Reliable Growth Projections for Business Valuation
Business valuations using the discounted cash flow method typically use two different phases of the surpluses of the firm, an explicit forecast period and a terminal value. The explicit forecast period may examine a period of up to 10 years, more often 3 or 5 years. It details the expected results from the value drivers of the firm. The terminal value builds on this detailed analysis and the assumption of stable future growth of the firm. This is also called the long-term steady state. In most such valuations, the terminal value is substantially higher than the value derived from the first phase. It representing two thirds or even more of the total valuation is not an uncommon occurrence.
Considering this, it is understandable, that the parameters applied for the terminal value require careful consideration and need to be reasonable. Even small changes may have a substantial effect on the quantitative outcome of the analysis. One of these parameters is the growth rate of the financial surpluses that the firm is expected to generate in its long-term steady state. Accordingly, valuators frequently face the challenge to determine a reasonable and justifiable value for this parameter.
A recent article by Gene Trevino in “The Value Examiner” addresses this challenge from an empirical perspective. This article outlines an empirical analysis of several frequently considered methods to determine the long-term growth rate of the firm. Using statistical methods, it examines several data sources for their forecasting quality.
Forecasting Approaches and Used Data
Dr. Trevino introduces focuses his analysis on macroeconomic trends, specifically the expected growth of the overall economy and the inflation, and historical averages. Firm-specific analyses are explicitly excluded, as they may not be suitable for projections up to 10 years. Overall economic growth and inflation are generally considered as the upper and lower bound of long-term firm growth. The argument for using historical industry growth averages is that these should contain information about the industry-specific uncertainties, also for the future.
As overall economic growth, the 10-year nominal gross domestic product (“GDP”) was used. For inflation, the 10-year inflation forecast based on the Consumer Price Index – All Urban Consumers. Both are published by the Congressional Budget Office. The data on industry growth used was the 10-year historical average industry growth rates for 21 selected industries published by the Bureau of Economic Analysis. These industries included broader as well as more specific businesses, such as Manufacturing – Durable and Accommodation.
Method of Analysis and Results
The forecasts were prepared using the data over the period from 2000 through 2010. It was tested against the respective data from 2011 through 2020. The 10-year forecast horizon was used as typical for lost profits calculations. To measure the forecast accuracy, the analysis used the mean absolute error (“MAE”), averaging the annual values. To test the statistical significance of differences between the three forecasts and to identify the differences, Mr. Trevino used an Analysis of Variance and a Tuckey-Kramer Post Hoc-Test respectively.
The lowest MAE, and thus the technically most accurate forecast was provided by the GDP, while the historical industry averages were slightly below this, but the difference seemed small. In line with the expectations, the inflation-based forecast provided the least accurate forecast. This on average as well as for the individual industries. The quantitative analysis confirmed that the differences between all three forecasts were statistically significant with regard to the inflation-based and the other two forecasts. The differences between the GDP- and the inflation-based forecast were found to not be statistically significant, leading to the recommendation that long-term growth rates should be based on overall economic growth or historical industry averages rather than inflation expectations.
Our Views and Comments
While the analysis focuses on growth for lost-profits calculations, we consider it also helpful for transfer pricing (“TP”). As business valuations for TP are mainly prepared for tax purposes, they require a higher emphasis of the verifiability of their parameters. Practitioners therefore are under a higher pressure using external data rather than professional judgement in their models. The analysis confirms that GDP-forecasts and historical industry averages may provide a reliable basis. Especially where the TP valuation is not related to the US-market, a independent confirmation may be necessary, but the methodology used by Mr. Trevino seems more than reasonable and should be relatively easily replicated, where the input data is available.
The article “”Forecasting Growth: Macroeconomic Indicators versus Historical Averages”” by Gene A. Trevino was published in The Value Examiner on page 16 through 20.