Unearthing Cash Forecasting Challenges and the Four Drivers of Best-in-Class Cash Forecasts

The four stages of the cash forecasting maturity model are classified as Laggards, Proactive, Strategic, and Best-in-class. Their cash forecasting practices are as follows:

Approach

1. Laggards: Perform cash forecasting based on past experiences and last year’s data along with some adjustments to create recent forecasts.

2. Proactive: Build forecasts through proper collaboration between managers, controllers, etc.

3. Strategic: Create top-down forecasts from FP&A forecasts and set distribution rules.

4. Best-in-class: Create forecast bottom-up from entity-level up to global forecasts.

Data gathering

1. Laggards: Have a manual process and use previous years’ data as a basis for the current year.

2. Proactive: Have a manual process, but use the previous years’ data and add forward-looking adjustments by incorporating the latest data.

3. Strategic: Automate their data gathering process and use previous and current data for cash forecasting.

4. Best-in-class: Use APIs for connectivity between disparate systems to gather real-time information from FP&A, sales forecasts, ERPs, A/P, and A/R.

Modeling

1. Laggards: Use spreadsheets for forecasting cash flow categories. Spreadsheet limits adding multiple variables, causing forecast inaccuracies in complex categories such as A/R and A/P.

2. Proactive: Use the average-based approach, which is helpful for categories like payroll, taxes but doesn’t give enough forecast accuracy for A/R and A/P.

3. Strategic: Use data from open ERP and banks, and adjust due dates to generate A/R and A/P forecasts.

4. Best-in-class: Use current and historical data, and use AI models for A/R and A/P and heuristic models for simple cash flow categories.

Variance analysis

1. Laggards: Don’t perform variance analysis due to lack of adequate data or tools.

2. Proactive: Perform variance analysis at the global level but for a single duration.

3. Strategic: Perform variance analysis at entity and cash category level with proper tools.

4. Best-in-class: Use better tools and models to perform variance analysis at the entity, region, cash flow category level over multiple durations.

To become best-in-class in cash forecasting, it is important to understand the cash forecasting performance levers and how to minimize the challenges associated with the performance levers.

Cash forecasting depends on three key factors such as:

1. Visibility: It refers to the ability to track daily cash movements and view cash forecasts by categories, regions, and entities.

2. Accuracy: It refers to the efficacy of the forecasts required to make confident decisions.

3. Frequency: It refers to the forecasting cadence required to make timely decisions. The forecast frequency can be daily, weekly, monthly, quarterly, or yearly.

Visibility

Visibility is lowered due to scattered data across various sources such as TMS, ERPs, bank portals, sales order systems, etc. Low visibility also stems from non-standard processes and currency fluctuations, which lead to forecast inaccuracy.

Accuracy

Forecast accuracy is lowered due to the following reasons:

1. Process inefficiencies such as:

• Top-down forecasting approach
• Unpredictability in A/R and A/P
• Disparate data sources

2. Dynamic factors such as:

• Changing FX rates
• Seasonality/ business cycles
• Macroeconomic fluctuations

Frequency

The frequency of forecasting is not timely enough due to the high turnaround time in generating those forecasts. The vicious cycle of cash forecasting leads to errors and data overload. As a result, treasurers rely on static and outdated reports for executing decisions.

Visibility

Visibility is improved by building local forecasts from entity-level such as geographies, company codes, currencies, and cash flow categories and rolling them to global forecasts.

Accuracy

Accuracy is increased by following these best practices:

• Leverage suitable models: Utilize specialized models for different cash flow categories: Heuristic models for payroll, taxes, etc., and AI models for complex cash flows such as A/R and A/P.
• Make adjustments when necessary: Make adjustments in forecasts for one-time events such as M&A, dividends, etc.
• Analyze variance: Track variance for multiple time horizons and adjust the models accordingly.

Frequency

Forecast frequency is very subjective since each company has different sizes, systems, technologies, and needs.

• Determine the right forecasting cadence: Focus on the outcome needed from cash forecasts and select the suitable frequency to meet the desired results.
• Automate data gathering: Use technologies such as RPA and API to gather data in real-time from the data sources.
• Leverage time-series algorithms: Incorporate time series to help in understanding fluctuations in cash flows by identifying trends and seasonality.

These are the 10 ways to deliver best-in-class cash forecasts:

1. Understand the cash forecasting purpose

The purpose of cash forecasting needs to be aligned with the company’s cash position and its priorities. Based on the cash position, companies fall under two categories:

• Net debtors: Since they struggle with cash crunches and focus on staying solvent, their purpose is to delay payables and check the balance in their revolvers frequently. Hence, they need to create cash forecasts with high accuracy and frequency.
• Net investors: Since their business is booming, their focus is to find initiatives for acquiring new businesses or expand their own businesses. Hence, they can do with reasonable accuracy and frequency in forecasts.

2. Map the processes in detail

Understand the key cash flow categories such as A/P, A/R, payroll, taxes, etc., and then determine what’s unique about the company and the category that needs to be focused on the most.

3. Understand the drivers of cash forecasting accuracy

Understand if the company’s current situation demands re-forecasting or continuous variance analysis and determine the forecasting cadence.

4. Understand required data vs. the available data

Analyze where data is captured from if it is gathered straight from the source or manually. Understand how you can capture data faster and limit errors. Evaluate centralization against decentralization to understand what suits the company.

5. Understand barriers in cash flow forecast accuracy

Identify if the forecasts are subpar or not timely enough, and identify the barriers.

The inherent barriers for inaccurate and untimely forecasts can be:

• Insufficient resources
• Poor data sources
• Poor alignment between direct and indirect forecasts
• Low priority by senior management
• Poor information access/exchange
• Organizational complexity
• Poor internal knowledge of cash flows

6. Identify the right size technology

Determining the most suitable technology that aligns with the processes, systems, and requirements helps generate fruitful results.

The various treasury technologies and their use cases are:

• API: Quickly and easily exchange information from sources at any point in time.
• RPA: Automate repetitive tasks.
• AI: Produce accurate forecasts and improve results with time.
• Data Lakes: Handle large amounts of data/ unstructured data.
• BI/Dashboard: Visualize data in a meaningful and actionable way.

7. Improve visibility into processes, supplier, and customer activities

There should be enough visibility into A/R, A/P, cash movements, FP&A activities, etc., to make effective decisions.

8. Establish accountability

Find ways to collaborate effectively and easily and gather data from the right people.

9. Harmonize cash forecasting procedures and systems

Establish a standardized process and follow the best practices around cash forecasting across data gathering, modeling, variance analysis, and reporting.

10. Develop a mindset for continuous improvement

Monitor time allocation, mitigate non-value-added activities and set the right incentives to enhance resources and processes.