Updated for the 2026 CFA® Level I curriculum.
Investment decisions often depend on variables that cannot be known in advance. Future returns, interest rates, inflation, cash flows, and default events may each take a range of possible values.
Monte Carlo simulation helps analysts study that uncertainty by generating many possible scenarios. For CFA Level I, focus on how the process works, what its output can tell you, and how the model’s assumptions affect the quality of the results.
Quick Answer
Monte Carlo simulation assigns probability distributions to uncertain inputs, generates random values from those distributions, and runs the financial model repeatedly. The resulting outcomes form a simulated distribution that analysts can use to estimate expected values, percentiles, loss probabilities, shortfall risk, and the possible range of portfolio or security values.
Key Takeaways About Monte Carlo Simulation
Monte Carlo simulation produces a distribution of possible outcomes rather than one forecast.
Each simulation run uses a randomly generated set of input values.
The model must define how the uncertain inputs produce the final result.
The output can be summarized using expected values, standard deviations, percentiles, and probabilities.
Monte Carlo methods can handle several uncertain variables and complex relationships at once.
More simulation runs reduce random sampling noise within the model.
The reliability of the results still depends on the model, input estimates, and probability distributions.
Simulation supports decision-making under uncertainty without removing that uncertainty.
What You Need to Know for CFA Level I
When reviewing Monte Carlo simulation, you should be able to:
Describe the main steps in a Monte Carlo simulation.
Explain why repeated random sampling produces a distribution of outcomes.
Identify common uses of Monte Carlo simulation in finance.
Interpret simulated outputs such as percentiles and shortfall probabilities.
Distinguish model-based simulation from methods that reuse historical observations.
Identify model risk, parameter risk, and distributional risk.
Explain why additional simulation runs cannot correct unrealistic assumptions.
What Is Monte Carlo Simulation?
Monte Carlo simulation is a numerical technique for analyzing uncertain outcomes. The analyst creates a model, specifies probability distributions for uncertain inputs, and then generates many random combinations of those inputs.
Each combination produces one simulated result. Repeating the process thousands of times creates a distribution that shows how often different outcomes occurred within the model.
Depending on the problem, the analyst may examine:
The mean or median result
Standard deviation
Best- and worst-case percentiles
Probability of a loss
Probability of failing to reach a target
Potential tail outcomes
Sensitivity to changes in assumptions
This approach is especially useful when several variables interact or when a direct mathematical solution would be difficult to calculate.
What Are the Steps of Monte Carlo Simulation?
A Monte Carlo simulation generally follows six steps.
1. Define the Problem and Desired Output
Start by deciding what the model should estimate.
The output may be:
An ending portfolio value
A security price
A project’s net present value
The probability of meeting a funding target
The probability that losses exceed a stated amount
A clearly defined output determines which inputs and relationships the model needs.
2. Build the Financial Model
Specify how the inputs combine to produce the output.
For example, a retirement model may connect starting wealth, investment returns, contributions, withdrawals, inflation, and the investment horizon. A security-pricing model may connect the underlying asset price, volatility, interest rates, and time to maturity.
3. Assign Probability Distributions to Uncertain Inputs
Choose a distribution for each uncertain variable.
Inputs may include:
Asset returns
Interest rates
Inflation
Exchange rates
Default events
Sales growth
Operating costs
Cash flows
The selected distributions should reflect the behavior the analyst expects. Correlations and other dependencies between variables should also be incorporated where relevant.
4. Generate Random Input Values
Draw one value for each uncertain input from its assigned distribution.
Together, these random draws create one possible scenario. A model containing annual returns over several years creates one complete simulated path.
5. Calculate and Store the Outcome
Run the model using the simulated inputs and record the result.
The result may be one ending value, one loss amount, or one measure of whether the target was achieved.
6. Repeat the Process and Analyze the Results
Repeat the random-draw and calculation process many times.
The collection of outcomes forms a simulated distribution. The analyst can then calculate summary statistics and probabilities that support the investment decision.
Increasing the number of runs generally produces a more stable estimate within the chosen model. Stronger results still require reasonable assumptions and an appropriate model structure.
Monte Carlo Simulation Example
Suppose an analyst wants to estimate whether a retirement portfolio will remain above a required value after 20 years.
