What is Sensitivity Analysis and How Can It Help Your Organization?
What if you had the power to see every possible outcome of your business decisions and confidently make decisions about your organization’s financial future?
Building an agile organization capable of withstanding and adjusting to ever-changing market conditions means being able to identify, with a high degree of certainty, how your organization could be impacted by unexpected events.
This is where sensitivity analysis comes into play. It helps organizations identify critical internal and external drivers that impact their choices, as well as how sensitive their financial models are to each of those variables.
Let’s go over what sensitivity analysis is used for, how it’s conducted, and how you can use it to steer your organization towards better decision-making.
What is sensitivity analysis?
All financial models are built with certain assumptions. These assumptions often take the form of input and output variables (also called independent and dependent variables). An analyst performing sensitivity analysis examines different combinations of these variables, their interrelationships, and how they impact business decisions and outcomes.
Also known as “what-if” analyses and “stress tests,” sensitivity analysis is often performed as a type of risk analysis and is very important in risk management and contingency planning. In a stress test, the sensitivity of a particular financial model to a set of independent variables is determined by running the model through extreme events (such as COVID-19).
When push comes to shove, can the model survive all possible worst-case scenarios?
The main goal of sensitivity analysis is to help organizations uncover potentially overlooked financial vulnerabilities. Sensitivity analysis is especially useful for complex “black box” scenarios that are very difficult to analyze using conventional methods.
Sensitivity analysis is also a reliable way to uncover the hidden levers that have the greatest impact on business decisions. Analysts adjust independent variables using one-at-a-time (OAT) analysis to uncover how each independent variable impacts the dependent variables.
What is Model Sensitivity?
The degree to which a dependent variable is affected by a change in an independent variable is called its sensitivity. The degree to which a financial model is susceptible to changes in independent variables is called model sensitivity.
Here are some examples of how a financial model may be sensitive to changes in specific inputs:
- What happens to operating profits if you raise the price of your best-selling product by $1.00 per unit?
- How will your bottom line be impacted if the construction of a new facility takes 3 months longer than expected?
- What happens to cash flow if your customer foot traffic drops by 25 percent?
In unprecedented or difficult times, answering critical operational questions like these as accurately as possible makes it easier for decision-makers to confidently reach a consensus on the best decisions for the organization.
Sensitivity Analysis vs. Scenario Analysis
It’s easy to mistake sensitivity analysis for scenario analysis and vice-versa. After all, both are used by analysts in risk management and contingency planning and share several similarities.
The major difference between the two types of analysis is the outcome of each analysis: scenario analysis reveals which scenarios are most optimal or most detrimental, while sensitivity analysis reveals how sensitive different scenarios are to changes in specific input variables.
For example, let’s say that a university wants to find out whether it makes more sense to build a new science center or dormitory during a partial capacity semester (due to COVID-19). Let’s assume that the cost of the science center is twice the cost of the new dormitory.
How will the university’s endowment determine which project makes more fiscal sense?
Using sensitivity analysis, analysts might determine that both projects, which require financing, are very sensitive to a longer timeline for a vaccine because of the lost room-and-board costs. But the new dormitory construction is much less disruptive to in-person and virtual science classes because no students or teachers will have to be relocated.
Using sensitivity analysis alone, the university might decide that building the new dormitory first may make more sense. But using scenario analysis, the university might review many other scenarios, including one in which neither project commences. They may actually decide that the best course of action, while COVID-19 continues unabated, is to simply do nothing at all.
Sensitivity and Risk Analysis
While no one can accurately predict the future, that doesn’t mean you shouldn’t try. Predicting points of failure and vulnerabilities ahead of time helps your organization stay solvent with plenty of cash equivalents on hand.
Unsurprisingly, sensitivity analysis and stress tests, in general, are very commonly used on Wall Street by financial organizations, such as banks, funds, and portfolio managers.
The most well-known example of sensitivity analysis is the Comprehensive Capital Analysis and Review (CCAR), conducted annually by the Federal Reserve Bank. The CCAR is used to assess whether the largest bank holding companies in the U.S. can weather worst-case market scenarios, such as sudden market crashes, high default rates for mortgages, and political upheavals.
That’s not to say that sensitivity analysis is only useful for risk management. Far from it. In fact, once sensitivity analysis is properly incorporated into your organization’s risk management strategy, not only can you dodge bullets, but you can also be poised and ready to take advantage of opportunities in times of crisis.
As an example, in the most recent COVID-19-driven market crash, some hedge fund managers (like Bill Ackman) made fortunes as the markets were nosediving by diligently conducting scenario and sensitivity analysis. As Warren Buffett is so fond of saying, “Be fearful when others are greedy, and greedy when others are fearful.”
How to do Sensitivity Analysis
Every sensitivity analysis can be simplified into three major steps:
- Establish a base case
In sensitivity analysis and scenario planning, the three most common scenarios are called:
- The best case, or the most optimistic scenario with the highest potential upside
- The worst case, or the most pessimistic scenario with the highest potential downside
- The base case, or the most conservative scenario with an outcome squarely between the best-case and worst-case scenarios (e.g., the outcome in which nothing amazing or disastrous occurs).
Once a realistic base case scenario is identified, analysts must decide which independent and dependent variables are most relevant to the outcome.
- Determine your variable inputs
Input variables may include things like cost of goods sold, debt financing, employee wages, customer foot traffic, etc. Output variables may include cash flows, internal rate of return (IRR), net present value (NPV), net profits, etc.
For example, net present value, which accounts for the time value of money, is often used to determine whether projects will be profitable. NPV takes into consideration initial capital, the acceptable rate of return, and the return on investment from cash flows.
