FP&A Professionals: Build a Financial Forecast Like A Boss
I sometimes envy my former classmates that went into equity analysis. When they build financial forecasting models it may be used by buy-side analysts as best guesses of what a company’s sales or earnings may be next quarter or next year but as a general rule, the company does not strive to hit them (at least they shouldn’t). In addition, they get to see if they were right.
Conversely, all internal financial forecasts are made with the intent to be used and manipulated. All forecasts are wrong even before they are built, but all internal financial forecasts are made to be purposely manipulated. For example, a target forecasted revenue of $100 million that has been dutifully supported by binders of assumptions and countless interviews will actively work to be out beaten by salespeople working to meet their goals. Conversely, this same behavior may work in the opposite direction in regards to hiring new employees, with or without the sufficient demand. Each may counterbalance the other, resulting in a wash, but the FP&A professional must be keenly aware that unlike the external financial analysts’ forecast, his/her forecast affects the behaviors of everyone within the organization.
So as the head of FP&A or an FP&A Analyst, Manager or Director, how do you build a financial forecast that can be held up 3, 6 or 12 months from its creation as a model of accuracy and success? You build a financial forecasting model based on the following forecasting methods.
Forecasting Methods
Of all the different ways to approach a financial forecast, most methodologies fall into two types: qualitative and quantitative methods.
Qualitative Methods:
Even in the computer age of data-driven analytics and corporate performance management systems, financial forecasts are still guesswork. Whether financial forecasts concern revenue projections, costs, inventory or industry economics as a whole they still reliant upon human judgment. All computers and algorithms are mere tools that rely on some type of human input in order for them to derive a forecasted output. Models derived using qualitative methods are often called Judgmental forecasts as well.
Financial Forecasts that use these methods are more opinion-driven, albeit sometimes using the opinion of industry experts. It involves weighing the conclusions of groups, but it’s considered more of a short-term solution since it relies on impressions rather than hard data. The usual qualitative models include the aggregation of similar, but from different sources. Examples include:
Consumer surveys of large numbers of people to predict responses to new products, services.
Historical analogy
The Delphi method which was pioneered by the RAND corporation and entails “a group of experts who anonymously reply to questionnaires and subsequently receive feedback in the form of a statistical representation of the ‘group response,’ after which the process repeats itself.”
Executive opinions from sales, operations, administration, finance are all averaged to generate a forecast.
Salesforce polling
Some of the major drawbacks to a qualitative approach is:
Can be time-consuming
May impose bias if inputs are required from several executives/employees at different rank in different parts of the organization
Group-think
Overly optimistic or pessimistic outcomes due to inaccuracies in view of broader economic events or strategic decisions
Quantitative Methods
As the name suggests, quantitative methods apply a data-driven approach that tries to take out the subjective human element. This involves not just aggregate income predictions but market variables such as slumping prices or foreign competition. Common quantitative approaches include:
Casual or Associative models explore cause and effect relationships and rely on the relationship being relatively stable over time, so that it can more reliably compare the leading indicators with the lagging indicators, i.e., if you work for Gerber and you are forecasting baby food sales which are probably highly correlated with birth rates.
Econometric modeling is a more data-intensive version of the casual model. Assuming the economic relationship still exists, econometrics tests the relationship over time-based on given data. It can be used to create custom indicators extrapolated from the data. It’s more often used on a larger scale to assess current policies.
Time series models use different techniques to evaluate past data in order to predict future results. These techniques often involve subtle factors such as giving more weight to newer or more comprehensive data sources. It’s the most typical approach because it’s fast and cheap, in that it can be easily applied to data companies already have at the ready.
Utilization of the quantitative methods of forecasting has both advantages and disadvantages. Although it does not take biases related to the human element out of the equation there are several opportunities that make its utilization of these models difficult under some circumstances, for example:
When insufficient historical data is available, such as new product launches and start-up businesses.
Historical data is subject to seasonality, cycles, and trends which may not repeat in the future to the extent they have in the past.
The underlying data available may be unreliable, or parts may be missing or incomplete. This happens due to systems failure or natural disasters.
Which Financial Forecasting Method Should I Use?
Both! You build a financial forecast based on both qualitative and quantitative methods, provide periodic variance analysis and continuously update your assumptions.
Unlike external financial forecasting, internal management financial forecasting is more about the process of risk and opportunities identification, capital allocation and strategy optimization.
An internal FP&A professional’s financial forecast needs to have multiple components and rely upon multiple methods, judgmentally weighting them and averaging them to come to a corporate financial forecast.
Things to Remember in Any Financial Forecast
As accounting continues to become more automated, financial planning and analysis has become an even more vital tool used in small business to large corporate entities. Financial forecasting allows companies to more accurately make decisions on trends such as marketing shifts, consumer tastes, production and marketing and advertising expenditures. Accurate and useful forecasting can make a huge difference to the bottom line. However, it’s still guesswork and involves risk:
There’s no guarantee that factors influencing prior data, especially old data, will still apply to future financial forecasts.
Remember uncertainty. It’s difficult to factor in surprises, such as hefty lawsuits, natural disasters, employee strikes, or other circumstances that dramatically affect data relevance.
Your financial forecast affects the future of the organization and people’s lives. If you forecast a 150% increase in sales and 200% increase in profits based on industry data, research and input from the field, operations, corporate administration and the executive’s input and that does not materialize, people get hurt. Conversely, if you under forecast people leave the organization, the company does not grow and possibly heads toward failure.
As Financial Planning and Analysis Professionals, we frequently rely on the fact that any financial forecast that we create was signed off on by senior management and therefore, at the end of the day, they own it (not us), but as business partners, consultants and professionals, we should seek to “level-up” and educate or clients and leaders with the information to help them make the best choices.
You may come to the conclusion that financial forecasting is more an art than a science, and you would be correct. Crafting your own solution means asking the right questions, understanding the data, selecting the right financial model, and coming up with answers that have actual value to decision makers. Financial forecasting has been a part of business since day one, only the tools have changed. By using the right data, coming to informed conclusions, and weighing it against other factors and consequences you can give yourself and businesses leaders insights they would otherwise not have gained.
What methods do you use to build your financial forecasts? Share with me in the comments or e-mail me.
