Price limits and same-day trading restrictions (the T+1 rule) are two market regulations that China Securities Regulatory Commission impose in Chinese A-share stock markets to improve price discovery and decrease volatility, but the effectiveness of these two regulations is debated. This thesis examines the effectiveness of price limits and the T+1 rule using an asset market experiment.
The main innovation of the experiment is that the dividend process follows a Markov process. The new dividend process makes this experiment to be the first asset experiment where participants need market information to predict the share's fundamental price in the future. This dividend process also makes the share's fundamental value to increase in the long-run, and to exhibit correlation in the short-run, which is more realistic.
The main experimental findings are the following. First, this new market environment is not immune to bubbles, but nearly half of these bubbles are negative. The traditional strong convergence of prices to fundamental values in terminal periods is not observed in some markets. Experience decreases bubble size in general, but negative learning is also observed. Second, the T+1 rule turns the negative bubble in the benchmark into a positive one, but price limits do not significantly affect miss-pricing. Third, participants respond to market information imperfectly, and price limits impede the price to fully absorb new market information.