First, and foremost our growth assumptions need to have realistic implications. For example, over a 30-year period two well known strong performing stocks, McDonalds Corporation and Coca Cola exhibited a 50 and 25 fold respectively increase in earnings. Similarly, each stock price exhibited approximately the same order of magnitude increase over this same period of time. On the other hand, stocks in industries that declined over 30-years exhibited stock price declines over the same time period.

Observations of this nature reinforce the assumption that growth in earnings and therefore cash flows, is directly related to stock market performance over the long run. An analyst, therefore, must be careful to ensure that growth assumptions are realistic relative to the industry wide and economy wide constraints upon the problem. We will consider the implications of economy wide constraints upon normal growth estimates in the next step. But first lets consider the problem of assessing abnormal growth and the number of years in stage I.

The larger the abnormal growth and the larger the number of periods assumed in stage I the more likely the cumulative cash flow projections grow unacceptably large. In addition, forecasting errors associated with the abnormal growth estimates become greater the longer stage I is assumed to be. As a result, it is usual to restrict the stage I forecasting horizon to a manageable length of time such as 5- or 10-years. For strong companies this is most likely being conservative. But part of this conservatism is overcome via the normal growth rate assumption that is applied in stage 2. This assumption is applied in perpetuity and therefore drives a significant (usually a majority) part of any intrinsic value forecast.

As a result, it becomes important to impose a consistency check upon one's assumptions for the number of periods in stage I, abnormal and normal growth in relation to earnings and free cash flow growth over a longer period of time such as 30-years. This quickly reveals whether the combination of assumptions is too conservative or too aggressive.

We turn to abnormal growth first.

Two approaches can be used to estimate abnormal growth: Accessing analyst forecasts and estimating fundamental growth from accounting statements.

Analyst forecasts are usually preferred if available because it is a competitive industry using timely information. Fundamental growth is useful for industries that are essential in a more stable normal phase or if analyst forecasts are not available.

The FTS Free Cash Flow Module supports both and we will illustrate both techniques.

In the above screen image you can observe two buttons. Consensus Growth Rate and Fundamental Growth Rate.

The top number was the consensus (5-year) analyst forecast at the time of the example. The second image contains the estimate of fundamental growth. Clearly, there is a large discrepancy (for IBM but often this is not the case). In addition, the number of years we are assuming for stage 1 is 10-years. So we reject the fundamental growth estimate in lieu of the analyst forecast. You may also want to contrast this with the implied growth rates in analyst forecast for earnings as an alternative check.

Fundamental Growth

Although this estimate resulted in a very high number for IBM it is not always the case that it is this far out. Many times it will result in a reasonable estimate (especially for firms that do not have high growth plus relatively stable earnings. It is defined as:

Fundamental Growth = (1 - Dividends Paid/Net Income)* Return on Equity (ROE) = 33.19% (December 2001)

Fundamental growth takes into account the major firm decisions, dividend policy, investment and financing decisions. Therefore, it is a sound method for estimating growth from especially if you can use economic as opposed to legal dividends. So an interesting variation is to replace the legal dividend number with the economic dividend number (Free Cash Flow to Equity per share).

Summary: In the base case we will accept the analyst consensus forecast of growth as our growth estimate.

Stage I Years: Assessment Issues

IBM has a strong track record in research and development which has resulted in it owning more patents than possibly any other company in the world. As a result, we will assume 10-years for stage I and then later we will also assume that normal growth is higher than the economy average.

For strong companies investors such as Warren Buffet typically assume 10-years for stage I. When assessing the number of years for stage I you should be careful to make the assessment based upon knowledge of the company and the industry. For example, a new startup may have a great initial product but enjoy zero barriers to entry. This means that the number of years for stage I will be equal to the time it takes potential competitors to offer very similar products. Of course, if there are patents in place to restrict this happening then the number of years assumed for stage I would be lengthened. As a result, knowledge of the firm and it's industry is key to resolving this issue.