14-8-2009 Executive Compensation in the Netherlands 1
Appendix I
In this appendix, a short overview of the principles and assumptions underlying the empirical study on Dutch executive combination will be given. As explained in section I, the population of the research consists of 71 Dutch listed companies. Data have been collected over 3 consecutive years, 2002, 2003 and 2004, and three positions within the executive board have been identified, the CEO, the CFO and the other members of the executive board. The data, as collected from the annual reports, however contained a significant amount of noise and needed to be converted in order to present consistent aggregate data. Therefore, a number of rules of thumb have been used to convert all data into comparable figures, which will be described below.
The study takes into account the full yearly remuneration of the executive board members.
Remuneration of executives that were inaugurated during the year is therefore converted into yearly amounts. If possible, this is done on a pro rata basis. However, in a few cases this has resulted in unrealistic amounts (compared to other years or other members of the board). In these cases was looked at remuneration levels in previous years, remuneration of other board members or bonus payout ratios to come to realistic yearly figures. When uncertainty about the reliability of the final figures still existed, these executives were excluded. This was however only the case in very few examples. Executives that left the company during the (financial) year were not included, since remuneration data concerning these executives in many cases contained too much noise. All executives should have been on a full-time employment contract and should have earned some kind of compensation in exchange for their labour. Interim-executives who were less than one year at the company were not included.
All remuneration is expressed in euros (€). Some companies have reported directors’
remuneration in dollars, pound sterling or any other currency. These amounts are converted into euros. Base salaries, pension benefits, other benefits and perks are converted against average annual exchange rates, since payouts are usually spread over the year. Bonuses are converted against end-of-year (ultimo) exchange rates, since payout is usually at the end of the year (or in the year following). Long term incentive grants are converted against the exchange rate at the date of grant. In all cases where remuneration data had been converted already by the company, these final figures have been used and no additional conversion (against other rates) has taken place 1 . Exchange rate fluctuations that have occurred in the period 2002-2004, have primarily affected executives that were paid out in dollars. Though these fluctuations did have a considerable effect on the income of these executives individually, these will only have minor impact on the aggregated data. Therefore has not been adjusted for these fluctuations, as we have not adjusted for differences in buying power for executives living abroad. The same goes for the fiscal handling of the income of individual executives, which in practice can be a lot favourable for foreign executives (‘30%
regeling’) or executives that receive their income via a private limited company (‘management BV’).
Also, corrections were made for different reporting principles among companies. In particular bonus plan payouts were reported differently by companies. Most of them reported the bonus corresponding to the performance in the financial year in question. This bonus is usually paid out in the year following the financial year. However 19% of the companies reported the actual cash payment in the financial year, which relates to the performance in the previous year. These were therefore not taken into account, which makes the number of datapoints in 2004 less than in earlier years.
1
Additional conversion against other rates, would not have resulted in significant different results.
14-8-2009 Executive Compensation in the Netherlands 2
All equity-based compensation grants were valued in accordance with IFRS 2, using a Monte Carlo Pricing Model. This model simulates future stock price movements on the basis of past stock returns. This method has led to fair value estimates that are more accurate than conventional methods 2 , in particular in the case of relative TSR as performance measure. The Monte Carlo Pricing Model has been discussed in closer detail in chapter 8.
All pension benefits were value using a standardized approach. This means that by means of actuarial methods, the value of the pension plan for the executive is calculated and expressed as a percentage of the gross income equivalent value. All benefits and perks were valued based on assumptions on the private use of the executive. Only the part of the benefit of perk, that has resulted in private gain for the executive (e.g. private use of a company car) has been valued. This value is therefore significantly different from the costs of the benefit or perk for the company.
The number of datapoints of the CFO (n=49 for base salary in 2004) and other members of the executive board (n=58) are in most cases less than that of the CEO (n=70), since not all companies have a CFO position on the board or no other members than the CEO and in some cases the CFO. The absent datapoints are quite evenly distributed among companies of different size and therefore it is unlikely that differences between functions can be attributed to the different number of datapoints. The number of datapoints per component or other sub- analyses of the pay package can also differ, due to companies that do not offer a specific plan or do not report specific information in their annual accounts. All component values are analysed without zeros 3 , unless this zero has resulted from a specific performance consequence (e.g. no annual bonus payout due to bad performance). All published analyses however have enough datapoints to be significant, or otherwise this is specifically stated. It is very unlikely that any biases have resulted from differences in publication or reporting. The number of datapoints per year, per plan, can also differ due to new introductions of specific plans or more companies reporting on a variable in more recent years. In cases this has distorted the information however, this has always been mentioned, or trend analyses have not been reported in these cases. Any differences in the number of datapoints have not affected the main trends that have been signalled in this paper.
2
For example adjusted Black & Scholes calculations. However, in the case of a traditional option plan without vesting restrictions and exercise assumptions, values closely resemble those of Black & Scholes.
3
