No market for lemons: artificial intelligence and the markets
Before the 1960s, not many economists thought information of any kind would make or break markets and didn’t put much effort into researching the roles of information.
A market consists of sellers and buyers. A seller may sell goods or services, and sellers and buyers have to take decisions about prices and conditions based on the information they have at hand. Almost all the time, the information at hand is uneven. One agent will always have more information than someone else. This particular effect is called information asymmetry, and causes distortion in the market.
Nobel prize for studying lemons
If a buyer or seller tries to manipulate a decision in his favour, the other party can try to signal right or wrong information. As a defense to this action, the counterparty of agents can try to screen for correct information to reduce the impact of asymmetric information on decisions.
Before the 1960s, not many economists thought information of any kind would make or break markets, and didn’t put much effort into researching the roles of information. In the late 1960s, George Akerlof wrote ‘The Market for Lemons’, which did just that - study the effect of information in markets and won a Nobel prize for economics in 2001.
Journalists and economic audience, in general, were shocked to see that you can get a Nobel prize just to observe the markets. The lemons paper that introduces the concept of information asymmetry was even a perfect description of a used car market that was used in the paper. Yet the findings of the paper were revolutionary and made many other economics models at the time redundant.
“This “information asymmetry” between buyers and sellers kills the market.”
Markets selling lemons
Let us see how information asymmetry changed economics and transactions. In an era where there is such a huge market for used things, such economics proves more than useful.
For example, the value of a good used refrigerator – let’s call it a “peach” – is Rs 3,000. A malfunctioning used refrigerator — or a “lemon” — is worth only Rs 1,500. Here, if buyers are able to identify the peaches from lemons correctly, the trade in both will rise. But, in reality, we, as used refrigerator buyers, are often unable to differentiate between good and bad refrigerators.
To account for this risk, what a buyer would do is cut his offer for a refrigerator to Rs 2,200. The buyer does this because he is unable to say for sure what he is buying - a peach or a lemon. However, a seller who knows for sure that he is selling a good refrigerator, or a peach, will reject the offer of Rs 2,200. As a result, a buyer will face an “adverse selection”. He would know that those sellers willing to sell for Rs 2,200 are selling lemons or bad refrigerators.
Great buyers can foresee all these problems, knowing they will only be sold “lemons” if they spend Rs 1500 for a refrigerator. At this price point, most of the “peaches” stay unsold. The buyer would have been happy to pay Rs 3000 for a good refrigerator, but unfortunately, he didn’t have the ability to judge the product at first glance. This “information asymmetry” between buyers and sellers kills the market.
Signals and Screens
You can take this idea of information asymmetry and apply it to any market. In the labour market, educational degrees and certificates are a signal for recruiters that the candidate is a peach rather than a lemon, and the particular candidate is eligible for a good salary.
At the other side of the table are recruiters, who know they will get all types candidates ranging from good to bad. They have to have the capability to screen candidates based on the information at hand, and still make salary offers well knowing that they might make mistakes.
Some of the undesirable consequences of information asymmetry are moral hazards, the monopoly of information and adverse selection. To think about it clearly, the difference in perception of goods and products is the basis of trade. Hence, we can safely assume and assert that markets are mostly run and ruled by biases and perceptions of participating buyers and sellers.
“An assumption that we can hold for sure is that Artificial Intelligence is making markets more quality oriented because of its ability to detect fakes and sub-standard products, hence making it difficult for lemons to exist in the market.”
Artificial intelligence and the markets
What happens when we introduce Artificial Intelligence in the market. Slowly, intelligent machines replace, or augment, humans and in effect become participants in the market. These machines have larger processing and logical capabilities than humans, and are able to digest large amounts of information and access data. Additionally, large amounts of data, better data science techniques, and advanced intelligent algorithms have only hastened and improved this process.
An efficient market hypothesis is a theory proposed by Nobel Laureate Eugene Fama, which states the market incorporates all the information such that it is impossible to beat the market. So automatically, what the hypothesis suggests is that the only way to beat markets is to make high-risk transactions. It is also a given that humans do not act rationally at all times.
Suppose we have a market where most or all participants are intelligent agents. The intelligent agents will dig for and make use of information from a variety of sources, the internet and experiences to make a decision. We can safely assume that their markets would be more rational than markets where humans are participating.
This will be a case because machines are more rational than humans. An absence of irrational elements in the market will create a largely efficient and rational market. Markets will be largely free of biases and perceptions. The traffic of transactions that will take place in this kind of market will be greatly reduced due to the reduction in information asymmetry.
This hypothesis or a scenario arises when intelligent agents augment or take over markets and become primary traders in the markets. But in the present situation, where AI agents are slowly being used as tools by humans to expand their information base or assist them in trades the exact effect of AI remains to be studied in detail.
An assumption we can hold for sure is that AI is making markets more and more quality-oriented because of its ability to detect fakes and sub-standard products. In the near future, I would be excited to see. Will there remain no market for lemons?
(Disclaimer: The views and opinions expressed in this article are those of the author and do not necessarily reflect the views of YourStory.)