- Test the variables
Once the inputs and outputs are determined, analysts perform sensitivity analysis on the assumed independent variables one by one to rigorously test how sensitive their base case is to even the smallest changes.
It’s important to use the base case as a frame of reference for the OAT analysis because it serves as a control. Without a realistic base case scenario, there is no way to reliably determine how the best-case and worst-case scenarios might be impacted.
After all, in the real world, multiple-input variables are likely to change all at once or one after the other, often in extreme and unpredictable ways.
You can test your input and output variables by following this simple four-step process:
- Determine which independent variable you will change first and which dependent variables you will observe.
- Record the output of your base case scenario after changing the first independent variable. Record the percentage change of both the input and output variables.
- Repeat steps 1 and 2 until you have tested all of your independent variables one at a time and recorded how they impact the dependent variables.
- Using the appropriate sensitivity analysis formula, calculate how sensitive your base case scenario is to changes in each of your independent variables.
Key Sensitivity Analysis Formulas
Here’s the main formula that analysts use to calculate the sensitivity of output variables to changes in input variables:
Sensitivity = Percentage change in output / Percentage change in input
If it’s possible to determine the exact mathematical relationship between each independent variable and dependent variable, outputs can also be written as a function:
f(x) = y
“x” is the independent variable (input), and “y” is the dependent variable (output)
The exact function formula will vary depending on the specific relationship between variables, which will be unique to the scenario you’re analyzing.
In order to determine whether a project or investment is worthwhile, analysts may look at net present value (NPV) as the output (dependent variable) in their sensitivity analysis. NPV helps you see whether a project is profitable using the following formula:
NPV = ( Cash flows / (1 + discount rate)t ) – initial investment
“t” is an incremental unit of time (such as # of years)
If the result of the NPV calculation is positive, the investment will yield the desired returns. If it’s negative, it won’t.
Real-world sensitivity analysis example
Let’s look at a real-world example of how sensitivity analysis might help a retailer decide where to focus their efforts in 2021. In 2019, they sold 225,000 units of their flagship product, Product A, at $49 per unit, which resulted in $11,025,000 in revenue. Some of their sales came from website visitors, while others came from in-store purchases.
The company’s financial analyst was tasked with determining the sensitivity of sales to website traffic vs. foot traffic. The analyst organized previous sales history in the following manner:
Website Traffic (+/-)
Store Traffic (+/-)
YoY Revenue (+/-)
At a glance, the analyst realizes that store traffic had actually decreased in 2018 before recovering slightly in 2019. But it was still below 2017 levels. Meanwhile, website traffic increased significantly in both years. He reasonably concluded that the YoY sales and revenue increases were both solidly due to growth in website traffic.
Then he looked at the numbers for 2020:
|Year||Website Sales||Store Sales||Website Traffic (+/-)||Store Traffic (+/-)||Revenue||YoY Revenue (+/-)|
Due to the COVID-19 pandemic shutdowns, store traffic plummeted to the lowest levels in the retailer’s history. Website traffic also fell, but not by nearly as much. Overall, revenue fell by 46.2% compared to 2019.
Clearly, website sales were less sensitive to the shutdowns than store sales, which makes perfect sense. More importantly, website overhead and associated expenses were significantly lower across the board:
|Year||Website Expenses||Store Expenses||Website Costs (+/-)||Store Costs (+/-)||Revenue||YoY Revenue (+/-)|
Without looking at the cost of certain fixed overhead expenses, which remained the same across both websites and stores (such as employees and vendors), the analyst was able to determine that although website costs rose faster relative to store costs, website expenses were still much lower as a percentage of website sales and revenue.
The analyst was also able to determine that despite ramping up advertising costs in 2020, website sales were still down, just like store sales. He reasonably determined that there was another variable the retailer’s customers had been particularly sensitive to in 2020: price.
The analyst’s final recommendation to the retailer? Continue lowering store costs (or close stores entirely), continue to increase online advertising budgets, and drop the price of Product A by 20% for six months, then by 10% for the next six months.
Will his recommendations work in 2021? Only time will tell. But the analyst can confidently state his case—and the retailer can more confidently make a strategic decision—after reviewing the results of this sensitivity analysis.
A Smarter Way to Perform Sensitivity Analysis
The most popular tool by far for conducting sensitivity analysis and building financial models remains Excel. Unfortunately, spreadsheets leave a lot to be desired. They require a lot of repetitive, manual entry, leaving little room for error.
There’s also no easy way to create a multidimensional analysis in a two-dimensional spreadsheet. Analysts often have to go back to the drawing board and build new spreadsheets from scratch every time stakeholders come up with a new set of questions that need to be answered.
This is why sophisticated organizations should consider using financial modeling tools built specifically for sensitivity analysis. Synario, for example, features patented Multiverse Modeling™ software, which allows users to run unlimited scenarios in a single model.
Not only does this help analysts uncover key drivers faster, but it also lets them stress test their models across countless scenarios in a fraction of the time it would take in spreadsheets.
Synario’s intuitive interface and vast library of prebuilt algorithms allow modelers to conduct advanced what-if analyses at the macro- and micro-scenario level without the need for hours of tedious and repetitive manual entry.
When it’s time for analysts to meet with decision-makers and present their findings in the board room, Synario’s Presentation Mode makes it easy to visualize the impacts of any executive decision—and to make changes that are automatically reflected mid-presentation.
Join hundreds of public and private organizations around the world and invest in your company’s financial future with Synario’s patented Multiverse Modeling™ solution.