Forecasting Comes of Age

Corporate finance departments have come to rely on forecasting in previously unimagined ways.


When Arby’s first opened its doors in 1964 in Boardman, Ohio, it had only roast-beef sandwiches and “Texas-size” iced tea to keep track of. But more than 3,000 stores and countless turkey and chicken combo meals later, the fast-food chain relies on its forecasting to keep one step ahead of the many forces that can affect food and labor costs.

David Pipes, CFO of Arby’s Restaurant Group, uses forecasting to stay current. “We need to know what our expectations are for the year at all times,” he says. “It can’t just be, ‘This is what we planned at the beginning of the year, so let’s assume that is right.’”

He is not alone. Corporations have come to rely on forecasting in ways that were never imagined previously. In turn, the kinds of forecasts businesses now employ have multiplied, including cash-flow forecasts, dynamic forecasts, financial forecasts, production forecasts, revenue forecasts, rolling forecasts, sales forecasts, soft forecasts, static forecasts, and all manner of other means of data-driven crystal-ball gazing.

But companies today don’t stop there. For some CFOs and other senior finance executives, forecasting is no longer just a supporting analytical tool. Instead, the process has become so central to companies’ overall growth plans that the way it’s done has become an essential part of the business itself. Companies have come to rely on their forecasts as working documents throughout the year, while updating them on a daily basis. No longer do multiyear plans merely molder in desk drawers, gathering dust.

Eileen Kamerick, CFO of Press Ganey, a health-care performance-improvement firm, says “the forecast, in many ways, is more valuable than the budget,” especially because forecasts are oriented to action. “What the forecast helps to do is get a sense of what the risks and opportunities are looming in front of you and what [you can] do about them.”

In line with such strategic uses, Kamerick says, the most appropriate question to ask before constructing a predictive program is, “What decisions are you going to make from this forecast?”

The answers, of course, vary. A highly leveraged company may need to figure out, from its cash-flow forecasts, how much cash it has on hand and would have on hand under certain borrowing scenarios. In contrast, a company that sells annuities may need to make forecasts of possible fee and investment revenue in order to price a product it is planning to offer.

How Arby’s Does It

In the case of Arby’s, Pipes says the company’s forecast considers future events like a change in the marketing calendar or in the timing of expenditures that will affect its financial results. How does this process begin? The finance chief starts with a detailed annual plan that has been built up restaurant by restaurant after he has received input from the company’s operations, marketing, tax, employee benefits, and other units. The plan is then adjusted each period, based on current expectations.

The company will drill down in its forecasts only so far, however. “We don’t forecast an exact product mix by restaurant. While it might add more precision to the process, I am far from convinced it would add more accuracy,” says Pipes. Arby’s current approach, he adds, identifies “what’s going on in sales, and not just in total sales for the company, but also by market and region.”

Similarly, Paul Schuster, treasurer and vice president of corporate finance at Trustmark Insurance, a health, life, and employee benefits management services firm, uses forecasts to continuously understand and manage the key drivers of his business. “We are forever looking at regulatory issues, product-design issues, competitive pressures, and so on,” he says. “We are continually looking ahead and only periodically peeking at the past.”

Managing the Future

Press Ganey’s Kamerick would agree with that thinking. Some corporations, she says, can spend too much time looking backward at their data for answers about their future growth. In contrast, a good forecasting process helps CFOs and other senior managers to think about offsetting any negatives a company may encounter and how to “further exploit the positives.”

Trustmark lives up to that philosophy. In its monthly meetings with the businesses that make up the company, including HealthFitness, Trustmark Voluntary Benefit Solutions, and the Starmark health-benefits group that targets small business, “90% of the meeting is talking about the future,” says Schuster. “We believe you manage and influence your future, not your past.”

To manage its future in that way, Trustmark employs rolling forecasts, which entail adding and dropping months to keep a constant 12-month projection. The Lake Forest, Illinois-based company continually recasts the remainder of the current year and the entire next year. Every month, it forecasts full profit-and-loss statements and then combines the data using Ontario-based Longview Solutions’s consolidated software platform.

