Numeraire independence and the measurement of mispricing in experimental asset markets

Mispricing (the difference between prices and their underlying fundamental values) is an important characteristic of experimental markets. The literature on the topic consists of many different measures. This state of affairs is unsatisfactory, since it is not clear to which extent results are sensitive to the choice of measure. This paper shows that numeraire independence is an important condition not satisfied by previous measures. Furthermore, under additional assumptions it can be shown that the geometric mean is the only such aggregation function to satisfy numeraire independence. This leads to the proposal of two new measures of mispricing, Geometric Deviation (for overpricing) and Geometric Absolute Deviation (for absolute mispricing). An application illustrates the potential impact of these new measures on previous experimental results.