Rataj and Yeanoplos provide some practical guidelines for calculating reasonable compensation

Today's session on reasonable comp started out with agreement that there is no standard definition that works in all situations.   Kevin Yeanoplos (Brueggeman and Johnson Yeanoplos) said his working theory is that reasonable comp is the number he'd have to pay to hire some one to do exactly the work an owner is doing.   He likes to think of it as "reasonable replacement compensation."   This eliminates payments to kids, or other related parties--essentially owner's comp disguised as employment.

He and Ed Rataj (CBIZ Human Capital Services) discussed some of the formulas and case law that support reasonable compensation calculations.   Rataj still feels like the IRS definitions are helpful to practitioners.  The IRS focuses on two issues when looking at what an owner is paid out of a business:

  1. Intent--is the payment in exchange for services provided?, and
  2. Amount--is the amount reasonable given the services rendered.

Tests for reasonable comp in the IRS service manual include:

  • the nature of the employee's duties
  • background and experience
  • the size of the business
  • the employee's contribution to the success of the business, and
  • the amount paid by similar size businesses in the same area to equally qualified employees for similar services

Yeanoplos reminds analysts that you need to consider all types of compensation before you start an analysis...stock appreciation rights, phantom stock, phantom family members, deferred compensation, restricted stock, grants, or stock options, industry-specific incentives (design wins in architecture, for example), etc.

It's even more difficult when the owner plays multiple rolls--the CEO, chief R&D officer, chief legal officer, receptionist.  Rataj includes all of these things as part of the CEO's overall ownership of all company activities (which generally means taking the highest compensation for all the different activities performed)--all, except for sales commission.

Trying to collect multiple points on each position helps, though "you can't blindly take the median," says Yeanoplos.  If you use ERI, or Watson, or Mercer, or some other source, you'll even get different numbers by demographic--industry, employee count, revenues, region, city, etc.  "In one apparently straightforward reasonable comp case, we had a range of $210,000 to $450,000 for the same position...from just two of our eight sources," Rataj commented by way of illustrating how difficult market pricing considerations can be.

Rataj believes the three most important considerations are:

  • company size (revenues is the best indicator generally)
  • industry (it often depends on whether industry knowledge is critical), and
  • geographic location (to retain executives, it's best to look at the higher of national or local rates)

After collecting data points from their multiple sources, Rataj then looks at the median; after first, absent any performance information (for instance, the CEO is in the top 10% of the market), eliminating the bottom 25% of their data set, and the top 25%. "But," reminds Yeanoplos, "the numbers themselves are meaningless.   If some one is making twice the market, there may be reason.  Perhaps they're working twice as hard, or they're more productive in some key metric like collections.  You can't take the median, or any other number, until you understand the individual's performance and role in the organization."