Have there been times when you step out of your house in the morning with your whole day planned down to every minute? However, along the line, one of five things go wrong, and the entire day goes down the drain.

The field of financial modeling can be a lot like this. There are various possibilities, and a good financial model is the one whose sensitivity you can stress-test against all these.

Sensitivity analysis can help give you appropriate insight into the problems related to any particular financial model.

**What is Sensitivity Analysis?**

Sensitivity analysis is the method used to find out how independent variable values will affect a particular dependent variable under a particular set of assumptions.

It is a technique that determines how the unpredictability in the outcomes of a model or system can be as a result of the different sources of unpredictability in its inputs.

Sensitivity analysis is used within specific boundaries, which is dependent on one or more input variables. Also referred to as the what-if analysis, it can be used for any system or activity.

From making decisions at corporate levels to planning a vacation with some variables in mind, you can do all these through sensitivity analysis.

Simply put, sensitivity analysis is a way by which you can foresee the outcome of a decision provided in the form of a specific range of variables. By creating a set of variables, the analyst can point out how changes in a variable affect the outcome.

**Types of Sensitivity Analysis**

Primarily, there are two types of sensitivity analysis, which are

- Local Sensitivity Analysis
- Global Sensitivity Analysis

Local sensitivity analysis is based on derivatives (numerical or analytical). The word “local” signifies that the derivatives are taken at a single point. This type of sensitivity analysis is great for simple cost functions but not practical for complex models.

Local sensitivity analysis is a one-at-a-time (OAT) method that assesses the effect of one parameter on the cost function at a time, holding the other parameters fixed.

On the other hand, global sensitivity analysis uses a global set of samples to analyze the design space. It is usually carried out using Monte Carlo techniques. Some of the more widely applied techniques include:

- Differential sensitivity analysis
- Factorial Analysis
- One at a time sensitivity measures

**Methods of Calculating Sensitivity Analysis**

Below is a step by step method of calculating sensitivity analysis:

- Firstly the base case output is established; say the NPV at a certain base case input value (V1) for which the sensitivity is to be calculated. Keep all the other inputs of the model constant.
- Then calculate the output’s value at another value of the input (V2) while keeping other inputs constant.
- Determine the change in the percentage in the output and the percentage change in the input.
- Divide the percentage change in output by the percentage change in input. This is how to calculate the sensitivity.

Keep repeating the process of testing sensitivity for another input while keeping the rest of the inputs constant until you obtain the sensitivity figure for each of the inputs.

The determination would be that the greater the sensitivity figure, the more sensitive the output is to any change in that input and vice versa.

**Uses of Sensitivity Analysis**

Sensitivity analysis is important for various reasons. Some of its uses include:

- Improves the understanding of the correlation between output and input variables in a system or model.
- Evaluates the strength of the output of a model or system in the presence of uncertainty.
- Look for the errors in the system or model by determining the unexpected relation of the inputs with the outputs.
- Lowers uncertainty by pointing out model inputs that generate uncertainty in the output.
- Boosts communication between decision-makers and modelers. It happens by making suggestions that are more credible, understandable, persuasive, or compelling.
- Enables model simplification by fixing the model inputs, which do not affect the outputs. It may also occur by ascertaining and removing unnecessary parts of the model structure.
- Attempts to identify vital connections between different observations, forecasts, or predictions and model inputs, which brings about the development of better models.
- Alleviates the calibration stage by bringing out the sensitive parameters. Sensitivity parameters should be known as without that, the result can be a total wastage of time being spent on the non-sensitive sections.

**Conclusion**

Sensitivity analysis is a useful tool that assists decision-makers with more than just a solution to a problem. It gives a reasonable insight into the problems related to the model under consideration.

It also provides the decision-maker with a decent idea of how sensitive the ideal solution chosen by him is to any changes in the input values of one or more variables.