An Information Theoretic Perspective on the Global Financial Crisis

DISCLAIMER: This is a post about the economy, but I am not an economist. I am an information scientist. My understanding of the global financial crisis from the economic perspective is amateur at best. The text of this post represents an informal account of formal models. I will present a more formal version when time allows.

Consider a barter market of people trading goods. Let’s call this a one-level system, because there is only one level of interaction between people, namely, the exchange of goods. Given a current state of the system (who has what goods), it is fairly straightforward to predict the future state of the system (who will have what good three years from now).

Now let’s add money into the picture: consider a money market of people exchanging currency for goods. There are now two levels of interaction between people. First, they have to agree on the price of goods (how much currency to exchange for each good) and then they have to exchange the goods. This system has just one more level than the first but there is a very important difference: due to the presence of the extra level, it is no longer straightforward to predict the future state of the system (who will have how much money and goods) from the current state (who has how much money and goods now). In particular, some people might get very rich and others, very poor, but it’s not clear, who.

Finally, let’s add information into the system. In the money market, people have to know about the price of goods so they know how much currency to exchange for how much good. This information is pretty cheap to exchange, generate, modify, and store, so it has almost no value. But let’s say people begin to place a value on this information: let’s say the piece of information that says bananas are worth $5 right now is itself worth something. Now there are four levels of interaction between people: they have to agree on both the price of information and on the price of goods, and they trade in both information and goods. Note that information is fundamentally different from other goods in this system: it has a very very low cost both to create, exchange, modify, and store, unlike any other good type.

This last system is most like today’s financial markets, where people can trade not just on goods but also on information about how much these goods are worth, or will be worth. It is also difficult to predict the future of this system from its present; however, we actually know one thing about it that we don’t know about the second, namely, the extremely low cost of information. We also know that the information is about what goods will cost how much money, which, as we know from the second system, is very hard to predict. So, it is both difficult to generate accurate information and extremely cheap to generate any information at all. So as long as information has any value at all, people will gravitate towards trading in it, and they will trade in generally inaccurate information.

Now let’s try to imagine what happens in the third system over a long amount of time. At first, a few people will trade in information, and presumably make a lot of money off it, since its cost is so cheap. Other people will imitate these early adopters, and drive the price of information up, but its cost will remain low. So, more and more people will begin to trade in information, and its price will go ever higher, while its cost remains minimal. New people who want to trade in information have two options: buy and sell information that’s already out there, or make up new information. Since the price of information goes up over time, while its cost remains low, the second option is much more lucrative, so people will make up new information all the time. As more information enters the system, more and more of it will be inaccurate. Furthermore, due to the difficulty of predicting the value of goods, this information has the possibility of being wildly inaccurate. The value of information can be much higher, or much lower, than what it is being traded for, and this value can change drastically over time.

So over the long term, most of the people in this system will be trading in information – because it is the most profitable option – that information will have a very high price, and it will be extremely inaccurate. When new information comes to light, the value of old information can change drastically. This can lead to rapid changes in the state of the system (who has how much money at what time). Furthermore, the overall value of information owned by people in the system can change drastically, leading to rapid changes in social welfare. We saw precisely such rapid and drastic changes over the last few years, during the global financial crisis.

Is there a way of preventing such crises? In other words, is there a way of changing things around so we either adopt the first or second system (or some other system of exchanging goods and services), or we shift the rules of the third system so in the long term it doesn’t lead to concentration of trading in volatile information? Let’s examine our options:

1. Use the first or second system of exchange. Possible, but unlikely, since both are less efficient at generating value than the third system, and people really like to generate value (everybody wants to be rich).
2. Use some other not-yet-invented fourth system. Possible, but certainly a more difficult proposition than just altering the third system slightly.
3. Use the third system with some changes:
3a. Eliminate trading on information completely. This is the simplest solution, but it’s equivalent to using the second system, and people would again reject this option due to simple greed.
3b. Make it so information is very accurate. Unfortunately, accurate information (that is, accurate prices for goods) is very hard to come by, and becomes much harder with each new actor and each new good that enters the system. So this choice boils down to not allowing any new actors or goods to enter the system, at which point the market would quickly become stale, not an attractive option.
3c. Increase the cost of information. This requires rebuilding our information infrastructure, as with the existing information infrastructure, it is almost impossible to make it impossible to trade or generate information cheaply. Again, a very expensive proposal.
3d. Regulate the value that can be assigned to information. This is very difficult to enforce.
3e. Add an alternative to trading on volatile information. This alternative will act as a “pressure valve” so that as amount of trading in information increases it becomes more valuable to trade goods, pay, and trade information, for this alternative. At the same time, the alternative should be self-limiting so that people don’t start assigning too much value to it. This is not trivial to design, but once designed, should offer an incentive to behavior that destabilizes the system and leads to rapid drops / rises in value. It should also be possible to provide this alternative in a cheap way.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: