The art and science of financial forecast reliability: How to get it right

Financial forecasting is the process of making estimates about the future performance of a business or organization. These estimates are based on past performance, current conditions, and trends. Financial forecast reliability is essential for businesses of all sizes. This allows them to make informed decisions about investments, operations, and strategies.

What does financial forecast reliability mean?

The art and science of financial forecasting involves balancing accuracy and reliability. Accuracy refers to how closely the forecast reflects the actual outcome. Reliability refers to the consistency and stability of the forecast. While it is important for financial forecasts to be as accurate as possible, it is also important for them to be reliable. This allows businesses to make more confident and informed decisions.

What factors affect financial forecast reliability?

Here are six factors that can affect financial forecast reliability. I’ll dive into each one more in-depth below.

1. Accuracy of data: The accuracy of the historical data used to create the forecast is crucial to its reliability. If the data is outdated, incomplete, or otherwise inaccurate, the forecast will not be reliable.
2. Assumptions: Financial forecasts are often based on assumptions about the future. This includes things like future economic conditions, exchange rates, and interest rates. These assumptions must be well researched and consider what the future may look like based on the most recent events.
3. Unknown external factors: External factors could be natural disasters, economic downturns, and changes in government policies. These can all have an impact on the reliability of a financial forecast.
4. Changes in business operations: Changes in a company’s operations, such as new product launches or changes to its business model.
5. Human error: Mistakes in data entry, calculation, or analysis.
6. Model complexity: The more complex the financial model, the harder it is to identify human error and the impacts to changes in formulas.

Accuracy of data in financial forecast reliability

Getting good data can be difficult. Most enterprises have some type of data warehouse to store system-generated data for business intelligence. However, that data can vary significantly in quality and quantity. Unfortunately, most financial professionals still spend about 50% of their time accumulating and cleaning the data they want to use in a financial forecast.

To help ensure that you are using the most accurate data available:

• Use a range of data sources to cross-check and validate the information. This can include financial statements, industry reports, and other external sources.
• Leverage technology to automate the data collection and analysis process. A great example of this is automated timesheet reports to help control hours spent and project costs. This will reduce the risk of errors and improve the reliability of the financial forecast.
• Regularly review and update the data used as the financial situation changes frequently and can have a significant impact on performance over time.

Just because data may be historically accurate, does not mean it will necessarily be useful. For example, during times of great economic stress, the US government has postponed personal income tax filing deadlines from the April 15th deadline. If you are analyzing the performance of a CPA firm whose revenue is concentrated in tax-related work, this shift in timing may be obvious in the historically accurate revenue numbers, but with the tax deadline now returned to April 15th, the revenue cadence of the past few years may not be the best support to estimate for the following years.

Assumptions in financial forecast reliability

Assumptions help to estimate how certain variables will change over time, and how these changes will impact the financial situation. However, it is important to recognize that assumptions are just that—assumptions. Assumptions can vary based on industry. For example, a retailer may have assumptions on sales per square foot, customer retention, or conversion, while a SaaS company will care about annual recurring revenue or the cost to acquire a customer. These assumptions are the cornerstone of financial forecast reliability.

All assumptions are based on a set of expected events and trends. However, these events and trends may not actually occur as estimated or at the time they are estimated. Let’s look at an example. Due to economic indicators, a downward sales trend may be assumed for an upcoming year, but a recession may not be as severe as assumed. In that case, your assumptions would have been wrong and your forecast would have not been accurate. But, it could still be considered reliable as the direction of the trend was correct.

Checking your assumptions

To ensure the assumptions used in a financial forecast are reasonable, question the underlying logic and examine the evidence supporting the assumptions. This can be done as part of a quality control review of the financial forecast by someone who did not have an integral part in building it but is familiar with business operations. For example, the head of FP&A could develop the assumptions and provide support for each one, while the CFO or Controller review the assumptions and their support to ensure they make sense. By thoroughly evaluating the assumptions, you can reduce the risk of errors and improve financial forecast reliability.

In performing scenario analysis, the oscillation of various assumptions can highlight where assumptions may need to be changed given the conditions presented in other parts of the financial forecast. In addition, the list of assumptions should include a range of forecast methods. Some assumptions will lend themselves to quantitative or statistical forecast methods due to a consistent historical distribution. Other assumptions may be best estimated using regression, qualitative judgment, or expert input. For example, unit output is first typically forecasted using a quantitative method such as time series and/or regression. That is then validated or adjusted through expert input. Combining multiple forecast methods depending on the assumptions used leads to a more complete and accurate picture of the businesses’ financial situation.

It is also important to regularly review and update the assumptions used in a financial forecast. As new information becomes available, the assumptions may need to be revised to reflect the changing circumstances. By continuously monitoring and updating the assumptions, you can ensure that the forecast remains reliable and accurate.

Forecast reliability with unknown external factors: Scenario planning and sensitivity analysis

External factors are elements that exist outside of an organization’s control, but that can impact the financial situation. These factors can include economic conditions, changes in laws or regulations, and the competitive landscape. Because they are external to the organization, they are often difficult to predict. Plus, they can have a significant impact on the financial forecast reliability.

The risk of unknown external factors affecting the financial forecast reliability can be highlighted through scenario planning. Creating multiple forecasts based on different assumptions about how the external factors may change will produce a range of potential outcomes. Creating distinct action plans for each scenario, or groups of scenarios, will minimize delays and increase the probability of more positive outcomes.

Similar to scenario planning, sensitivity analysis can also address the impact of external factors. Sensitivity analysis involves analyzing the financial forecast to understand its sensitivity to changes in specific variables. This involves changing a single assumption while holding all else equal. For example, if you could decrease revenue by 10% but keep all other costs steady, what would the effect be on the bottom line? By understanding which variables have the greatest impact on the financial forecast, organizations can focus their efforts on managing those variables and mitigating their impact.

No matter how many iterations of either scenario or sensitivity analysis are done, it is impossible to identify everything. It is also important to note that the future does not mirror the past but often rhymes. Assuming that the future will unfold exactly as it has done before is a very risky proposition. The future is probabilistic and should be looked at in that way.

Changes in business operations in forecasting reliability: Rolling and flexible finance forecasts

Changes in business operations can include the following:

• Shifts in strategy
• Changes in the product or service offering
• Changes in the organizational structure.

During prohibition, many breweries across America switched to brewing non-alcoholic beverages. For example, the brewer, Anheuser Busch, produced more than 25 different non-alcoholic products, including soft drinks, corn syrup, frozen egg products, and malted milk products.

Changes in business operations may be known or unknown. A known change in business operations is the introduction of a new product line or service offering, new locations, or acquisitions. Others can be caused by completely unknown events. To mitigate the effects of changes to business operations on the financial forecast reliability one can implement a rolling forecast or a flexible forecast.

A rolling forecast approach

A rolling forecast involves updating a forecast over a set horizon in each iteration. Traditional financial forecasts start with some type of Budget. This is used as a measuring stick to assess performance over a set period. This Budget is then updated with actuals as time passes while the Budget for future periods stays static. A rolling forecast continuously rolls the financial forecast over a set period (typically 12 or 18 months). For January a traditional forecast may include 1 month of actuals and 11 months of Budget as the forecast. A rolling forecast would include January actuals but would update the forecast for February through January of the next year.

A rolling forecast is a lot more work and is typically used with a less frequent bottoms-up forecast which solicits input from business leaders. Rolling forecasts can use trends, statistical methods, and seasonality as a starting point. One can then layer on any new information that becomes known (e.g. product launches, new locations, acquisitions, etc…).

A flexible forecasting approach

Another way to address the impact of changes in business operations is to use a flexible forecasting approach. Flexible forecasts are adjusted based on revenue and cost changes through a fiscal period. This can account for expected unpredictability in demand for certain products or services. A flexible forecast includes a fixed cost component that does not change throughout the year and a variable cost component that fluctuates based on revenue, reviewing costs periodically to make real-time adjustments. This type of forecast is best suited for organizations with a small or homogeneous offering where the gross margin per offering is relatively steady.

