Uncovering Root Causes: The Role of Decomposition Trees in Financial Reporting

Imagine you’re faced with millions (or even billions) of journal entries, and someone asks, “Why did we miss our budget?” Where do you even begin?

Answering the why question!

Answering the why question in finance is rarely straightforward. It usually involves identifying the cause behind a variance, not just restating the reason. For example:

• Reason: We spent more on electricity than budgeted.

  
  

• Cause: There was an unexpected cold spell, increasing our heating needs in factories.

Decomposition trees help us identify the reason

Decomposition trees can help us get to the reasons quickly. From here we can use our own knowledge, or ask the right questions, to get to the cause.

Decomposition trees allow you to start with a high-level variance and progressively break it down into granular components. For example, imagine a $391k positive variance versus plan. The decomposition tree shows this was driven by higher than expected revenues (+$1,441k) offset by higher than expected Expenses (+$1,050k).

You can then drill further into Revenue or Expenses by Account, Region, Country, or Period. In the example, you can see that $76k of the overperformance came from extra revenue in account 10100, generated in Mexico; this was strong in some months (e.g., Period 5) and weaker in others (e.g., Period 7).

This gives us a more nuanced understanding of the reasons behind our performance.

Crossfiltering/Drilldown with decomposition trees

At any point in the decomposition tree, you can cross-filter or drill down to the underlying journal entries. For example, you might uncover a $19,020 journal posted on May 6th.

With this level of visibility, you can start asking valid questions:

• Was there a marketing campaign that week?

• Did we adjust our pricing?

• Were competitors facing supply chain disruptions?

The answers to these questions enable us to provide more nuance in our response to the why question