Abstract: Based on market microstructure theory, a Kyle-type market maker model is developed to explore the completion between traders who bear cognitive biases on the same private information they both observe and the consequent impacts on market outcomes. Firstly by analyzing informed traders’ trading pattern and asset pricing in the base model where both informed traders process the private information rationally and correctly, it is found that informed traders adopt identical trading strategy in equilibrium and divide profits equally. Then, the main model assumes that one of the two informed traders is rational and has objective and right view on the information while the other one is not rational and have subjective bias on that information, either over-optimistic or over-pessimistic. On the premise of those cognitive biases, these informed traders’ trading strategies, competition effect between them and the consequent asset pricing are examined. Then, it compares informed traders’ strategic behaviors, and the market outcomes of asset pricing with and without cognitive biases, and finds that a unique equilibrium exists when and only when the cognitive biases are within certain limits, or else the market breaks down. In equilibrium, the rational informed trader trades on this private information only, whereas the subjective informed trader trades not only on this private information but also on the difference between his views on the shared information and the rational informed trader’s. As a result, the rational trader’s trading intensity on the private information, expected trading volume, and expected profits are larger(smaller) than those of the subjective informed traders when he is over-pessimistic (over-optimistic), respectively. That is, the over-pessimistic (over-optimistic) informed trader competes against the rational informed trader in this trading game and gains more(less) than the rational informed trader does due to his cognitive biases. In respect to market liquidity and efficiency for price discovery, they are smaller when the subjective informed trader has pessimistic (optimistic) cognitive biases than those when there are no cognitive biases, respectively. Furthermore, the relations between cognitive biases and asset price bubbles are explored and it is found that informed trader’s over-pessimistic (over-optimistic) bias would lead to underpricing (overpricing) and thus trader’s optimism would generate price bubble gradually.

Key words: strategic trading; over-pessimistic/optimistic biases; limited competition effect; price bubble