Trustmark focuses on obtaining 50/50 forecasts. In such predictions, a sales plan, for example, would have a 50% chance of achieving its targets and a 50% chance of failing. But when the next year’s plan is getting close to finalization in the fourth quarter, the key drivers will then be adjusted to ones that will have a 70% chance of achieving their targets and only a 30% chance of missing them. Using the two assumptions works well for the company, notes Schuster.

Beyond Excel

Gerard Chiasson, president of corporate performance management at Longview Solutions, says CFOs today want forecasts to help in their current decision-making. “What they really want is a way to analyze their metrics,” he says. Years ago, simple expense forecasting was done in Excel, a process “[that] was restricted by what you could and could not do by the limitations in Excel.”

Indeed, before Trustmark applied Longview’s software to the forecasting process, Schuster says using Excel to consolidate financial data from its various business units was “just dreadful.” It involved heavy cutting and pasting, and using various links between the numerous spreadsheets. Since there were so many links used, some would invariably break, and the entire process would fail.

But CFOs and their staffs are by no means giving up on Excel; they just plan to apply it differently when it comes to forecasting. In Schuster’s case, his Excel workbooks are linked to a single database that has actual results alongside the forecasting estimates.

Similarly, Michael Giglio, CFO and director of acquisitions at Kaufman Organization, a New York–based real estate owner and leasing company, uses Excel in tandem with more-advanced forecasting tools supplied by Argus Software, a real estate–focused forecasting-solutions firm. While Kaufman has raw data on the properties it owns, the Argus program adds data on many other buildings the company doesn’t own. Thus, by using the software, Giglio can derive intricate financial forecasts not only from Kaufman’s own properties, but also from a broad range of properties the firm might be interested in buying.

The software can be used to develop forecasts with a great deal of specificity, including such data as the amount of a tenant’s income, lease clauses as they pertain to rent, real estate taxes, wage raises, and cost-of-living increases as they relate to a particular property’s tenants. The data on both the Kaufman properties and the global market are then output to the finance department’s Excel spreadsheets. The process enables Giglio to determine realistically achievable rents for his properties, to better understand the increases in a property’s value, and to set market leasing assumptions, according to the finance chief.

More CFOs and other corporate executives are coming to rely on such specific detail in their forecasting. Models have “become more complex because of the users’ and the authors’ need to run them as close to reality as you can get,” says Giglio. Using forecasting tools is “really telling yourself the truth,” he adds, rather than just guessing what a property is worth — particularly when the company considers an acquisition.

Overforecasting the Forecast

Determining which details are relevant in forecasting is still a challenge. A forecast always tries to include as many variables as possible; it wouldn’t be much of a forecast if it didn’t. But some corporate executives take that too far and put extraneous variables into the mix. Such overage can clog the forecasting processing, leading to mistakes and misuse of a CFO’s time.

“You want to be forecasting on things that make a difference,” says Kamerick. If CFOs have too many variables, they are going to drown out or dilute the variables that really matter, she explains. “You should have the discipline to say, ‘What really moves this business?’”

Inclement weather, for instance, affects everyone to some degree, though it doesn’t necessarily belong in a general law firm’s financial forecast, according to Kamerick. Of course, if a law firm caters to insurance-related customers only, then weather could have a bigger effect, she notes.

Still, by including a plethora of factors in its forecasts, a company may overlook obviously critical perils. “What you have are companies that are focused on such a wide variety of sophisticated financial measures when, frankly, that [critical] risk is right in front of [their] face,” says Kamerick, citing the example of a mortgage company not accounting for the possibility of a downward turn in the housing market in a forecast.

“Forecasting numerous different scenarios and assumptions can bury you in numbers,” says Trustmark’s Schuster. He advises that you start by focusing only on those that have a 50/50 chance of happening. “What if we have an asteroid hit? That’s overkill. To the extent that something does surface that we didn’t see or anticipate, then you do have to quickly and effectively respond to it,” he says.

Being nimble enough to adapt one’s forecast when severe changes do occur is one more hurdle corporate executives face; another, of course, is if the forecast is wrong. Every corporation, says Kamerick, should have an “escape hatch,” or contingency plan, for their forecasts. “It’s great to have a forecast,” she says, “but don’t lull yourself into thinking that just because you spent a lot of time on it and put a lot of thought into it, that it’s necessarily going to be right.”


comments powered byDisqus

Read next:

Becoming a Value Integrator: A CFO Journey