Technology

Using Analytics to Improve the Capital Budgeting Process

As industries and economies converge, corporate leaders are tasked with the challenge of making complex decisions using imperfect information.

Capital and resource allocation processes are complex and time consuming in large organizations because of the diverse choices from R&D, Marketing, Manufacturing, HR, IT and other areas. These choices show uncertainties in investments needed, timelines over which such investments are made, technical and regulatory risks as well as ultimate benefits accruing to the firm.

Often, companies utilize qualitative judgments and negotiations to optimally allocate the limited resources and capital available to them. There has been wide agreement that the status-quo is both inefficient and sub-optimal. However, lack of technologies and methodologies to aid this important process, has perpetuated what Jack Welch characterized as the “bane of corporate America.”

Real investments into products, technologies and people, have an economic value, downside risk and upside potential, just as simpler financial instruments. However, real investments are not single decisions without future flexibility, but rather a basket of interacting options driven by many different uncertainties. Organizations who have recognized this and attempted to improve their budgeting processes have been mired in methodology confusion, long and protracted analysis-paralysis and heavy consulting fees. Many have relapsed back to what they have been doing for decades – driven by an annual and negotiated budgeting process. In a regime of accelerating information and dynamic investment choices, a year is too long and such processes too rigid to assure best allocation.

Recent advancements in big data, predictive analytics and artificial intelligence provide an additional level of sophistication in the representation of technical and regulatory risks embedded in investment choices. Gut feel is not sufficiently robust to predict the chance of success of a pharmaceutical product in R&D or its regulatory approval, assessing the probability of success of an emerging technology or idea, forecasting demand for products driven by internal and external factors, capacity and availability of production lines, timelines for product launches and market entries, the timing and quantity of synergies that could emanate from M&A and most such evolving uncertainties that have profound impact on the value, risk and potential of choices facing the firm. Much of the noise in predictive analytics and big data has been in consumer markets that produce information commensurate with the over-inflated term. But businesses can certainly take advantage of these ideas in the enterprise.

The market for analytics has been confusing to say the least. In the 90s, Enterprise Resource Planning (ERP) was the technology of choice and the focus was fundamentally on the collection and warehousing of data. The leaders seem to have accomplished it, albeit, at a cost that has lower shareholder value enhancement to the users. Then came Business Intelligence, a neat concept that promised pretty pictures of historical data – something they accomplished with great fanfare. Whether such constructs impacted decisions in the enterprise is an open question but they did introduce some excitement to the blue and dreary screens of ERP systems.

Lately, it has been big data but the term “analytics,” seem to have been caught up in a time warp. Almost everybody is doing analytics but the real question is whether it is contributing to shareholder value. Data scientists, the contemporary rock stars, are expensive and the data modeling software, technical and unapproachable. This has resulted in companies plunking large amounts of money down the analytics chute with little to show on the other end. Just as disappointments arrived in predictable fashion in the ERP and BI waves, it is possible that we are nearing a break point that may give big data and predictive analytics a bad name in the enterprise.

To impact decision processes related to resource and capital allocation, organizations need to incorporate predictive analytics in an economic framework. After all, analytics is not the end game, better decisions that enhance shareholder value, are.  Decision Options is a methodology and technology developed over the last two decades based on work in industries such as life sciences, energy, aerospace, hi-technology, manufacturing and financial services. It allows organizations to represent investment choices of any type and in any area with a set of consistent constructs that represent interacting decisions. Uncertainties that evolve over time as well as those expected to arrive at future times or even unexpectedly can be represented using a few factors.

More importantly, the platform can be consistently applied across the enterprise, encompassing all departments and external collaborations, providing a systematic view into the value, risk and potential of the entire portfolio of investment choices. It can provide relief to corporate finance and planning professionals, struggling with spreadsheets for most of the budgeting season with thoroughly dissatisfying results. It also incorporates the latest in machine and deep learning to characterize technical and regulatory risks in decision processes and factors that influence time varying uncertainties such as volatility and term structure in interest rates. Such inputs drive the economics of decision choices and analyzing that using market based principles assure that selected portfolio allocation maximizes shareholder value with due consideration to downside risk and upside potential.

The budgeting process is supposed to not only allocate resources optimally but also allow communications with shareholders on metrics that are most important to them. It has been a prevalent scene in shareholder meetings that senior leaders of technical companies, propose large investments because it will cure a disease or invent a new technology, but often fail to articulate the value, risk and upside potential of such choices. More importantly, rigid allocation processes driven by requests for resources and capital from various departments appear to consider investments as binary choices – invest or not. Conventional finance, requiring deterministic cash flows and discount rates have forced such an outcome. However, investment opportunities are typically not binary nor are they now or never propositions and so what is important is to consider the options that are available to the firm – in the quantity and timing of allocation into opportunities that are dynamic in nature. Implementation of a mechanism that is facile, quick and capable of analyzing many choices in a systematic and consistent framework that spans the enterprise, could substantially enhance value. A dynamic budgeting process and more frequent communications with shareholders on the value, risk and potential faced by the firm, can multiply it further.

We are no longer constrained by technology, analytics and processes that make all of this possible. And, we can do that without high resource burns, consulting fees, analysis-paralysis, methodology confusion, systems implementation and many other issues that give senior managers pause. The acid test for any emerging technology or idea is how it impacts shareholder value.

Gill Eapen is Managing Director- Predictive Economics & Management Consulting at Stout Risius Ross. Mr. Eapen’s presentation, “Using Analytics to Improve the Capital Budgeting Process,” will be featured at this year’s CFRI conference on November 14, 2016.