- from the "1st Annual Executive Compensation Conference" (10/04)

Talking Points - Incentive Plan Sandbagging

Mike Kesner is Head of the Executive Compensation Practice for Deloitte & Touche LLP

1. Introduction
   1. Some golf matches can be won upfront by negotiating for additional strokes.  The same is true in negotiating performance hurdles and metrics used in bonus plan calculations.
   2. And, while most Compensation Committee members know performance measure sandbagging when they see it, it sometimes does not become evident until the performance year is over and maximum bonuses are about to be paid.
   3. Most companies link the incentive plan performance targets to the corporate budget, which is often approved by the entire Board, not just the Compensation Committee.
      1. Most budget processes begin in October and finish by December, just before the beginning of the budget year.
      2. In many cases, the budget is considered aggressive based on management’s visibility of the upcoming performance year.  Oftentimes, the budget may be deemed out of reach by the end of the first quarter.
      3. In other cases, management is very cautious about the outlook for the performance year and sets extremely conservative financial targets for the upcoming year.
2. Front-end Due Diligence
   1. In order to approve incentive plan financial targets, Committee members should take a number of steps in confirming the reasonableness of the performance targets:
      1. Request historical financial data for the company (and divisions) for the last three years.  How do the performance year targets compare to actual performance the prior two to three years?  Is the growth rate reasonable?  What has the historical trend been the last 3-5 years?
      2. Review analyst earnings reports.  Where do the budgeted numbers fall within the range of earnings estimates provided by analyses?  If the budget is at the low end of expectations, why?
      3. Obtain past comparisons of actual and budgeted performance.  Does the Company have a pattern of consistently exceeding budget?  Or has performance tended to average around budget each year?
      4. Request a ten year summary of incentive payouts (expressed as a percentage of target).  Companies that tend to crush the budget in some years but fall way short in others may be averaging close to target payouts.  This is usually a sign there is real risk baked into incentive targets.
      5. Review peer group historical and projected data.  Does the Company rank at or near the peer group median in terms of past and projected earnings?  If not, why are you paying target bonuses for below median performance?
      6. Compare 1, 3, 5 and 10 year total shareholder return to peers.  Is the Company performing at or above median during these time periods?  Were bonus payouts at target during this period?
         1. If TSR has lagged to peers, but bonus payouts have been at or above target, it could be an indication of budget sandbagging.  But, it could also be a sign that you are using the wrong financial metrics to measure performance.
         2. EPS, which is widely used in incentive plan formulas, is notoriously disconnected from TSR, even over long periods of time and can be legitimately manipulated (e.g., stock buybacks can help boost EPS).
         3. Similarly, Return on Equity (ROE) can fluctuate wildly, especially where companies are writing off goodwill or paying out large, special dividends (a sure way to boost ROE the following year, as the denominator is reduced).
3. Plan design can also help minimize incentive plan sandbagging
   1. Consider using relative financial performance measures to determine bonus pools.  Median performance yields median pay.
   2. Retain discretion to adjust bonus payouts up or down by 30% of the target amount once the year is over.  That way, the Committee can consider a number of quantitative and qualitative factors in determining final payouts.  (See Campbell Soup and KeyBank’s 2004 proxies for examples of this technique.)
   3. Limit the number of financial measures used to determine bonuses.  Beware of plans that use multiple corporate and divisional measures.
   4. Cap payouts at 150% of the target bonus amount.  Based on experience, companies that far exceed their targets either sandbagged the budget or benefited from events far beyond management’s control (e.g., unexpected rebound in the economy, major competitor has a product recall or employee strike, sharp drop in cost of raw material or sharp rise in commodity price charged for the product).
   5. Rigorously test the “performance curve”
      1. Where does the threshold financial performance begin?  What level of payout is being proposed?  70% to 90% of budget and/or prior year results are common thresholds.
      2. Similarly, how far above budget before the bonus maxes out?  If the max is only 10 to 15% above target, the steepness of the performance curve will result in significant increases in bonus payouts for above target performance.
      3. Is it more lucrative (in bonus dollars) to exceed target than it is punitive to fall short of target?

4. As a general rule, flatter performance curves remove some of the incentive to “game” the system
   6. Consider using a uniform standard year over year.
      1. “Bonus pool equals 6% of profit in excess of 9% return on capital.”
      2. “Target performance is established at 10% EPS growth over prior year; threshold is equal to the prior year results.”
      3. While fixed standards remove the guesswork out of where to set performance targets, it could result in over or under compensation depending on changes in the economy, interest rates, etc.
   7. Make sure incentive plans are self-funding i.e., financial results are net of the accrued bonus.