Quantative Methods
Simulation Methods
Monte Carlo Simulation
Definition and uses
Monte Carlo simulation involves generating
random samples from a probability distribution
using computer software.
Widely used for risk and return estimation in
investment analysis and valuing complex
securities.
Steps in Model Sensitivity
Analysis
1. Specify underlying variable(s) and initial
values. Define the value of the option at maturity
(CiT) for each simulation trial.
[Link] the time period and distributional
assumptions for key risk factors.
[Link] a model to estimate underlying variables
based on random observations.
[Link] the option value at maturity (CIT)
and discount it to obtain CiO.
[Link] the average stock price and the value
of the contingent claim.
Accuracy
Precision depends on the number of trials; more
trials increase accuracy.
Random Number Generation
Generate a uniform random number (T)
between 0 and 1.
Use the inverse of the cumulative distribution
function F(x) to obtain a random observation on
variable X.
Role in Analysis
•Monte Carlo simulation provides statistical
estimates, not exact results.
•Used as a complement to analytical methods in
modeling and analysis.
Bootstrap Resampling
Resampling
A technique for statistical inference of population
parameters by repeatedly drawing samples from
the observed data.
Used in situations where traditional analytical
formulas are difficult or impossible to apply.
Bootstrapping
• A specific type of resampling that draws samples
(with replacement) from the observed data.
• Allows for statistical inference without relying
on traditional formulas.
• Mimics random sampling from a population
using the observed sample as the population itself.
Differences between
Bootstrapping and
Monte Carlo Simulation
Source of Random Variable:
a) Bootstrapping: Drawn from the original data.
b)Monte Carlo Simulation: Drawn from a known
probability distribution.
Purpose:
a)Bootstrapping: Infer statistical distribution
parameter values (e.g., mean, variance) based on
the observed data.
b)Monte Carlo Simulation: Generate random
data based on known statistical distribution
parameters.
Steps in Implementing
Simulation Using
Bootstrapping:
[Link] the quantity of interest and set
starting values.
[Link] the time grid and subperiods.
[Link] the data generation method based
on observed stock price changes or returns.
[Link] stock prices by drawing random
values and converting them to a sequence of stock
prices.
[Link] the average stock price and the
contingent claim's value, then calculate its present
value.
[Link] steps 3 and 4 muftiple times to obtain
summary values and statistics.
The mean value from all trials is the bootstrap
estimate for the contingent