On the impact of semantic framing in experimental asset markets

This paper studies how semantic framing affects price efficiency. In an experimental asset market subjects are provided with an overly positive, overly negative or no description of the asset traded. This description provides no information about the asset's value. Prices are neither lower when subjects are negatively framed nor higher when subjects are positively framed compared to a treatment without framing. Furthermore, learning effects and price dynamics are comparable across treatments. I discuss two possible explanations from individual choice experiments, namely, that completely described problems and ratings and judgments are less prone to framing. Furthermore, I discuss an alternative possible explanation that asset markets are able to prevent biases to occur.