Many companies are sidestepping the consequences of earnings misses by holding back on guidance and, instead, focusing on improving performance management for the long haul.
Coca-Cola made a splash in December when its CEO informed investors that the company would no longer provide any quarterly or annual earnings-per-share guidance. Critics dismissed the move as foolish, but "daft like a fox" may be a more accurate characterization. Coke's new stance doesn't necessarily indicate that the company lacks visibility into its short-term performance. Rather, by focusing on long-term expectations, Coke can duck some of the blows Wall Street dishes out to organizations that miss short-term earnings forecasts by even small margins.
Forecasting is certainly more art than science, most corporate financial analysts agree, and it's a difficult art at that. The markets have recently been walloping companies for anything less than perfect vision. Yet focusing on the reluctance of Coca-Cola and other public companies to share earnings expectations masks the fact that businesses are using the downturn to significantly improve their forecasting and business performance management (BPM) capabilities.
Only 34 percent of financial services companies provided specific earnings guidance during the past 12 months, according to a late-2002 Ernst & Young analysis of that industry. But the companies that withhold forecasts may have sharper insight than they're letting on, says David Axson, senior vice president of The Hackett Group, an Answerthink company based in Hudson, Ohio. "With the stock market's short-term volatility, companies are getting punished -- even though they may be performing well -- simply because they missed their estimate," Axson claims.
The current economic and geopolitical environment has exacerbated long-standing problems with generating accurate top-line forecasts. When revenues and profits soared in the late '90s through early 2000, missing projected earnings per share by one cent was less of an issue. "Two years ago, the world changed," says Dave Catrambone, director of corporate financial planning and analysis for Network Appliance Inc., a network storage company in Sunnyvale, Calif.
"I think the challenges of forecasting have also changed," Catrambone adds. Companies now must address the potential impacts of unexpected events that could have major economic consequences, such as a 20 percent reduction in imported oil, major companies dropping off the face of the earth and gruesome terrorist attacks on civilians.
"There are a lot of external variables at play in today's economy, adding to the unpredictability of the sales cycle," says Chris Scherpenseel, president of FRx Software, a Microsoft Business Solutions company based in Denver. "The economy has driven companies to focus on demand creation through performance and productivity." That makes revenue -- which is always more challenging to predict than costs -- even more difficult to forecast. And, Scherpenseel notes, "incomplete or unsupported information makes organizations, employees and investors nervous."
Most CFOs and their finance managers were jolted by the April 2001 announcement that Cisco Systems would write down $2.5 billion worth of parts. "That triggered CEOs to wander down the corridor; bang on the CFO's door; and say, 'How the heck do we get a better forecast?' " Axson notes.
The lifeboats most companies turned to in the downturn (which John Chambers, Cisco's president and CEO, has referred to as a "100-year flood") consisted of the variety of processes and technologies that the term BPM now comprises. Two years ago, software producers began trumpeting performance management systems' potential to provide X-ray visibility into supply and demand chains. BPM applications haven't quite lived up to the hype, but many public companies have dedicated substantial amounts of time and resources to the goal of sharpening their forecasting -- and budgeting, analysis and reporting -- capabilities.
On the process side, corporate financial analysts today spend more time identifying and tracking trends and evaluating the likelihood and impact of external events -- a larger-than-expected rate hike by the Federal Reserve, for example, or protracted instability in the Middle East. Alan Yong, director of financial analytics for Aberdeen Group in Boston, points out that companies with leading forecasting and planning practices frequently run through scenarios. "They're taking a page from the risk manager's playbook," Yong says. "They're using predictive forecasting to assign probabilities to their forecasts." So a business-unit or product-line manager might submit two forecasts. The rosier of the two -- let's say a scenario that calls for earnings of 6 cents a share -- may be assigned a 20 percent likelihood, while the other forecast -- predicting 3 cents a share -- has a 75 percent chance of becoming reality.
On the technology side, BPM systems save time and deliver enhanced flexibility when effectively aligned with strong business processes. "The technology helps us carry out our vision of how the analysis and forecasting should function," says Dave Stack, manager of corporate financial planning and analysis for RSA Security, an IT security company based in Bedford, Mass. "We enhanced the efficiency of our reporting and our ability to distribute those reports. Specifically, the technology from Comshare automatically produces a profit-and-loss report by product line. We look at that each month," Stack notes. "It allows us to make sure we're on track and, if necessary, reallocate resources."
Catrambone explains that Network Appliance's forecasting processes were in solid shape before the introduction of technology. "We didn't see the quantum leap you might see if you move from nothing to something," he says of the Longview Solutions software his company installed in October 2002. "We already had a very fast planning process. But we have a lot more flexibility now. I'm not sure the planning cycle is any faster, but the amount of time that someone needs to spend collecting, manipulating and reporting data has significantly decreased."
Like Cisco and many other technology companies, Network Appliance experienced an inventory surge and revenue drop when the Internet bubble ruptured in 2001. Yet unlike many of its peers, Network Appliance remains in growth mode. "Over the past two years, our planning process has focused on prioritizing everything we want to do," Catrambone says. "We put our emphasis on the things we think will bring us the most short-term return as well as on what best positions us, strategically, for several years out." The challenge, he says, is not so much about forecasting the top line as it is about identifying and prioritizing all of the activities the company can invest in to maximize success in a down economy.
