Artificial Intelligence can Reduce Information Asymmetry
In Tshilidzi Marwala and Evan Hurwitz’s study regarding information asymmetry, they made an interesting discovery that artificial intelligence can be used to greatly reduce the amount of asymmetric information, thus enhancing efficiency in trades in the market while decreasing the volume of trades.
Information asymmetry is a common phenomenon in a market due to the unequal amount of knowledge about the trading item. For example, frequently sellers know the quality of the items they sell and have reasonable expected values from selling the items, while the buyers have no way to identify the quality of the item. In a market that has high-quality and low-quality items, knowing the percentage of each in the market and having in mind the prices they will be willing to pay for a high- and low-quality item, buyers will pay the expected value of the item regardless of quality since they can’t tell the quality of the items. Thus if this single price is still higher than the value of sellers of the high-quality item, then all sellers will be willing to sell; otherwise only the sellers of the low-quality items will be willing to sell, and the high-quality items are driven out of the market. If both sellers and buyers are humans, it is very likely that the buyer is unable to gather enough information on the quality of the trading item from the seller, and it’s very possible that he or she makes the decision to purchase the item irrationally. So the trading occurs when the sellers feel like they are able to make a profit, and the buyers think that they are saving money from paying less. Therefore, although the market is very inefficient since the trade happens under the condition of lack of knowledge of the quality of the items from the buyers, there will still be a high volume of trading happening.
If instead of human agents, the buyers and sellers are artificially intelligence, then the assumption of lack of ability to gather information will be meaningless. Both buyers and sellers will be able to have as much information as possible on the trading item from different means, and both will not be willing to pay extra money for a given item knowing the quality of the item. Thus the market is very efficient, since the sellers will not be willing to sell for a price below the value of the items, and the buyers, having known the quality of the items, will not be willing to pay a price higher than the value of the items. Using artificial intelligent agents reduces the amount of information asymmetry in the market, and makes the market resemble more closely to a market with symmetric information. However, it is not guaranteed that information asymmetry will disappears in this market, but it is clear that it will be reduced. In this case, Marwala and Hurwitz argue that the trading volume will decrease due to the fact that there is no room for both parties to make profit since they know the true value of the item.
Link to the article: http://arxiv.org/pdf/1510.02867.pdf.