Chapter 20 – Management Compensation, Business Analysis, and Business Valuation
20–40
Source: Pradnya Joshi, A Rich Year and More to Come,” The New York
Times, April 10, 2011, p. B7.
20-44 Compensation; Regression Analysis (20 min)
1. Given the available information, the most reliable regression appears to
be regression one, as it has the highest R squared. Regression three is
next best, and regression two is poor, an R squared of only 11%
2. The regressions have some common patterns:
• return on assets (ROA) is not a significant predictor of CEO pay for
any of the dependent variables.
• stock price volatility is significant at the lower level of reliability (.05) in
consistent with the expectation that lower levels of ownership require
stronger incentives, and thus higher CEO pay.
There are mixed results for the remaining variables, passenger load,
CEO tenure, and book to market value. These variables are not
significant in at least one of the regressions, indicating there is
potentially some relationship there but the nature of the relationship
depends on the dependent variable chosen, and that further analysis
would be necessary to understand the nature of the relationships
between these variables and CEO pay.
3. The principal goal of the study was to identify a potential relationship
between non-financial performance, as measured by passenger load, and
note that since many of the carriers included in the study were in financial