Financial stress testing: A COVID-19 survival technique for business

The massive economic shock from COVID-19 has made planning for the next normal increasingly complex. Financial stress testing can bring much-needed clarity.

The onset of the COVID-19 crisis immediately rendered obsolete every organization’s carefully constructed revenue projections and budget planning. That has left executives scrambling to come to grips with not just the pandemic’s immediate impact but also its implications for their viability.


Financial stress testing, or scenario-based forecasting, is a corporate planning tool that helps executives manage under pervasive uncertainty. This method starts by first identifying and cataloging relevant macroeconomic and microeconomic risks and then correlating major factors that affect a company’s balance sheet income statement with external drivers.

Arvind Govindarajan, a McKinsey partner focused on financial institutions, risk, and advanced analytics, is well acquainted with how to apply financial stress testing to home in on the most relevant metrics and indicators for companies. In this podcast, Govindarajan walks through applications of financial stress testing and highlights several key elements of the current crisis, including:

  • How financial stress testing can inform and support business strategy
  • Why previous assumptions and traditional forecasting methods are insufficient to cope with the crisis
  • How the pandemic has severed the link between supply and demand

For more details, please listen to the podcast or read the full transcript below.

Podcast transcript

Amanda Schmitt: Hello and welcome. You are listening to a podcast by McKinsey & Company focused on COVID-19 response. I’m Amanda Schmitt, global risk learning manager at McKinsey & Company. Today I’m joined by Arv Govindarajan, who is a partner in McKinsey’s Boston office and a leader of McKinsey’s risk work in digital and analytics. He’ll be walking us through the process of financial stress testing and how using this approach can help a company survive this crisis and thrive in the future.

Arv, thank you so much for joining us. To start us off, can you explain what exactly stress testing is and what it’s used for?

Arvind Govindarajan: Financial stress testing, otherwise known as scenario-based forecasting, is a way for companies to plan under uncertainty. Essentially, the world has a number of risks right now—some macroeconomic, some not macroeconomic. And companies are asking us, “Well, how do I know what’s going to happen to me under these different circumstances?” Stress testing allows them to basically model out their P&L under a wide range of different economic and noneconomic scenarios.

The outputs of this are things like the balance sheet as well as the income statement, but also the eventual impacts on their labor force, operational impacts on the amount they can pay out in wages, et cetera. It really is a corporate planning tool that allows them to manage under a wide range of uncertainties, such as the crisis we are facing today.

Amanda Schmitt: OK, Arv. How should clients come up with different scenarios to test?

Arvind Govindarajan: The first step in any of this is coming up with the various risks that the company faces. Some of those will tend to be global macroeconomic ones—things like GDP and employment. Some of them are going to be very industry specific.

So, for example, if I’m thinking about agriculture, I’m very concerned about oil costs because that, in turn, impacts my cost of fertilizer. If I am a bank, I’m very worried about unemployment for a retail bank. And I’m very worried about trade flows, for example, to payments. So the key question “What are the major drivers of my business?” is the first step.

Second step is correlate those major drivers that impact your balance sheet and income statement with external drivers: for example, macroeconomic ones, for example, government regulation, for example, labor-force size, et cetera. For these, we tend to use statistical techniques in order to get to the link between the two. In terms of “How do you develop scenarios?,” most of the time, people look at the demand side of the equation—they design scenarios where, for example, the demand for the products are down, et cetera.

The key is to understand each individual step separately. COVID-19 is a good example. It’s not just demand that’s going to be cut out. It’s also supply—because of the supply chain impact, because people are not allowed to travel. So even if demand is there, they actually cannot access the demand because of supply-side issues. So key elements of it, again, are to think about what does drive your business and then, based on that, see what are the various ways to stress your business using those specific drivers.

Amanda Schmitt: How can clients use this approach now in the midst of this crisis? And how is this one different from other crises that you’ve seen?

Arvind Govindarajan: Clients should definitely be using this approach now, even though they’re in the crisis as we speak. A number of CSTs [client-service teams] have already asked us the question, “How can I tell my client, help my client, figure out how this crisis is going to impact them?” Why is that? The reason is because this crisis is unlike any of the other ones we’ve seen before, which means pure experience won’t help. Unless they were alive in 1918, there’s never been a pandemic that has been global before.

What is the difference with this pandemic? Three things: the velocity of change, second one is the uncertainty of when it will end, and the third one is that it does not only impact demand like most recessions do, it also had a number of other hits.

The velocity. Now, I’m in the US. And in the last week of March, the number of jobless claims on a weekly basis jumped five times larger than ever seen before since the Second World War. That is a month of jobless claims that we have never seen before. The impact of that is not even clear, as a result of which just relying on experience does not work.

Second is we’re seeing something unique—lockdowns on a global level—because the uncertainty on when the virus will actually be controlled is not fully clear. What else is not clear is how the economy will rebound back to normal. There is huge uncertainty in how it is that the economy will recover once we end the lockouts, and so on. There are possible ranges, but you can’t get to the ranges from experience again. You have to get the ranges by modeling different options.

The third one is that the areas of the economy that are impacted are unlike the ones before. This is not pure demand. Even if people had jobs and money, they actually cannot spend the money the way they want. You can’t go to the car shop and buy a car because you’re in lockdown. You can’t go to the restaurants. You can’t fly in many places. Even if you wanted to, you couldn’t do it, which means that it has impacts beyond the classic demand.

Some industries are even more impacted—essentially, the airline industry, the hospitality industry, the hotel industry is completely down for the count. Anybody with a supply chain that’s impacted that goes to the countries with high lockdowns has a problem.

There is a spat between Saudi Arabia and Russia, which is pushing oil prices to levels not seen since the ’90s. These are not just pure demand problems. They’re more than that. And so figuring out how this is going to work requires more than just pure experience.

As a result of this, even if companies are prepared, they should use the preparation but also, at the same time, reexamine which assumptions in their preparation are no longer valid. In clients we have talked to—the demand side, like I said, is fine again—but did not really plan for a longer period of shutdown: workers to be working from home, a number of folks with relatives who are sick. Even the best-laid business-continuity plans tend to last a few days, not eight weeks to 12 weeks.