The disbursements side has proved more problematic. Currently, McCuiston is working with SAP consultants.

In 2004, McCuiston was able to cut in half the frequency with which forecasts missed their mark, down to around 14%. "Most months, the variation is under 10%," says Henry Graef, the parent's treasurer, "and that directly translates into more effective borrowing and investing. We don't over-borrow nearly as much now."

According to McCuiston, the exercise has put the subsidiary on the parent's radar screen. "Nobody from treasury asked us to do this, [but] it's elevated our position with treasury," says McCuiston. "They came over to see what we've got. We've got better numbers but we also have a better relationship with treasury." Another big benefit: Syngenta has managed to pay off its debt earlier and raise its debt rating.

Cash flow forecasting projects do not always pay off so quickly, but there's no doubt it takes less effort to sell them these days. "It's gone from nowhere to the top of the charts as a management priority," says Philip Say, director of solution marketing for MySAP ERP financials. "Planning is now a three-dimensional exercise. The result is longer, more accurate forecasts and therefore smarter decisions by treasurers and by management in general."

TOO MANY MOVING PARTS

But just because executives seem ready to green-light forecasting projects doesn't mean they're a snap to pull off. In fact, even at the best treasuries, cash flow forecasting–particularly when you get into long-term forecasting–is still considered dangerous territory by most treasurers. The problem: There are an awful lot of variables that can derail an otherwise perfectly reasonable forecast. "Short-term forecasting is like looking out the car windshield to see which way the road is turning. Long-term forecasting is like looking at a road map," says consultant Craig Jeffery, managing director of Atlanta-based Strategic Treasurer LLC. "The farther out you go, the more it becomes an art rather than a science."

Today's best practices view of cash flow forecasting recognizes that the information derived from an accurate forecast goes beyond how efficiently cash is used. It can serve as a deciding factor in whether to pursue a merger, extend a franchise globally, introduce a new product line or, more fundamentally, plot the course of future business. It is a demonstration of financial prowess–not as in "my forecast goes out longer than yours," but as a display of efficiency and control that ultimately reflects a company's ability to anticipate and plan. Its most appreciative audience: top management, of course, but also shareholders and the investment community. "We recently spoke with a large technology company that traditionally never worried about its operating cash because it is highly liquid. Now they need to start showing shareholders that they're overseeing funds efficiently, so they're managing liquidity more tightly with centralized cash pools," reports Lisa Rossi, head of U.S. liquidity management for Deutsche Bank Global Cash Management.

There are two kinds of forecasts, explains Glen Solimine, vice president of sales for XRT Inc. One forecasts receipts and disbursements. With a little work, this forecast can be accurate out as far as 120 days, he says. It presumes that history repeats itself and merges a monthly cycle with a pattern of payments such as taxes, dividends and leases. Imported data from A/P (disbursements) and A/R (receipts) are used to show how the cycle is running ahead of or behind last year's receipts and disbursements. Business units report unique capital expenditures.

The other forecast–a longer-range liquidity plan–starts with the business plan and budgets and then filters out accounting conventions (accruals), so that the forecast reflects projected cash in the bank for treasury purposes. Theoretically, a forecasting system can be hardwired upstream to variables like prices in the commodities futures market and correlation algorithms embedded in the disbursements forecast, but that's seldom done, Solimine explains. "Sometimes the view is not worth the climb," he observes.

Forecasting based on modification of historical patterns is well developed and now can be done largely by software, reports Veena Gundavelli, CEO and founder of Emagia. Future progress now lies largely in forecasting based on market intelligence from people in the field who are in touch with customers and who have learned from experience to make shrewd judgments. Those forecasts require teamwork and discipline, aided by technology that makes reporting easier so that various reports can roll up automatically into one consolidated forecast, Gundavelli explains.

David Holland, the treasurer of Cisco Systems Inc. and himself an agile forecaster, agrees, noting that good forecasts are as much about the data as they are about the follow-through to find out why forecasts are off and fix the problem before the next forecast. "In the end, a good forecaster develops a gut instinct for what's going to happen, based on what the forecast tells him or her," Holland concludes.

IT'S A SLOG EVEN FOR THE BEST

Better supply chain management, through such innovations as electronic invoice presentment and payment (EIPP) and purchasing card programs, has enhanced forecasting, adds Emagia's Gundavelli. Supply-chain integration is even giving some sellers direct access to their customers' A/P, planning and sales systems, making big strategic relationships more predictable, she notes.

Nonetheless, at the end of the day implementing a long-term forecast is complicated and time-consuming even for very sophisticated and experienced forecasters. Take, for instance, EDS Corp. It has long done a rolling 12-month cash forecast that doesn't circulate outside of treasury. But last year it implemented a new treasury-led forecasting process that takes free cash flow–as defined by Generally Accepted Accounting Principles (GAAP)–through the next fiscal quarter. The new forecast is used to help management make business adjustments to keep the company on track with what it has told investors, explains David O'Brien, assistant treasurer. "It tells management in what direction we're headed and recommends action to correct variances," he says.

Despite EDS' storied abilities in cash forecasting, the process of long-term forecasting still took almost two years to build and implement. If it takes EDS that long, what hope is there for those still well behind their best practices forecasting colleagues?

It's tales like that one that make some companies hesitate. "It takes a lot of resources to figure out where all the sources of cash are and to build a good cash forecast," concedes Tom Morrison, vice president for product development at Thomson/Selkirk. "Good tools are not a ready-made solution. You have to [know how to] use the tools, and there's the rub. There's a lot of desire to get good cash forecasts, but it will take an economic event like a serious liquidity crunch to get the mandate to spend the time."