The model includes:
Current portfolio value
Annual contributions
Planned withdrawals
Expected returns
Return volatility
Correlations among asset classes
Inflation
Investment horizon
For one simulation path, the model draws a return and inflation rate for each year. It then updates the portfolio after investment performance, contributions, withdrawals, and inflation.
The same process is repeated across many possible paths. After 10,000 simulations, the analyst could review:
Median ending portfolio value
5th percentile ending value
95th percentile ending value
Probability that wealth falls below the required target
Probability that the portfolio is depleted
Years in which depletion most often occurs
How the results change under different withdrawal rates
The 10,000 paths are illustrative rather than a fixed CFA requirement. The exam-relevant idea is that repeated random draws create a range and distribution of possible outcomes.
How Is Monte Carlo Simulation Used in Finance?
Monte Carlo simulation in finance can support several types of investment analysis.
Portfolio Planning
Analysts can estimate the probability of reaching a wealth target, funding future liabilities, or maintaining a portfolio through retirement.
The output may reveal how often the portfolio succeeds under the assumptions and which variables have the strongest effect on the result.
Security Valuation
Monte Carlo methods can value securities with complex or path-dependent payoffs.
For example, the value of an option may depend on the sequence of prices followed by the underlying asset rather than only its final price.
Risk Management
Simulation can produce a distribution of potential gains and losses.
Analysts may use that distribution to examine:
Loss percentiles
Tail exposure
Probability of breaching a risk limit
Probability of a portfolio falling below a minimum value
Capital Budgeting
A project’s value may depend on uncertain sales, costs, exchange rates, commodity prices, or financing rates.
Simulating those variables produces a range of possible project values rather than one estimate based on fixed assumptions.
Asset and Liability Management
Financial institutions and pension funds can simulate asset returns and liability growth together.
This allows them to estimate funding shortfalls under different economic and market conditions.
Testing Alternative Strategies
The analyst can rerun the simulation after changing asset allocation, withdrawal rates, hedging strategies, or other assumptions.
Comparing the resulting distributions can show how each decision changes expected outcomes and risk.
How Should You Interpret Monte Carlo Simulation Results?
The simulated output should be interpreted as a distribution conditional on the model’s assumptions.
Suppose a simulation reports:
Median ending value: $1.2 million
10th percentile ending value: $600,000
90th percentile ending value: $2.1 million
Probability of falling below the target: 18%
The median shows the middle simulated outcome. The percentile values describe the spread, while the 18% result estimates how often the portfolio missed the target within the simulation.
A useful interpretation explains the center, range, downside risk, and assumptions together. Reporting only the average can hide a wide range of possible results.
Monte Carlo Simulation vs Historical Simulation
Monte Carlo and historical simulation both generate possible outcomes, but their scenario sources differ.
Area | Monte Carlo Simulation | Historical Simulation |
|---|---|---|
Scenario source | Random values drawn from assumed distributions | Actual historical observations |
Model assumptions | Requires distributions, parameters, and relationships | Relies more directly on observed historical behavior |
Possible outcomes | Can generate scenarios not previously observed | Uses outcomes contained in the historical sample |
Flexibility | Can incorporate forward-looking assumptions and complex relationships | Closely tied to the available historical period |
Main risk | Poor model or input assumptions | Historical data may not represent future conditions |
Historical simulation applies past observations to the current portfolio or model. Its usefulness depends heavily on whether the selected historical period reflects the risks that matter today.
Monte Carlo Simulation vs Bootstrap Resampling
Bootstrap resampling repeatedly draws observations from an existing dataset, usually with replacement.
The distinction is:
Monte Carlo simulation draws values from a specified probability model.
Bootstrap resampling draws values from observed sample data.
Historical simulation reuses actual historical scenarios, often without creating a theoretical distribution.
Bootstrap methods can preserve features of the empirical sample that a simplified theoretical distribution may miss. Their results remain limited by the quality and representativeness of the observed data.
Detailed bootstrap and jackknife methods belong in their own study note.
Advantages of Monte Carlo Simulation
Monte Carlo simulation can:
Handle several uncertain variables at once.
Model nonlinear relationships.
Model path-dependent outcomes.
Incorporate correlations and dependencies.
Produce a full distribution of possible results.
Estimate probabilities that may be difficult to calculate directly.
Test how results respond to changes in assumptions.
Support decisions based on ranges and probabilities rather than one forecast.
These features make the method useful for complex investment problems where uncertainty affects several parts of the model.
