Agent-based computational economics

Agent-based computational economics (ACE) is a special way of studying how economies work by looking at them as groups of interacting parts, called agents. These agents are not real people, but computer programs that follow certain rules to make decisions and interact with each other. This helps us understand how whole economies change and evolve over time.

Unlike older economic models that assume people always make the perfect decision, ACE allows agents to have limited knowledge and learn from their experiences. This makes the models more realistic, as it reflects how real people often make decisions based on what they know and can figure out.

ACE uses computer simulations to study complex economic problems that are hard to solve with traditional math. By starting with set conditions, researchers can watch how the economy changes as agents interact and learn from each other. This method is similar to game theory but focuses more on how agents act independently and adapt over time.

Thanks to better computer technology and improved modeling techniques, ACE has become more powerful. It has been used to study many areas, such as how prices are set, energy use, market competition, and economic welfare. Recently, it has also incorporated advanced artificial intelligence to create more realistic agents and predict how they might behave in complex situations.

Overview

In agent-based computational economics (ACE), "agents" are like computer characters that can be people, groups of people, plants, or even transport systems. These agents follow certain rules and interact with each other over time. Scientists set up these characters and watch how they behave without stepping in to change things.

ACE is a special area of study for the Society for Computational Economics, and many researchers from the Santa Fe Institute have helped develop this field.

Agent-based finance

Agent-based computational economics (ACE) has been used to study how prices of things like stocks change. In these models, many people, called agents, try different ways to guess what stock prices will be. They pick the methods that have worked well recently. The success of these methods depends on what’s happening in the market and which methods people are using. These models can show that big rises and falls in stock prices can happen when people switch between different guessing strategies.

Recently, some researchers used this kind of model to suggest that adding new financial tools might make markets less stable. Some papers have even suggested that ACE could help us understand big financial problems, like the 2008 financial crisis. For more details, you can look at the sections on Financial economics § Financial markets and § Departures from rationality.

Agent-based macroeconomics

Agent-Based Macroeconomics (ABM) helps us understand the economy by starting from the very smallest parts. Instead of looking at big, general numbers, ABM looks at many different people, each following their own rules. These people, or "agents," make decisions and learn in their own special ways. This idea comes from the work of Herbert Simon. In ABM, big economic changes happen naturally from how these agents interact with each other, not just because of outside events. ABM uses real-world information to create these rules and also shows how people can be very different from one another. It includes details about how institutions and markets actually work.

Main article: Agent-based model