published an influential study that illustrated how the observed equity premium in the US had been substantially above what standard utility theory would predict. This was labeled the equity premium puzzle, and it has been a source of debate ever since. This paper investigates a potential approach to solving the puzzle, namely myopic loss aversion, as presented by Benartzi & Thaler in 1995. The approach builds on elements from the field of behavioral finance, along with a non-standard investor preference structure developed Tversky & Kahneman (1979; 1992) as part of their renowned prospect theory. Firstly, the study extends the original analysis of Mehra & Prescott (1985) to the three Scandinavian countries, along with six other EU countries of varying size, selected to provide a strong and robust comparative European context. For the sample period 1970-2010, it is found that an equity premium puzzle is present throughout the nine countries included in this study. Next, the study applies the myopic loss aversion approach of Benartzi & Thaler (1995), to investigate whether it can serve as a broadly founded solution to the observed puzzles. The analysis shows that this is not the case. While myopic loss aversion can explain the observed equity premium in Sweden, and to some extent in Norway and Denmark, the approach fails to provide a viable solution for the six non- Scandinavian countries in the sample. Consequently, the overall conclusion from the comparative analysis is that myopic loss aversion cannot provide a solution to the equity premium puzzle observed within the nine countries over the period 1970-2010. To understand these results, sensitivity tests are performed for changes in the key model input, loss aversion, as well as the sample period. While changes in loss aversion are shown to have considerable impact on the results, the magnitude of change needed for the approach to work, makes the solution implausible. Meanwhile, the change in sample period provides a stronger picture, as an investigation of the return data shows a change in the pattern of equity returns over the last 10 years. Adjusting for this period provides strong support for the myopic loss aversion approach across all countries analyzed. However, as 10 years cannot be described as an outlier, the overall conclusion remains unchanged. Within this comparative analysis, myopic loss aversion fails to provide a convincing solution the equity premium puzzle.

EducationsMSc in Applied Economics and Finance, (Graduate Programme) Final Thesis
LanguageEnglish
Publication date2011
Number of pages97