CHARACTERISTICS OF A SIMULATION

Regardless if one is entering into a military simulation or creating a code-based simulation, there are similarities. The starting point for most simulations is the assumptions that go into it. For a military simulation that is preparing for urban warfare, this might include the number of soldiers per unit, the weapons and supplies that each solider carries, the standard and unique training of the soldiers, and the possible buildings, enemies, weather, and so forth that they could encounter. In a financial simulation, such as the corporate default example, you might have characteristics of the companies, such as the industry, regional operating location, historical asset levels, historical liability levels, and so forth.

Once the assumptions of the topic that we are trying to simulate are understood, a method for assembling the system and rules for how the system works are required. In our military simulation example, we would have a training area where the soldiers arrive with all of the training and gear one would expect, and then have an area with buildings and enemies they would expect to face. A mission with an objective would be established, and certain rules might be integrated to help make the simulation as real as possible. For instance, even though a soldier could theoretically leave the simulation area to get around an obstacle, a rule could define the simulation area and state that soldiers are not allowed to go beyond its perimeter. Similarly, in a financial simulation we would need a medium in which to conduct the simulation, which in modern times is done within the confines of a computer application. We program rules to guide our assumptions' behavior through processes that simulate how real-life events might unfold.

Another characteristic of simulations is that they may be repeated to determine varying outcomes. In the military situation, soldiers may choose one path through the buildings in one iteration of the simulation and then choose a different path in another iteration. The outcomes in both scenarios could be markedly different. Similarly, in a financial simulation asset levels for the same company in a future period could be assumed to be different from one simulation iteration to the next. This could mean that the default outcomes are also different.

At the end of the simulation, there should always be an analysis. Multiple aspects of the military simulation would be analyzed, such as speed of completion of the simulation, effectiveness at achieving the mission objective, supplies used, and so forth. In the financial simulation, we would want to see the frequency of companies defaulting, which types of companies defaulted, the characteristics of those companies, the balance of exposures for the ones defaulting, the time at which they defaulted in the future, and so forth.

FIGURE 1.1 Most simulations will follow a similar process of selecting or creating assumptions, constructing a simulation environment with rules, analyzing the outcome, and possibly repeating the process.

Finally, we should be concerned about the validity of our results. Numerous flaws could occur in the construction of the military simulation. Perhaps the individuals posing as enemy soldiers are not as aggressive as in real life or the equipment used is different. In the financial simulation, perhaps we assumed lower correlation than really exists or measured historical volatility wrong. All of these could lead to error that should be taken into account. See Figure 1.1.