When do employee and corporate incentives line up?  Ideally, incentive schemes should reward people with a fraction of the additional profitability that resulted from the additional work that they did.  Difficulties: measurement impossible in many cases, people could receive a bonus when the firm is not profitable, neglects synergies (both positive and negative).

Though I wrote that in 2002, I formulated the idea in the late 1980s.  The concept of how bonus/incentive systems should work intrigued me.  Part of it also stemmed from Warren Buffett’s observation that he would never hand out stock options, because employees can’t control P/E expansion or shrinkage, but employees have some impact on profits, if fairly measured.  So Buffett would offer profit incentives, rewarding employees with a share of the profits over a given threshold.

The first time I mentioned the idea publicly was at the Fellowship Admission Course for the Society of Actuaries in 1991.  The first case study was on a misuse of employee incentives, and I commented something close to “rule” that I mentioned above.  After I said that, a female consulting actuary based in Australia said that it was one of the stupidest comments she had ever heard.  But beyond that, she didn’t explain.  The discussion moved on.  I didn’t make too much of what she said, because she offered no reasons for her opinion.

In 1994, my best boss came to me, and said, “Well, you drew the short straw.  You get to try to redesign our compensation system for our representatives.”  He described to me the current system, and what the overall goals were.  I assented, and he left.  Shortly after that, the division’s Radar O’Reilly, “Roy” came to me and said, “You got the compensation project?”

I told him that I had been given the project, and he told me not to put too much time into it early, or it would suck up gobs of time, and besides, no compensation scheme over the past five years had lasted longer than a year.  I thanked Roy, he was a loyal friend, and never told less than the truth.

But then I had to think.  Surely there had to be a way that would work here, and maybe putting in some development time in on the front end could pay off, maybe?

I had been playing around with reduced discrepancy point sets with my free time.  Like Assurant, my boss gave me eight free hours per week to come up with new ideas, and temporarily, I created the best method of creating structured randomness — how best to have “r” points represent an n-dimensional unit hypercube.  The practical upshot was that I could create scenario analyses that were far more accurate than any others around.  (Note: better methods emerged within 10 years, and I never published my work, because my insights were intuitive rather than provable… but it enabled me to do some amazing things for the next ten years.)

I set up my profit model, and chose my criterion: Try to pay commissions equal to 1.25% of the Present Value of the Gross Value Added.  My model had four components.  I can’t remember all of them now, but the last one was the most significant, an item called the “revenue bonus.”  Over a certain threshold offices (with multiple representatives) would receive extra compensation for exceeding targets.

It leveled out the amount paid versus the Present Value of the Gross Value Added.  Success, except that my best boss ever had one of our two fights over it — he thought it was a horrible idea — we could be paying out too much money in a bad year, or too little in a good year.  I argued  that it was better than what we currently had, and that we could tweak it in future years.  We will learn from the errors of the method.  He told me that it was fine for me to present it to the chief marketing officer and the CEO of the division, but he would not be behind me.

Much as I respected him, I had done my work, so I presented it two days later to the CMO and CEO.  They went gaga for the idea, and in the meeting my boss said “I see it now.”  Later, he came to me and apologized, and as is usual with me, I accepted it, saying it was no big thing.

So what happened?  Not only did the compensation scheme work for a year, it stayed in place for four years without modification, while sales and profitability grew dramatically, and the division grew to be the star of the company.

That said, after the CEO retired, the CMO became the new CEO, and I got transferred to a different division to clean up operations andfinancials there.  After four years, the representatives complained that the scheme was too tough, and they needed some low hanging fruit to motivate them.  And so my scheme was abandoned, and sales did not improve, but they were worse.

Profit-based incentives work if they are structured right.  You want representative to write good business, and should incent them to do so.  Offering them a percentage of the expected improvement of the value of the company is a smart thing.

By David Merkel, CFA of Aleph Blog