Information Cascades and Bubbles

Investment markets are highly determined by the opinion of a goods valuation in the eyes of others. For this reason, information cascades can be highly influential, especially when it comes to the formation of a bubble. Bubbles are inflated by the speculation that comes from one investor thinking not that a good is worth more than it is priced in a real value kind of way, but that someone else will buy it for more than it is priced sometime soon. This article from the Economist discusses this in relation to a Netflix stock bubble that occurred about two years ago. A more recent example (and one with a much shorter time frame) is the Bitcoin bubble and collapse which took place over the course of about 3 weeks. Both of these rises in value were driven by reporting of the rapid rise in value taking place. This signals to others that these are “good” investments and assure them that others will think so when the time comes to sell. At first, rising trade numbers alone is enough to raise the value. 

This is ultimately unsustainable as eventually bubbles reach a point where some people turn away and refuse to push the price higher. This is another signal which kicks off a frenzy of selling which deflates the price, generally past the levels of actual valuation as calculated by other means. What would be interesting to examine is whether a proportion level of say price to trade volume, share amount or some other metric could be calculated for the bubbling and collapsing tipping points. For the price to keep rising, investor number will have to rise, but is a collapse triggered by too many people signaling “Sell” or not enough continuing to buy? More often than not, investing is a game of expectation identification, being able to put any kind of coefficient to the cascade would give someone a significant leg up in knowing when to hop on the new fad, and when to get off before it’s too late.


1 thought on “Information Cascades and Bubbles”

  1. This is an interesting read. I work in finance and was actually just discussing the Netflix bubble the other day with one of our firms traders, he was basically upset that he didn’t see the bubble coming and couldn’t take advantage of it. Im no sure how one would form a coefficient for a bubble though. Especially since there are so many causes of bubbles, how could there be one uniform coefficient to explain them all? For instance, a stock market bubble like you described can be based off of the amount of people trading and intrinsically over valuing a stock. More traders could see the rise in volume, and price and base their own valuation off of the valuations of the traders who have been active in the stock. This cascade could prove detrimental if the grounds for the valuation of the stock is unfounded, and it is being over valued market wide. Which then triggers a fire sale of the stock and people lose money. But in the case of a nations housing bubble the causes of the burst can vary. Take Spain for instance, Spain experienced a housing boom, and this caused spanish wages to inflate exponentially relative to other euro zone countries. Then the financial crisis happened, and home sales went stagnant due to wage inflation. Spain’s other means of production such as manufacturing were uncompetitive because of the inflation and the housing bubble burst. Because of the euro Spain could not do anything about its uncompetitive wages and is still in bad shape today. So I guess what I am saying is that the coefficient would have to vary from case to case when dealing with bubbles, which may make it too difficult to even consider adopting.

Comments are closed.