|New Mexico Supercomputing Challenge|
Challenge Team Interim Report
In our problem we are modeling two economic systems. One economy will be controlled by a fixed traditional fiscal and monetary system which monitors only the GDP, and changes the interest rates. The second system that controls the second economy will monitor all parts to the economy that interact in the program, such as GDP, money market, labor output, etc. The second economy will be controlled by a genetic algorithm that monitors each of the individual parts to the economy and makes changes accordingly.
We are programing this model in C. The first step is to make a very simple very basic working economy. This simple economy will simply be a combination of the different parts to the economy that interact whenever a change is made to the interest. The first three interacting components will be interest, money market, and GDP. We will from those first three add on supply and demand, labor output, and labor supply market. After we get a working simple economy that responds to change working we will program the fiscal and monetary system, which for the most part will be a large list of condition statements to determine how much to change, increase or decrease, or whether to make any change at all to the interest rate. For the second system we will randomly create several codes to control the interest rate looking at each individual part of the economy and looking backwards to see how much the interest would have to be changed and in what direction to create the recommended change in that component, which would to generate the correct change in the economy. Each economy is made into a structure. Each piece of the economy has one function that passes at least one value necessary to recalculate that part of the economy. Each of the randomly created systems will be tested. We will take the codes that worked best and mix and match code among them in different fashions. Once we get what we think is the ideal code or the best working for this system we will just let it run against the set fiscal and monetary policyto see which one works better and by how much.
Progress has been slow in the actual project. Up until very recently I have just been doing research, mostly a lot of reading on economics. I have also needed to learn a lot about math as well. We have yet to make a working model. Another thing I have been doing is searching for time lags between when different parts of the economy are affected after another. We are modeling an economy assuming that it isn't at the present time. Just recently with the aid of computers the time lags have been reduced from months to a matter of minutes. We feel that it would be more interesting with the real time lags. As of now I am starting to really get started on the programming.
In the results it is hard to tell what will happen. We expect the computer generated system will hold its own or even do better than the original fiscal and monetary system. This depends obviously, a lot, on how well we program the fiscal and monetary system. With all of the other available data to the computer generated code it should be able to maintain better control over the economy.
For questions about the Supercomputing Challenge, a 501(c)3 organization, contact us at: consult1516 @ supercomputingchallenge.org
New Mexico Supercomputing Challenge, Inc.
80 Cascabel Street
Los Alamos, New Mexico 87544