Information Cascades with Stocks and Financial Markets (with applications to lecture references)
An information cascade occurs when individuals make decisions one after the other based only on some limited observable information, with no direct verbal communications between each other. We often model this with a series of individuals that sequentially need to make a decision to accept or reject something (a bet, a purchase, an action). While the first couple of people may make their own decisions based on their personal knowledge, there comes a point where people later in the sequence begin ignoring their personal information in favor of the actions of the people before them and thus make the same choice as them, and thus the cascade begins.
This article and page from Investopedia, a financial website that provides information on investment and financial products, discusses a real-world application of an information cascade, as it is applied to a financial market, or in particular, buying a certain company’s stock. In their example, a beginner or even an average person that buys and sells stocks may take note of the tendencies another person buying stocks may exhibit, and then imitate, simply because they think that this person might make good picks for investments. Perhaps one of these people’s friends observes that they picked that specific stock to invest in, so, thinking that these other people’s opinion matters more than your own, they copy them and also invest in that stock. At this point, an information cascade has begun, and the people investing in that stock didn’t even have much information on how successful the stock could be in the first place; they were simply following the suit of others.
This can go awry if the person whose decisions created the cascade isn’t even that qualified to give stock advice in the first place, or even if they made a poor decision in that one instance. If the stock flops, financial harm can be brought upon all of these people.
This application has several ties to concepts explored by the professors tangentially during lecture. One example brought up was that during staff meetings, as the leader, it might be best to select someone new or inexperience first for any advice or comments; if a more experienced person is selected, then the inexperienced people will tend to copy what the more experienced person said just because they are scared of being wrong, and we end up with a very narrow scope of ideas in the meeting. Calling on newer faculty members thus prevents cascades and circulates more diverse and original ideas, and this is perfectly analogous to the example brought up by the aforementioned article about information cascades in financial markets.
Another in-class example brought up by the professors, which incorporates concepts from cascades as well as prediction markets, was an example about sports betting and horse betting platforms. It was said that in general, if people are aware of the level of expertise of the people involved in betting, and this level of expertise is high, then more amateur betters should know not to place any bets, as they’d likely get defeated by more knowledgable players. We can see that ranges in expertise are thus applicable to both cascades and prediction markets/betting: more knowledgeable players control cascades, and they also can take advantage of inexperienced players in prediction markets. However, inexperienced people need to be in the mix with respects to prediction markets: if they’re not, it’s difficult for the platform to break even, since if many people bet right, they have to distribute more money back to them, with less funds from people that bet incorrectly.
Source: https://www.investopedia.com/articles/investing/052715/guide-understanding-information-cascades.asp