# The advantages of simulation modelling

This entry is part 6 of 6 in the series Monte Carlo Simulation

All businesses need the ability to, if not predict the future; assess what its future economic performance can be. In most organizations this is done using a deterministic model, which is a model which does not consider the uncertainty inherent in all the inputs to the model. The exercise can best described as pinning jelly to a wall; it is that easy to find the one number which correctly describes the future.

The apparent weakness of the one number which is to describe the future is usually paired with so called sensitivity analysis. Such analysis usually means changing the value of one variable, and observe what the result then is. Then another variable is changed, and again the result is observed. Usually it is the very extreme cases which are analyzed, and some times these sensitivities are even summed up to show extreme values and improbable downsides.

Such a sensitivity analysis is as much pinning jelly to the wall as is the deterministic model itself. The relationship between variables is not considered, and rarely is the probability of each scenario stated.

What the simulation model does is to model the relationship between variables, the probability of different scenarios, and to analyze the business as a complex whole. Each uncertain variable is assessed by key decision makers giving their estimates for

• The expected value of the variable
• The low value at a given probability
• The high value at a corresponding probability level
• The shape of the probability curve

The relationship between variables is either modeled by its correlation coefficient or a regression.

Then a simulation tool is needed to do the simulation itself. The tool uses the assigned probability curves to draw values from each of the curves. After a sufficient number of simulations, it will give a probability curve for the desired goal function(s) of the model, in addition to the variables themselves.

As decision support this is an approach which will give you answers to questions like:

• The probability of a project NPV being at a given, required level
• The probability of a project or a business generating enough cash to run a successful business
• The probability of default
• What the risk inherent in the business is, in monetary terms
• And a large number of other very useful questions

The simulation model gives you a total view of risk where the sensitivity analysis or the deterministic analysis gives you only one number, without any known probability. And it also reveals the potential upside which is in every project. It is this upside which must be weighted against the potential downside and the risk level which is appropriate for each entity. The S@R-model

The S@R-model is simulation tool which is built on proven financial and statistical technology. It is written in a language especially made for modelling financial decision problems, called Pilot Lightship. The model output is in the form of both probabilities for different aspects of a financial or business decision, and in the form of a fully fledged balance sheet and P&L. Hence, it gives you what you normally expect as output from at deterministic model, and in addition it gives you simulated results given defines probability curves and relationships between variables.

The operational part of the business can be modeled either in a separate subroutine, or directly into the main part of the simulation tool, called the main model. For complex goal functions with numerous variables and relationships, it is recommended to use the subroutine, as it gives greater insight into the complexity of the business. Data from the operational subroutine is later transferred to the main model as a compiled file.

Series Navigation<< Decisions – Criteria for selection

Tags: 