Sensitivity Analysis (Including Best Practices and Examples)

By Indeed Editorial Team

Published May 31, 2022

The Indeed Editorial Team comprises a diverse and talented team of writers, researchers and subject matter experts equipped with Indeed's data and insights to deliver useful tips to help guide your career journey.

Financial analysts employ a range of different modelling methods when forecasting the impact changing variables have on an organization. Sensitivity analysis is one of these modelling methods and focuses on the measurement of uncertainty as a result of a changing variable. If you're looking to pursue a career in financial analysis, understanding the uses for this tool and when to apply it can help you advance your career. In this article, we define this type of analysis, compare it to scenario analysis, discuss the best practices, explore its benefits, and provide useful examples.

What is sensitivity analysis?

Sensitivity analysis is a financial modelling tool that provides a means of measuring how the impact of uncertainty in one or more input variables results in uncertainties in the output variables. Some teams refer to this type of analysis as a what-if analysis. The process aims to predict the outcome that may occur after a team takes a specific course of action or performs certain behaviours. For example, a team may wish to evaluate what happens if they change their supplier for a particular manufacturing process. Undertaking this analysis involves changing a model and then observing the subsequent behaviour.

The process is also useful in the analysis of data in scientific research. An organization completes an analysis using different assumptions than those that they used in the primary analysis. This enables the development of a set of comparative results that provide an excellent test as to the robustness of a set of results.

It's also a tool that researchers use when there have been violations of good clinical practice or protocols, or when there is missing or ambiguous data. Medical and pharmaceutical organizations use sensitivity analyses extensively in their research.

Sensitivity analysis vs. scenario analysis

Although there are similarities between these two approaches, there are some significant differences. The former looks at the impact a set of independent variables has on dependent variables, with those doing the evaluation looking at one variable across a broad range of values.

Scenario analysis examines a specific scenario in depth. The specific scenario is usually a significant event that causes a major economic impact, such as a global recession that causes a change in the markets. Those completing the evaluation focus on all variables for a limited range of values. Financial analysts use the sensitivity tool when considering the impact of one input, such as an increase in the price of a resource, while the same analysts use scenario analysis for financial investments, such as the purchase of capital.

Related: What Is Quantitative Analysis?

Sensitivity analysis best practices

The following best practices are worth considering when using this tool:

Consider the layout

This type of analysis can be complex and, unless it's clear and well formatted, it can be difficult to interpret. To overcome these complexities, use a spreadsheet style presentation, as it assists with clarity and readability. There are several key considerations when formatting the analysis, including:

  • Group all assumptions together in one area of the model.

  • Use a unique font colour to separate the assumptions and make them standout.

  • Give the assumptions you plan to test some thought, only test the key assumptions.

  • Create a separate area for the analysis that utilizes grouping to make it easier to read.

  • Make sure you understand the relationship between the independent and dependent variables.

Make use of visual aids

Spreadsheets can be difficult for those who aren't overly familiar with that style of presentation. To assist with visualization, consider presenting some of the data in charts and graphs. Try not to overdo it. Instead, focus on using visual aids to present key data. Not only does this assist with understanding, it also helps emphasize key points from the analysis.

Tornado charts are ideal, as they can illustrate the changes to multiple variables on the one chart. Instead of listing the data horizontally, list it vertically and detail the results from the most impactful to the least impactful. This typically results in the chart taking on the appearance of a butterfly.

Prepare for results to change

There's always some uncertainty and possibility for change when predicting future outcomes. When the available data changes, it's possible that the results may also change. These results can give you useful insights into future possibilities, but there's always some potential for error. The best way to ensure the results are as accurate as possible is to use high-quality data for the model and ensure that you analyze a range of different outcomes, as this helps account for variances.

Direct vs. indirect methods

When considering sensitivity modelling, give some consideration to whether you wish to apply a direct or indirect method of analysis. A direct method substitutes different numbers into the analysis, while an indirect method changes the percentage value that the formulas in the analysis use rather than changing the value of the assumption.

For example, if the assumption for sales growth is 10% each year, the formula is (revenue previous year) (1 × 1.10). Using a direct model, different percentage changes might be 5%, 15%, or 20%. Replace these figures in the analysis to give a total sales value each year. Using an indirect method, the 10% growth each year remains the same. Instead, the value of the assumption, the formula becomes (revenue previous year) 1 × (1.10 + x).

Related: Understanding How to Complete a Risk Analysis

Benefits of conducting sensitivity analyses

There are several benefits that come from applying this type of analysis in a business setting, including:

  • Informs decisions: Sensitivity gives the decision makers in an organization clear data regarding the consequences of changing certain variables. This means they know the likely consequences of their decisions so can base these on data.

  • Adds credibility: Testing a model against a range of distinct possibilities adds credibility to any results and the decisions these results yield.

  • Allows forecasting: The benefit of using historical data to forecast future outcomes results in more accurate forecasting, to the extent the organization can rely on the results.

  • Allows better resource allocation: The detailed nature of this type of analysis means that organizations can test different combinations of resources. This means that it's easier to find the optimal scenario when allocating resources.

  • Enables fact-checking: This type of modelling allows an organization to fact-check and determine the likelihood of a certain outcome.

  • Leads to improvements: The nature of sensitivity modelling and the in-depth analysis it entails allows decision makers to focus on the areas that they can make the greatest gains in over the coming years.

    Related: What Is Forecasting? (With Definition and Different Methods)

Examples of sensitivity analysis

Here are some practical examples of how you can use sensitivity analyses in practical scenarios:

Example 1

Jamie owns a sports store in Calgary, Alberta. When the local hockey team makes the playoffs, Jamie expects the company's revenue to increase through the sale of team merchandise. He wants to find out the likely increase in revenue in relationship to how far the Calgary team advances in the playoffs. A team jersey retails for $150. Last year, the team advanced to the third round of the playoffs and the store sold 200 jerseys, resulting in revenue of $30,000.

After conducting sensitivity modelling, Jamie discovers that for every round of the playoffs that the local team makes it through, there's a 10% increase in the sale of team jerseys. This means that Jamie can forecast a 40% increase in revenue from Calgary team jerseys if they win the championship. With this information, Jamie can manage the in-store inventory and determine their return on investment.

Example 2

Dawson City, Yukon, is conducting an environmental analysis this year to help determine areas to open up to gold mining. Some input variables it considers include its natural capital, such as lakes, local parks, and forests. It also considers stakeholders, such as citizens, gold miners, and tourists, and the amount of interest stakeholders have, such as the areas that the mining is likely to impact.

After conducting its analysis, Dawson City creates a map outlining the possible mining of various areas based on the degree of natural capital and stakeholder interest in preserving it. This allows it to make informed decisions when determining what to do with a particular natural capital. It also helps Dawson City determine what areas it can open up to gold mining.

Explore more articles