Probabilistic Financial Planning and Capital Allocation

Situation Analysis

Long-term financial planning is crucial for organizational stability and sustained growth. It guides decisions around:

• Resource Allocation
• Investments
• Budgeting
• Information gathering
• Operational planning

However, traditional forecasts often rely on single-point inputs. While informative, this approach leads to more questions:

• “Are all inputs assessed under a specific scenario?”
• “What is the likelihood that the assumptions for each input are met?”
• “How optimistic are the results?”
• “How do correlations impact the results?”

The financial team of a fertilizer producer asked these same questions of their 5-year planning process. The financial team realized that probabilistic financial forecasting could offer additional insights into risk and probabilities, addressing the questions above. However, their 5-year plan rested on several separate and complex models. They wanted to keep the existing process in place while integrating probabilistic results. Strategic Decisions Group (SDG) was called on to structure and implement this probabilistic functionality.

Discovery & Solutions

SDG worked with the client to audit past performance of their 5-year plan forecasts against actual performance. This analysis identified the factors contributing the most risk and uncertainty to the forecasting process, which helped SDG create a fit-for-purpose scope for the project. The team pinpointed several inputs that required probabilistic functionality, with commodity prices being the most significant. Commodity prices can be volatile, but they also have varying degrees of correlation, a key factor to capture when evaluating portfolios and financial forecasts.

SDG then conducted a full examination of the 5-year planning models to understand how probabilistic inputs would behave in each one. It is vital each model produced a realistic scenario in every iteration of probabilistic simulations. This involves upgrading some model logic to handle changing inputs.

Given the need for multiple Excel-based models to work together, SDG worked with the client to explore the most suitable implementation methodology. SDG ended up building a probabilistic engine that would be held in its own file which connects to all of the other files’ inputs through external connections. With all the models open simultaneously, inputs flowed from the probabilistic engine to each 5-year planning model in parallel and ultimately condensing into company-level outputs.

Results & Impact

The probabilistic results have been integrated into the annual 5-year plan outputs. Tornado sensitivity charts and a distribution around cash from operations are two key outputs the client incorporated into their communications. Probabilistic outputs have allowed the client to put their deterministic 5-year plan forecasts into context and assign their traditional outputs to a percentile value.

The success of this work led to another project focused on allocating investing capital. Leveraging the probabilistic 5-year plan, the capital allocation team wanted to overlay potential investment projects to evaluate their existing business and new potential projects together. Probabilistic financial modeling made this possible, as risk cannot be quantified without defining the range of uncertainty and correlations between key inputs.

SDG returned to create a new tool for the client. It extended the 5-year plan to a 20-year outlook on the existing business and included the ability to mix-and-match any number of investment project into that 20-year outlook. The probabilistic engine was enhanced to handle any number of portfolio combinations. This allowed the client to create an efficient frontier of potential portfolios according to full enterprise value and full enterprise risk.

A key value-add was quantifying the impact of price correlations. Some projects were viewed as “risky” in isolation but had a minimal impact on the overall portfolio risk because of insulation from commodity prices.