An Experimental Research on Information Efficiency of Securities Markets

YANG Xiao-lan,JIN Xue-jun

DOI: https://doi.org/10.3785/j.issn.1008-942x.2005.06.010

2005-01-01

Abstract:Information efficiency of securities markets is always an important coomponent of financial theories.Although a great many propositions have studied on this area,how to test them with field data is still a big problem.Information efficiency requires to be measured by the degree of price derived from the fundamental value which cannot be evaluated accurately in the real world.Experimental methods allow us to investigate into propositions on the information efficiency in a controlled environment and evaluate the fundamental values accurately. With reference to researches of experimental economics in the West,we apply the computer experimental system developed by Zhejiang University and recruit participants to build simulation laboratory securities markets.We set three sessions of experiments with different information environment including information symmetry,pre-trade information asymmetry and post-trade information asymmetry.Each session consists of three trials.According to the experimental data,we find that when the laboratory environments are information symmetrical,traders cannot use the information to form common rational expectations.It implies that symmetrical information is not enough for traders to form common expectations and markets cannot realize equilibrium instantaneously.Compared with information symmetrical environments,pre-trade information asymmetry will induce prices to derive from fundamental values more seriously and produce more bubbles,while prices in post-trade information asymmetry markets are very close to the fundamental values.Those experimental results imply that information efficiency is reduced by post-trade transparency and improved by pre-trade transparency.We use trader behavior patterns to explain those results and propose that transparency of efficient information is important to securities regulation.