For that reason, Network Appliance's planning and forecasting has grown even more efficient and flexible. "The key is having a technology work for you," Catrambone says. "It allows us to maintain fairly consistent processes. The people who need to do the analysis have a tool that limits the amount of time they spend collecting, manipulating and reporting the data. ... It lets us have data faster and helps decision-makers make decisions faster. It also helps us not be so rigid in the planning process. So it's not a big deal if we want to shorten our planning cycle or if we want to look at things a little bit differently."
If a financial analyst senses an opportunity to save money through a more efficient allocation of resources or is asked by a business unit manager to confirm a trend, "we have a tool that allows us to do that very, very quickly," Catrambone adds. "It no longer takes 30 minutes to compile the information. And I think where you see the benefit of the technology is by adding up all those half hours of time."
Nevertheless, performance management improvements don't always go smoothly. Not all corporate cultures are prepared for what deeper analyses, more agile trend detection and better forecasts herald.
"Suppose things are going great, your shares are doing very well and you're forecasting fabulous profits for the next quarter," says Yong. "You have an appointment to go down to the Mercedes dealer and pick out the colors for your S-Class. And then your monitor says, 'Hey, this thing is going off a cliff.' There's a tremendous tendency at that time to either find fault or disbelief. That's an entirely different set of problems in forecasting that technology cannot address."
That's not the only stumbling block. In his new book, "Best Practices in Planning and Management Reporting" (John Wiley & Sons, 2003), Axson examines several other challenges for BPM software implementations. He also identifies three forecasting approaches companies can pursue to sidestep those obstacles:
1. Match your desire for detail with your predictive capability. Axson believes companies need to be more rational about forecasting. "When you look into the future, by definition you deal with uncertainty," he explains. "Many people make the mistake of confusing more detail in a forecast with a more accurate forecast. That is a myth."
Rather, the more details a forecast -- particularly a long-term forecast -- contains, the greater the likelihood that it will be off base. Pinning down next year's budget details today is similar to trying to decide now whether you will want a small or large side of French fries with your hamburger next New Year's Day. More companies are dropping their attachment to small-fry details in long-term plans, Axson says.
2. Move toward a forecasting process that balances financial and operational drivers. Whether they call it a driver-based forecast, a leading forecast or a predictive-indicator forecast, more and more companies are turning from pure financial modeling to an approach that incorporates the effects of possible changes in business drivers.
Ford, for example, recently "built a set of information that enabled it to better understand what types of customers responded to what types of incentives," Axson explains. "Whereas GM had blanket incentives across all models, Ford took a slightly more scientific approach and mapped the incentive to the customer." As a result, Ford increased its profit per vehicle, despite the fact that it was offering substantial discounts and cheap financing. "That's an example of getting out in front of the business not only to improve profitability," Axson says, "but also to get a better forecast."
Other organizations are following in Ford's footsteps. Dell Computer tracks the types of packages people configure on its Web site, whether or not those visitors complete their purchase. This visibility enabled Dell to spot a shift away from traditional CD-ROM drives to CD/DVD rewritable drives. Such insight, when acted upon correctly, has positive ramifications for production schedules and inventory levels. "It's a good example of taking a data point at the very front end of a business cycle," Axson says, "and using that to improve the financial performance. But also it gives you a better sense in terms of the rigor around your forecasting activity."
3. Forecast fewer things more often. The deluge of IT investments has resulted in a severe case of data overload at many companies. "The danger is if you try to analyze and model every piece of data, you'll never finish the forecast," Axson says. "There will always be one more analysis you can do."
Better, more flexible -- and more accurate -- forecasting processes identify which information is important. "So instead of doing a full financial forecast for all P&L and balance sheet line items every quarter, more companies are now moving to a much smaller set of variables that they will forecast but doing them monthly, weekly or sometimes daily." That makes sense. It takes roughly a year and a half to spot a viable trend from quarterly forecast metrics. But if a company tracks a set of data points on a weekly basis, analysts can spot the same trend in six weeks. "Companies will therefore gain much better insight into short-term trends," Axson says, "but they'll also get sufficient forewarning into whether those short-term trends may foreshadow more significant change in the business going forward."
As companies progress along those lines, the traditional budgeting process may fade. Many budgets, derived from lengthy exchanges between corporate financial management and dispersed business units, are obsolete the day they are finalized. "In some companies," Axson adds, "the budget is effectively being replaced by the forecast process."
Why Silence May Be GoldenDan Oakley, director of thought leadership for Ernst & Young's financial services practice in New York City, notes that the most common pitfall of providing hard numbers in earnings forecasts is that the company is making a public commitment regarding future results. "Although the information is accompanied by proper legal disclaimers, analysts invariably accept the numbers as the most likely scenario, on the assumption that the companies have the best visibility into their future results," Oakley notes. "Typically, analysts adjust their own earnings estimates to bring their models more closely in line with company estimates, after factoring in the company's track record for hitting the forecasted numbers. "Providing specific estimates draws a line in the sand for future performance, increasing the pressure on companies to hit the numbers. This kind of interaction is often called a game, but it is a very serious game in which billions of dollars of market capitalization are at stake," Oakley adds. He summarizes the trade-offs of providing specific guidance on future earnings as follows: |
|
|
Perceived Benefits |
Perceived Drawbacks |
|
Improves transparency |
Creates incentive to manage earnings |
|
Reduces volatility |
Creates short-term management goals at the expense of long-term goals |
|
Simplifies communication |
Allows external factors (e.g., interest rate changes) to overwhelm forecasts |
|
Manages investor expectations |
Creates duty to update guidance when material changes occur |