Using the Cycle to Your Advantage

There are trustworthy forecasting indexes CIOs can use to time their business decisions accurately.

Two summers ago, Martin Tractor Co. ended its third-best fiscal year ever. Some top executives at the privately held distributor of Caterpillar products felt in their guts that the following quarters would be even better. "Compared with last year's first half, this year looks like a barn burner," the company's president had said that summer.

Still, after looking at several forward-looking economic indexes, the executives grew concerned that the construction sector would soon weaken. Heeding the indexes - rather than their instincts - managers at the Topeka, Kan., distributor cut back product orders and aggressively reduced inventory. Sure enough, the next six months proved difficult for the commercial construction sector, but not for Martin Tractor. "Because of the leading indexes, our inventories were in far better shape than they would have been if we had listened to our [instincts]," CEO Harry Craig said.

For centuries, people have been trying to improve their timing. In Shakespeare's Julius Caesar, Cassius speaks of the "tide in the affairs of men" in which the timing of a decision can lead to either fortune or misery. Can business managers do what Cassius suggests and ride the ebb and flow of the business cycle? As Martin Tractor discovered, the answer is a resounding yes.

Yet, many people believe it's impossible to predict turning points in economic cycles, the moments when expansion halts and the economy begins to contract, and vice versa. One such nonbeliever is Cisco Systems CEO John Chambers, who says, "The brightest people in the world didn't see [the last recession] coming."

I beg to differ. I believe recessions and recoveries can be predicted by state-of-the-art tools of business-cycle analysis. Making sound business decisions based on the expected direction of the economy isn't the sole purview of CEOs and CFOs. CIOs and other senior business technologists can use forward-looking indexes to time the implementation of major enterprise systems, the purchase of costly hardware and software, and the hiring of key knowledge workers. Knowing when to act is as important as knowing what to do.

Two fundamental reasons explain why many have failed in this area. First, most forecasters use models that are severely flawed for predicting the timing of economic turning points. The second, vested interest, is less technical but just as dangerous.

Many economists forecast by extrapolating economic trends. The regression models they use are based on the average relationships among variables over many years, which often break down near turning points. Such models tend to be autoregressive, meaning their forecasts rely heavily on patterns seen in the recent past. The assumption that the recent past is a good guide to the near future may hold true for much of the time periods between turns. But as economic turning points approach, the relationship among variables changes. Hence, the gap balloons between such forecasts and reality, resulting in large forecast errors (see chart, above).

You read or hear views from experts and pundits every day, predicting where the economy is headed next. But especially near turning points in the economy, these predictions are likely to be wrong. Near a peak in the economic cycle, most people see clear skies ahead and are loath to forecast rain. That's why a CIO who listened to sunny predictions as the most recent recession approached might have overinvested in capacity. Likewise, near a trough in economic growth, business leaders are prone to errors of pessimism that are apt to hold back CIOs from pursuing investments that would soon prove critical to meeting rising demand in a recovering economy.

Predictive model

At the Economic Cycle Research Institute, we use a cyclical approach to forecasting, developed by three generations of researchers, that doesn't rely on linear extrapolations. The cyclical indicators that we use are designed to predict future changes in the direction of the economy. They turn before the economy does. The focus is on the timing of a change in direction.

The last expansion is an example. The 1990-1991 recession ended in March 1991, and then we didn't have another recession for 10 years, until March 2001. A forecasting model using economic data from the 1990s will be dominated by the relationship among variables during the 10 years between recessions, and not by the relationships that exist in the relatively short, but critical time around recessions.

Today's forecasting models are likely based on data over longer periods of time, even back to the end of World War II. There have been 10 recessions and recoveries since then, and econometricians have tweaked their models so they appear, in hindsight, to have forecast all those economic twists and turns.

Economists often choose their model parameters to provide the best fit over past decades, periods mostly devoid of turning points. Yet, the parameters break down near turning points, and the precise relationship among variables used in such models changes over time anyway. This is what econometricians call parameter drift. It's no wonder such tools are late in catching up to the reality of a change in economic direction.

Nevertheless, this approach - which does little to forecast the next recession in real time - remains a business mainstay. Because of the data fitting, most decision makers were grossly misinformed of the risk of the first business-led recession in the post-war period, which arrived in 2001. Adopting a cyclical worldview is the best way to avoid these problems, along with timing management decisions to gain competitive advantage.

This cyclical worldview is rooted in our understanding of business-cycle drivers. At each cycle turn, a durable sequence of events takes place, where leading indicators turn in advance of coincident indicators, such as employment, income, production, and sales...

There's no Holy Grail of economic forecasting; no magic indicator that's infallible. Business-led recessions are a good example. As we were recently reminded, the fact that one doesn't occur for decades is no assurance it won't happen. It's critical to monitor a broad range of economic indicators that have proven their worth over time. Trying to anticipate which one will best forecast the next turn in the economy is a fool's errand. No company can rely on only one indicator to anticipate the future.

In the same vein, one leading index, made up of many indicators intended to capture the key business-cycle drivers, isn't always likely to be right. We monitor an array of indexes, each with its own strengths. We use long leading indicators... to make a Long Leading Index.

Another important set of indicators... make up the Weekly Leading Index, which leads the economy by about three quarters. Separate, shorter leading indicators... turn just six months before the economy itself.

