TRS Linked to Changes in Expectations

In Chapter 4, we made the case that total returns to shareholders are linked more to performance against expectations than absolute levels of performance. For example, on October 15, 1997, Intel reported that its earnings were up 19 percent compared with the previous year. Intel's share price declined 6.3 percent on the announcement because analysts had forecast a 23 percent increase in earnings. Over horizons of at least 15 years, TRS will be linked to earnings because earnings growth will track cash flow and returns on capital. Over shorter periods, however, we would expect performance against expectations to be more important for TRS than the level of earnings growth.

We conducted a statistical analysis of TRS against various performance measures. We correlated TRS with traditional earnings and earnings growth measures, as well as economic profit and economic profit growth. We also

1 A. Jackson, ''The How and Why of EVA® at CS First Boston," Journal of Applied Corporate Finance, vol. 9, no. 1 (spring 1996), pp. 98-103.

Exhibit 5.1 TRS Linked to Performance against Expectations

Exhibit 5.1 TRS Linked to Performance against Expectations

correlated TRS with the difference between actual economic profit and expected economic profit using consensus earnings forecasts from the Zacks Research System database.

As the theory would suggest, there is a strong relationship between TRS and deviations from expected performance but almost no relationship between TRS and the various earnings measures. Exhibit 5.1 summarizes the results of the analysis. The R2 (in this instance, a measure of how much of TRS is explained by each of the tested measures) for the deviation from expectations measure is 40 percent, high for such a regression.

Valuation Level Linked to ROIC/Growth

It is clear from the above analysis that changes in value (TRS) over short periods are more linked to performance compared with expectations than to absolute performance. A company's value at a point in time, on the other hand, is more linked to the absolute level of performance (i.e., expected sales and earnings growth and ROIC).

As you will see next, the evidence from the market supports this idea. We compared the market value of 340 of the largest U.S. companies with their five-year growth in sales and five-year average spread (in percentage points) between return on invested capital and opportunity cost of capital. The companies' market values were divided by their book values to adjust for size differences.

We then grouped the companies into cohorts with similar sales growth and spread (for example, all companies with average sales growth between 9 percent and 13 percent and spread between 2 percent and 6 percent). We then calculated the average market/book value for each cohort. Exhibit 5.2 shows the results of this analysis. You can see that for any level of growth, higher spread leads to a higher market/book. You can also see that higher levels of sales growth are associated with higher market/book, except for

Exhibit 5.2 Relationship between Market Values, Spread, and Growth

Exhibit 5.2 Relationship between Market Values, Spread, and Growth

low or negative spread companies. This analysis lends support to our argument that the market values companies based on sales growth and spread.

To test the significance of these relationships statistically, we conducted a regression of market-to-book values against various measures, including spread and various growth measures. The results are summarized on Exhibit 5.3. The regression of market-to-book versus sales growth and spread resulted in an R2 of 46 percent, which is very high for such a test. The results were similar whether we used operating profit growth rates or sales growth.

Analyzing Exhibit 5.3 indicated that spread was more important than sales growth and that perhaps growth was not important at all. The theory suggests that growth should matter most when spread is high. (This is apparent from Exhibit 5.2.) We next separated the sample of 340 companies

Exhibit 5.3 Market-to-Book Regressions

Dependent va ri a ble Va ria ble 1

Variable 2

R2 (percent)

12/98 Market-to-book 12/98 Market-to-book 12/93 Market-to-book 12/93 Market-to-book 12/93 Market-to-book 12/98 Market-to-book 12/98 Market-to-book 12/93 Market-to-book

Spread 94-98 Spread 94-98 Spread 94-98 Spread 94-98 Revenue growth 94-58 Revenue growth 94-98 EBITA growth »4-98 EBITA growth 94-98

Revenue growth 94-98 N/A

Revenue growth 94-99

0 0

Post a comment