273: M6 - Forecasting

Question 1

What is the primary focus of forecasting company revenues?

Answer 1

Primary focus is to forecast in company revenues

Question 2

Two categories of forecasting approaches

Answer 2

1) Quantitative
2) Qualitative

Question 3

Quantitative Forecasts

Answer 3

Quantitative forecasts are based on statistical models using historical data to generate numeric predictions. The obvious benefit to this approach is it is tangible and easy to visualize. The downside is that the past is rarely an accurate predictor of the future. If it were, the process wouldn’t be called forecasting, it would be called extrapolation.

Quantitative models are useful when there is a long timeseries of data available and the company being valued and its industry are reasonably stable.

Question 4

Qualitative Forecasts

Answer 4

Qualitative forecasts are predictions are based on the opinions and views of experts or a panel of experts. Essentially these predictions are a judgement call by an informed person or persons. These types of forecasts are particularly useful with start-ups or new product segments of an existing company where little representative, historical data is available.

Question 5

Top-Down Forecasting

Answer 5

In top-down forecasting, we start at a very high, macroeconomic level and work down to a given company’s forecast of revenue. The advantage of this approach is the easy access to macroeconomic data which are the inputs into the forecast. The only company-level data considered is typically market share. The downside is the approach doesn’t consider the process by which market share is captured.

Top-down forecasting is useful for large established firms in reasonably mature industries

Question 6

What is TAM

Answer 6

TAM is referred to as the “Total Addressable Market) (TAM) for a given product segment. The TAM is impacted global macroeconomic GDP forecasts or other macro factors. Once a forecast for a product’s TAM has been derived:

Total Addressable Market (TAM) -> Market Share -> Revenue

Question 7

Bottom-up Forecasting

Answer 7

Product/Service -> Sale Volume and Prices -> Revenue

Question 8

Hybrid Forecasting

Answer 8

If one were to survey equity analysts in an attempt to determine which form of forecasting is more popular they would invariably find that many analysts would describe their forecasting process as top down/bottom up. Essentially, the analyst is combining both methodologies to derive a forecast. It is a rare product that is not impacted by the underlying economy in which the product is being marketed and sold.

Question 9

Statistical methods which we can use to forecast revenue given historical data:

Answer 9

1) Naive Approach
2) Moving Average Approach
3 Exponential Smoothing Approach
4) Trend Analysis Approach
5) Regression Analysis Approach

Question 10

Naive Approach

Answer 10

In the naïve approach, we assume that revenue growth in the future will be the same as the most recent period. The advantage here is the simplicity of the approach, however, unless the most recent past is most representative of the future, this approach is not accurate.

Question 11

Moving Average Approach

Answer 11

In the moving average approach, we assume that revenue growth will be the average over a number of recent years, typically 3-5 years. This approach is slightly more complicated and has the advantage of smoothing growth over a number of years.

Question 12

Exponential Smoothing Approach

Answer 12

The exponential smoothing approach is very similar to the moving average approach, however, when the average is calculated, the different years in the average are weighted differently

Question 13

Trend Analysis Approach

Answer 13

trend analysis relies on the assumption that the data will follow a predictable trend in the data

Question 14

Spurious Relationships

Answer 14

A potential concern that we must consider with using the regression approach is the potential for spurious relationships or correlations between variables. When we run a regression, we are hypothesizing that the independent variables influence the dependent variable

Question 15

Qualitative Analysis

Answer 15

Qualitative factors have the advantage of being entirely forward looking and can be adjusted specifically to the context of the firm and the industry. The downside is that qualitative factors are driven by the opinion of experts with incomplete information and inherent biases that are not observable.

Question 16

Executive Opinion

Answer 16

Nobody understands the firm and its prospects better than the management of the firm. Thus, there is valuable information in the commentary of the executives of the company you are seeking to value. The obvious challenge with this information is that management is inherently biased to be optimistic in their forecasts and is prone to understate of even overlook key challenges. The sample is also very small.

Question 17

Delphi Method

Answer 17

The Delphi Method is a process of gathering opinions from a group while mitigating the potential for one group member to dominate and potentially have his or her opinion overly influence the outcome.

The process uses a sequence of rounds where experts are given questionnaires and the anonymous responses are shared after each round.

The advantage of this process is you obtain an unbiases consensus or near consensus from a panel of experts without one strong opinion dominating the outcome. However, the process it is time consuming and potentially costly, identifying and paying experts to participate.

Question 18

Surveys

Answer 18

Information can be collected via direct surveys of consumers either conducted by the analyst’s firm or by external surveyors. Surveys can also be conducted of retailers.

Question 19

Forecast Items: Sales Revenue

Answer 19

Projecting income statement line items naturally begins with the top of the income statement at the revenue line.

first, you can model sales revenue as a simple growth rate from previous years. This means that any subsequent year is the past year’s sales revenue multiplied by one plus the growth rate.

It is also common to model revenue as a factor of GDP or some other macroeconomic metric. This means that revenue for each year will depend on a regression formula based on historic revenue and the input of that year’s GDP (or other metric).

Question 20

Cost of Goods Sold (or Gross Profit)

Answer 20

Cost of Goods sold is normally forecast as a percentage of revenue. Use historical figures of COGS (or GP) over sales revenue and use these percentages to predict future percentages.

Question 21

Selling, General, and Administrative Expenses

Answer 21

Generally speaking, SG&A expense can be forecast with reasonable accuracy by calculating it as a percentage of revenue on a historical basis.

Question 22

Depreciation Expense

Answer 22

The proper way to determine depreciation expense is to develop a depreciation schedule. The schedule will incorporate new capital expenditures and apply depreciation to the balance remaining. On simplified DCF’s, analysts will often estimate depreciation as a percentage of revenue.

Question 23

Tax Expense

Answer 23

To properly account for future tax expenses, one should default to the statutory rate for the jurisdiction in which the company is domiciled. There can be a considerable difference between the historical rate and the statutory rate but ultimately the company will be subject to the statutory rate.