John Frechette is an economist and the founder of Sourced Economics, an Arlington, Va.-based research firm that provides advisory and analytics services to corporate strategy, finance, and supply chain departments. Views are the author's own.
Today’s volatile economic conditions are raising the costs and level of effort required from CFOs’ finance teams to prepare forward-looking financial plans, and generally making the task of forecasting extremely difficult.
Enterprise forecasts that previously assumed stable conditions in, say, commodity markets are today faced with significant input cost inflation and large unpredictable market shocks, highlighting the need for more
specialization in the forecasting function.
That leaves finance teams struggling to get a handle on the complicated web of spreadsheets and software models which together produce supply and demand forecasts, affecting all aspects of their businesses.
But companies can weather the storm by honing in on the “build versus buy” decisions within those forecasting processes.Too often it seems, these teams opt for build for a variety of reasons and end up overly vertically integrated in their forecasting operations.
Let’s step back. Build versus buy decisions are analyzed daily in the context of manufacturing and development operations. However, much like the transformation of raw materials into products, forecasts go through a similar transformation. Information on national output forecasts is procured, implicitly or explicitly, and internal knowledge is used to transform that information into business unit forecasts. An over- or under-reliance on external “suppliers” then can have serious consequences on the accuracy of a finance team’s forecasts and by extension, its planning and budgeting decisions.
Suffice to say, especially in times of volatility and economic uncertainty, the forecasting process deserves more
attention. It is, believe it or not, common for finance teams to maintain their own forecasts based on proprietary, albeit thrown-together, methods for obtaining global commodity prices.
Understanding where market forecasting, using external projections already available, should end, and enterprise forecasting, using proprietary internal knowledge, should begin can drive significant improvements.
Transaction cost theory
This build versus buy decision then is key to optimizing the forecasting process, and to establishing operational resilience ahead of uncertain economic conditions.
Per transaction cost theory, enterprises “build” where internal transaction costs are low relative to the cost of contracting and exchanging with suppliers, and vice versa for “buy.” In other words, markets can be costly and in certain processes, internal governance is more economical. Where forecasts do not require enterprise knowledge, they should be sourced.
The goal of finance teams, regarding this “buy” decision, is to source increasingly well-informed forecasts, for increasingly relevant markets. Rather than general predictions of energy prices, reliable sources forecasting the prices of sweet light crude oil, if applicable, is preferable. The source of a forecast could be as simple as a website
or government publication, among many other possibilities.
Where forecasts require specialized knowledge of enterprise operations, customers, or products, they should be developed internally. Here, the role of finance teams regarding this “build” decision is to continue to discover ways to combine external forecasts with internal proprietary information, to produce increasingly accurate predictions that are tailored to the company.
A clear example is in demand planning, where technology systems may use only internal historical records to make forecasts. Another example is in cost forecasting where finance managers may maintain forecasts for raw materials with large external markets, with no reference to external forecasts.
Speed and frequency
While in practice predicting prices and quantities is hard especially in certain markets, finance managers can still better prepare themselves for the future by analyzing best- and worst-case scenarios. These can
be performed to estimate the impacts of these scenarios on an enterprise’s overall portfolio supply and demand, products, procurement categories, and even specific geographies.
Finally, there’s the importance of speed and frequency. While one source of competitive advantage is having the best-informed forecasts, another is having the most consistently updated ones.
Determining the implications of supply or demand shocks quickly will become a source of arbitrage. For example, by adjusting inventory management forecasts more rapidly than competitors in response to an event, companies can capture additional market share, or alternatively lower opportunity costs from otherwise idle inventory.
In short, accurate and cost-effective forecasting doesn’t come from a single methodology or data set. Instead, it is an ongoing process to bring together large amounts of disparate knowledge. By understanding the build versus buy decisions involved in developing those foreecasts, CFOs and their companies can better
plan, and prepare for future uncertainties.