As accountants face an ever-growing volume of data and challenging audit environment, the ability to identify outliers and better understand financial statement fluctuations is crucial. In these examples, I used R, a programming language designed for data analysis and visualization, to build simple but powerful tools for accountants to use. Let’s go through a few analytic examples that accountants can deploy in their processes:
- Outlier identification using box plots and histograms
- Correlation plots for financial statements
Outlier Identification
Here is an example of journal entry analysis for cost of goods sold between two different products.
Box plots: Box plots clearly show the median, lower, and upper quartiles allowing a user to see any transactions labeled as outliers. For Product A, the majority of the values are between $8K-$6K. But in Product B, although the majority of the amounts are around $4K, there are outliers around $8K.
Could this be due to misclassification of cost of goods sold between Product B and Product A?
Histograms: Although histograms don’t call out specific outliers, they help understand the general shape of the data including data skew. In the case of Product B, we can see that there are almost two separate populations, which relates to the potential misclassification between Product B and A noted in the box plot.
Box plots and histograms are examples of how you can use analytics to identify outliers!
Correlation Plots for Financial Statements
As an accountant, have you considered adding a statistical dimension to your monthly or quarterly forecasting? Correlation plots can contribute to this by helping you explore and visualize the relationships between different financial accounts and external market indicators.
Consider the plot below, where the plot reveals the linkages between various ledger accounts and the Consumer Price Index (CPI). Correlation plots yield a numerical coefficient that ranges from -1 to 1, reflecting the nature and intensity of the relationship. A coefficient near -1 denotes a strong negative correlation, while a figure close to 1 signifies a strong positive correlation. Grasping these connections enhances your insight into trend analyses, bolstering your forecasting confidence.
In the graph referenced above, it’s observed that the CPI (Consumer Price Index) and COGS (Cost of Goods Sold) share a positive correlation of 0.69, indicating that an increase in one is likely to be accompanied by an increase in the other. Conversely, the correlation between CPI and Intangible Assets is -0.12, suggesting a weak and negligible relationship between these two variables. Correlation plots serve as a valuable tool for visualizing such relationships, aiding in the accurate interpretation and explanation of financial trends during the reporting process.
Summary
Leveraging analytics such as box plots, histograms and correlation plots allows accountants to uncover detailed insights into the distribution, trends and relationships within financial and journal entry data, fostering enhanced decision making and more precise reporting. Diving into R by experimenting with these graphs on your dataset can serve as an excellent entry point into the programming language. Stay tuned for upcoming posts, where I’ll provide a custom dataset and step-by-step guidance on crafting these plots.
If you have feedback, questions or want to learn more, email me at mike@theledgerproject.com.