A blog about using cellular automata to model financial and investment markets.
This blog intends to explore, experiment, and discover the use of cellular automata in modeling financial phenomena. This includes financial markets, trading markets, stocks, exchanges, demand, supply, currency, price fluctuation, and much more. The potential applications of CA far exceed the sparse implementations of CA known today. Conway’s game of life, and other classical CA only attempt to simulate a fundamental set of rules. Not ones which can really relate to the world we live in today, or financial scenarios.
Why pick finance, money, and investment as a topic? Why bother applying automata to a discipline like this one?
Business and financial markets are a great example of a system where past data cannot directly indicate a future pattern. The past can offer clues about, say what stocks might go up or down in price, but it cannot strongly guarantee an outcome. Machine learning and artificial intelligence succeed when the past stays consistent and forms a detectable pattern. However, if the future changes from the past, they will fail. If the future constantly changes, both machine learning and artificial intelligence won’t be useful. They both, by definition, work off the assumption a pattern exists.
Financial markets also prove to be an excellent application for CA modeling because of chaotic behavior. The behavior of, a stock market for example, does not follow any functional domain(security) and range(price). Although a moving average can be calculated over a given interval of days for a stock, whatever approximations made to determine the change from that average fails to consider one incredible aspect:
Changes in financial markets are driven by speculation on the current state of conditions.
At any point in time, the price of a stock is determined by the demand, and the number of shares outstanding. However, what determines the demand, and the shares being bought and sold, is the state of many conditions in a system. Different factors influence the reluctance or preference of traders to buy or sell. These factors range from the individual trader, to how an individual trader views other stocks or other traders, and much, much more.
Cellular automata allow us to model chaotic systems and understand the conditions and forces that drive them to certain states. It gives greater insight than machine learning and artificial intelligence do, because it conveys a greater explanation for the current state beyond patterns and trends. Exploring and investigating the uses of CA in financial modeling will provide a way to study scenarios like market crashes or spikes without having to have the pre-existing data for them, all that’s needed is setting up the automata with the desired properties, members, and conditions.