Subject 1. Steps in simulation
1. Determine "probabilistic" variables. Focus on a few variables that have a significant impact on value.
2. Define probability distributions for these variables.
There are three ways to develop distributions:
- Historical data. This approach is possible for variables that have a long history and reliable data over that history.
- Cross-sectional data.
- Statistical distribution and parameters.
3. Check for correlation across variables. For example, stock prices tend to go up on average. Are two variables (e.g., stock A price and stock B price) correlated? If so, pick only one, or build the correlation into the simulation.
4. Run the simulation. The number of simulations you run is determined by:
- Number of probabilistic inputs.
- Characteristics of probability distributions.
- Range of outcomes.
Practice Question 1In developing a simulation model, you should be concerned if you find that two variables:
I. have a strong, positive correlation.
II. have zero correlation.
III. have a strong, negative correlation.
A. I only
B. I and III
C. I, II and IIICorrect Answer: B
A strong correlation can be either a positive or a negative correlation.
Practice Question 2Your company is considering starting a new production division. In the simulation model one probabilistic variable is future sales. The number of simulations will be largest if the future sales amount is assumed to:
A. be normally distributed with a mean of 50k and a standard deviation of 10,000.
B. have a uniform distribution.
C. be based on historical sales distributions of competitors in the industry.Correct Answer: C
Future sales amounts based on the historical sales distributions of competitors will indicate the greatest diversity in future sales and thus the number of required simulations will be the largest.
Study notes from a previous year's CFA exam:
1. Steps in simulation