For example, instead of developing a revenue point forecast, a forecaster can provide a probabilistic revenue forecast that fluctuates with resultant impacts to profit. Flexible forecasts demand a bit of imagination and require time being spent on hypotheticals. But they allow businesses to adapt to changing external factors relatively quickly. Unfortunately, many external investors don’t like probabilistic forecasts as they like to hold management to a specific number.

Financial forecasting & human error

Mistakes happen. Regardless of the size of an organization, a financial forecast will most likely be consolidated into an Excel workbook. As it’s been estimated that virtually every Excel model has some type of error, this is less than ideal. Because financial forecasts are created by humans, they are subject to human error. Of course, this can affect your forecast reliability.

Human error can be minimized through the use of check figures, automation, and independent review. Check figures validate that a financial forecast rolls from one worksheet to another. Let’s look at an example. If an organization’s total revenue is $1 billion, and is broken down into 15 business units, then the sum of the 15 units should be $1 billion. It seems so simple, and it can be. But in deriving a financial forecast it is easy to overlook 1 cell or have 1 formula off. And when this happens, it causes it not to roll when you are aggregating and slicing different parts of an organization in different ways for different audiences and purposes.

Mitigating human error

There is a plethora of FP&A software on the market today, and all work basically the same way. By using a multi-dimensional database and Online Analytical Processing (OLAP), they aggregate distributed data from around the organization, consolidate it into one or several views and (if desired) perform online calculations to distribute costs or revenue to various business units or departments. If your financial forecast requires input from many constituents throughout an organization, instead of emailing spreadsheet templates back and forth, the use of one of these systems will help mitigate the possibility of human error in consolidation.

It is also important to have multiple people review the financial forecast to catch errors and identify areas for improvement. People that built the financial forecast should be able to effectively explain it to someone else familiar with the business who can then independently look at the proverbial forest instead of the trees. This adds an additional step and a delay in an already tight deadline environment, but I have found it critical to catching errors or anomalies that were previously going unnoticed.

Model complexity within financial forecasting

Model complexity overlays the five other factors above. The more complex the financial forecast, the higher probability of an assumption, human or formula error, and/or a critical component being overlooked. This is because complex financial situations often involve a larger number of variables and a greater level of uncertainty.

Model complexity can be minimized through the application of the basic economic formula of price times quantity. In deriving a financial forecast all revenue must be generated by taking the price of a good or service times its quantity. There is a practically infinite number of these combinations, but it all comes down to Revenue = Price * Quantity.

Variable expenses are much more straightforward and can be estimated using revenue and assumptions around gross margin. All other operating expenses can be derived based on historical patterns, contracts, estimates based on price, and volume being paid for.

Taking a step back and making a high-level summary of the financial forecast, while keeping the above in mind will help to minimize any inherent complexity within a financial forecast. The developer of the financial forecast should be able to communicate the total financial forecast using 1-2 slides to organizational leadership. If this can’t be done, the financial forecast may be overly complex and warrant simplification.

A final word on financial forecast reliability

Although every effort should be put forth to maximize financial forecast reliability, a practitioner will never be able to mitigate all risk factors. As the owner of a financial forecast, you must be able to accept a certain amount of risk. Some of these risks, identified in the factors above, will be offset with unknown opportunities, while other risks may become blaring obvious only after the fact. A financial forecast can only take into account the best information at the time of its creation.

RELATED POST

How to use Power BI for planning, budgeting, and forecasting

Read more

—
Photo by J W on Unsplash

Facebook

Twitter

Linkedin

Whatsapp

Reddit

Telegram

Kenneth Fick

Kenneth Fick is a seasoned finance leader with more than 20 years of experience in helping companies by leading their financial planning, forecasting, and analysis. He specializes in budget/forecasting, FP&A/CFO solutions, risk analysis & financial modeling, business process optimization, and more much. You can find more of his work and finance thought leadership published in publications like Argyle Group, CFO Dive, CFOUniversity and CFO.com.