All papers
How Does Learning Makes Market Efficient? The Role of Market Microstructure
The Zero-Intelligence (ZI) agents in the double auction market ensure efficiency by
providing the highest surplus and trade volume but have been unable to ensure price
convergence. Afterwards, focusing on the importance of rationality, the ZI agents are
allocated limited intelligence to learn from the market and improve their behaviour over
time. These agents with limited learning power are known as Zero-Intelligence Plus (ZIP)
agents, and they ensure allocative and volume efficiency and make price convergence
possible in the presence of the Marshallian trading sequence. However, these ZIP agents
possess assumed learning rules to guarantee the desired efficiency level, and these rules
are not based on how humans learn in real markets. To overcome this downside, this
paper replaces the assumed learning rule of ZIP agents with new learning rules extracted
from real human behaviour, i.e., introducing ZIPH agents. In addition, the market
efficiency of these ZIPH agents is examined for different trading sequences and varying
demand and supply schedules, i.e., symmetric, asymmetric, and box-shaped demand
and supply schedules.