Savvy companies use such leading indexes to anticipate big shifts in demand. One of DuPont's major divisions is a case in point. In fall 1998, in the wake of the Russian default and the long-term capital-management or hedge-fund debacle, in what President Clinton characterized as the country's worst financial crisis in 50 years, the Fed made three emergency rate cuts. Recession fears became widespread. But while apprehensive competitors cut prices, the DuPont division held its ground because the leading indexes it monitored didn't show a recession ahead. As a result, the division outperformed its competitors.

In late 1999, with the nation enthusiastically grasping the dot-com explosion and believing the boom was back on track, complacency set in again. Yet, the Weekly Leading Index pointed to a clear cyclical downturn in late 2000. Knowing this, DuPont's division managers took pre-emptive steps, aggressively pushing sales and reducing inventories in the first half of the year, anticipating a lean second half. When industry demand began to slump in late 2000, DuPont was well-positioned, with inventories already pared down.

In another example, executives at a Midwestern machinery maker, with tens of billions of dollars in annual revenue, knew their business was highly sensitive to the economic cycle. Success was clearly linked to two gauges solidly connected to the business cycle: corporate profits and capital-goods orders.

Executives monitored leading indexes, continually adjusting the timing of production and pricing. If they wanted to slow production, they'd order fewer components or schedule time to retool the process. When they wanted to reduce inventory of finished goods, they'd offer more-aggressive incentives to customers.

Nearly a year ago, the growth rate of the Weekly Leading index soared to a 20-year high, giving the machinery maker's executives an unambiguous message that economic growth was set to surge. But they didn't believe it. They weren't alone. Forty percent of CEOs surveyed at the time by PricewaterhouseCoopers thought the economy was stagnant or shrinking. Later, the gross domestic product for the third quarter grew 8.2%, the strongest pace in two decades.

Still, the executives couldn't decide whether to trust their instincts or the weekly index, so they hedged their bet. They modestly increased production capacity. By early this year, production ran full tilt, but demand grew even faster, and as a result, their shipping docks were being emptied as soon as the finished goods arrived. Don't feel too sorry for this company; its competitors are worse off and losing market share.

With advance knowledge of economic turns, people can make tactical decisions with great effectiveness, going against the grain with conviction at just the right times. Knowing what direction the economy will take helps executives decide employment levels. In the latter stage of an expansion, the cost of maintaining or increasing a workforce can be high, with competitors fighting over the best workers. If the leading indicators show no impending downturn, it's advantageous to aggressively pursue employees you'll need. But if a new downturn appears on the horizon, let your competitors keep hiring as you begin to cut back your workforce.

Despite gains in employment in recent months, many managers still are loath to hire because of uncertainty about the durability of the recovery. But, with leading indicators for most industries pointing to solid growth for the rest of the year, you can hire the best available workers without getting into a bidding war with other employers and knowing you'll have more than enough demand to warrant the expense.

Forewarnings of cyclical turns can also be useful in timing the implementation of longer-term, strategic moves. This is especially important to CIOs as they evaluate investments in large IT projects.

Many businesses have explicit plans to make continuing capital investments, including IT, as the business grows. Executing such plans without regard to the cyclical outlook virtually guarantees recurring cash-flow shortfalls and credit crunches. Alternatively, postponing plans for expansion as a cycle peak approaches will leave you with the cash or credit needed to weather the downturn without as many problems.

The same can be said for mergers, acquisitions, and divestitures. Like stock prices, the cost of buying and selling companies changes with the business cycle. Too many companies pay record prices for acquisitions, only to find the economy shortly turning south. Take Time Warner's merger with AOL in late 2000, in which the mass-media company invested top dollar for a company that it no longer includes in its name and reportedly is looking to sell at what's almost sure to be a loss. This is an example of great timing on AOL's part and terrible timing for Time Warner. What a difference a cyclical worldview made in the fortunes of these two companies. Maybe AOL knew something Time Warner didn't? Perhaps the Time Warner executives, like many B-school grads, were properly educated in how to make good tactical and strategic decisions but not how to time those decisions for an optimal outcome.

Those who don't think cycle forecasts can improve decision timing face being stranded on dangerous shoals by a sharp turn in the economy. Clearly, the timing of decisions often spells the difference between success and failure. Lee Iacocca is credited with saying to his chief economist following Chrysler's near-death experience in the early 1980s: "I just want to know six months before the next downturn!"

Sidebar: The Index Upside For CIOs

With the responsibility to plan for major technology investments, CIOs are well-advised to use forward-looking indexes to time those investments to coincide with the vagaries of their industries and the broader economy.

If your industry is poised to suffer lower-than-expected demand, that's not the optimal time to start a large ERP or CRM initiative, when software, hardware, and, more important, talent are at a premium. Rather, begin investments in IT projects when talent is widely available and prices are low. For example, now is a good time to pursue such projects, because the economy is set to grow solidly for the next few quarters, even though lagging pessimism has many managers afraid to make commitments.

You'll likely have to convince your bosses of the merits of a cyclical worldview. The person signing the checks will express an interest when the potential cost savings become apparent. As your industry approaches a peak, the cost of adding capacity will generally be much higher than near a trough. Indeed, you should go the other way and move to lower capacity once a peak in demand becomes apparent in leading indicators such as the Weekly Leading Index. When it's clear no new downturn is looming, move quickly to ramp up to meet rising demand.

As reasonable as this strategy seems, most people don't use it. Instead, they tend to make their investment decisions based on recent events. This approach practically guarantees that they'll pursue a large investment even when their industry is near a peak, and that they'll delay one when the worst is already behind them. Managers who actively use cycle forecasts can improve the timing of their IT investments, and you can, too. -Lakshman Achuthan