Disadvantages of Monte Carlo Simulation
Monte Carlo simulation has several limitations that candidates should recognize.
Model Risk
The model may simplify or omit relationships that materially affect the outcome.
A detailed simulation based on an incomplete model can produce internally consistent but misleading results.
Parameter Risk
Expected returns, volatilities, correlations, and other inputs must often be estimated.
Small errors in these estimates can materially change the simulated distribution.
Distributional Risk
The assumed probability distributions may not reflect actual market behavior.
For example, a model may understate risk when returns have greater skewness, heavier tails, or changing volatility than the chosen distribution allows.
Dependence Risk
Inputs may be modeled as independent even though they become strongly related during stressful markets.
Incorrect correlation or dependence assumptions can understate the probability of joint losses.
False Precision
A model may produce exact-looking percentages and values. That level of numerical detail reflects the simulation process rather than guaranteed forecast accuracy.
Computational Requirements
Large and complex simulations may require significant processing power and time.
This concern is smaller for basic models but becomes more important when the model contains many variables, time periods, or securities.
Sensitivity to Assumptions
Reasonable changes in return, volatility, correlation, or other inputs may produce materially different outcomes.
Analysts should therefore test a range of plausible assumptions rather than relying on one model specification.
Common Exam Traps
Common mistakes include:
Treating a simulated outcome as a guaranteed future result.
Assuming that more simulation runs remove model or parameter risk.
Confusing Monte Carlo simulation with bootstrap or historical simulation.
Ignoring correlations and other dependencies among inputs.
Choosing convenient distributions without checking whether they fit the variables.
Interpreting the average outcome without reviewing dispersion and tail results.
Assuming a complex model is automatically more accurate.
Treating simulated probabilities as facts rather than estimates based on assumptions.
Forgetting that every simulated result is conditional on the model structure.
Practice Question
An analyst uses Monte Carlo simulation to estimate the future value of a portfolio. The model generates 50,000 possible outcomes based on assumed distributions for returns, volatility, and inflation.
Which statement most accurately describes a limitation of the analysis?
The simulation cannot produce a probability distribution of possible portfolio values.
The reliability of the results depends on the model and input assumptions.
The simulation can generate only outcomes already observed in the historical sample.
Correct Answer: B
Monte Carlo simulation produces a distribution of outcomes by repeatedly drawing values from the specified input distributions. The usefulness of that distribution depends on whether the model, parameter estimates, relationships, and distributions reasonably represent the investment problem.
Option A. describes one of the method’s main strengths. Monte Carlo simulation is designed to generate a distribution of possible results.
Option C. applies more closely to historical simulation or empirical resampling. A Monte Carlo model can generate values that did not appear in the historical sample.
Continue Your CFA Level I Prep With KeyPoint
Use structured lessons, practice questions, mock exams, and progress tracking to focus on the time you have left
FAQs About Monte Carlo Simulation
What Is Monte Carlo Simulation?
Monte Carlo simulation is a numerical method that models uncertainty through repeated random sampling. It generates many possible outcomes based on a financial model and assigned probability distributions.
The resulting distribution can be used to estimate expected values, risk, percentiles, and the probability of reaching or missing a target.
What Are the Main Steps of Monte Carlo Simulation?
The main steps are to define the output, build the model, assign probability distributions to uncertain inputs, generate random draws, calculate one outcome, and repeat the process many times.
The completed set of outcomes is then summarized and interpreted as a distribution.
How Is Monte Carlo Simulation Used in Finance?
Monte Carlo simulation is used in portfolio planning, security valuation, risk management, capital budgeting, retirement analysis, and asset-liability management.
It is particularly useful when several uncertain inputs interact or when outcomes depend on the path followed over time.
What Are the Main Disadvantages of Monte Carlo Simulation?
The main disadvantages include model risk, parameter-estimation errors, unrealistic distributions, incorrect dependence assumptions, computational requirements, and false precision.
More simulation runs can make the numerical estimates more stable, but they cannot repair weak assumptions.
What Is the Difference Between Monte Carlo Simulation and Bootstrap Resampling?
Monte Carlo simulation generates random values from a specified probability model. Bootstrap resampling repeatedly selects observations from an existing dataset, usually with replacement.
Monte Carlo results depend more directly on the chosen theoretical model, while bootstrap results depend on the available empirical sample.