Ten Things to Remember About Non-GAAP Metrics in Incentive Design

Since hitting peak sensitivity in early 2016, non-GAAP earnings have raised nonstop questions in financial reporting and compensation circles alike. We’ve discussed this subject quite a bit—most recently for a Reuters story. Here’s a summary of the top 10 things you need to remember about these alternative metrics as they relate to long-term incentive (LTI) design.

1. When it comes to non-GAAP earnings, SEC disclosure is different from incentive design.

Rules govern the use of non-GAAP metrics in financial reporting, the most recent being a number of Compliance & Disclosure Interpretations (CDIs) issued by the SEC in May 2016. In contrast, there are no rules that limit the incentive metrics that companies can use in designing LTI awards. Quite simply, the SEC is not in the business of telling companies which metrics to use for executive performance evaluation.

2. Even so, non-GAAP metrics are a concern to compensation committees.

For one thing, the SEC’s May 2016 CDIs have created enough concern among audit committees that compensation committee members are naturally examining a commensurate response in LTI design. Compensation committees also worry that if they use a non-GAAP metric that’s substantially different from a GAAP metric, and a high payout occurs despite weak GAAP performance, they’ll have some explaining to do. After all, the basic concern about non-GAAP metrics is that they create a misleadingly positive view of firm performance. Finally, proxy advisors like ISS rely solely on GAAP metrics in their assessments of performance.

At the same time, compensation committees remain sensitive to the importance of line of sight, as well as making sure LTI metrics are something that executives can understand and control.

3. Not all non-GAAP metrics are equal in the eyes of shareholders.

In a CFO.com article I wrote with Eric Hosken last year, we ranked non-GAAP adjustments in order of acceptability. The most acceptable is to back out the effect of accounting standard changes, since it’s objectively untenable to set metrics based on current GAAP and assess performance using future GAAP. A more controversial adjustment is to strip out the effect of economic volatility, such as interest rate or foreign exchange gains and losses. Either way, non-GAAP metrics don’t have to be an all-or-none decision.

4. Which financial components should you exclude from an incentive design? That depends on your answers to a few basic questions.

Are they actually normal, recurring business expenses? If they are, plan for more shareholder controversy. It’s hard to defend the omission of a very normal business expense from an evaluation of executive performance.

Could it possibly harm executives to exclude the earnings component in question? If so, that makes the adjustment more defensible. It supports the argument that you’re excluding the components for line of sight reasons and not simply to benefit executives.

What’s the tradeoff between executive line of sight and shareholder value alignment? On the one hand, executives should view the metrics as something they can control. That’s why adjusting for accounting standard changes is unequivocally acceptable—there’s no reason why an executive should benefit or be punished by a change in GAAP. On the other hand, shareholders should agree that the metrics are valid measures of executive performance. That’s why some investors dislike adjusting for foreign currency gains or losses, since shareholders aren’t shielded from these events.

5. Whatever the decision, your grant agreement must be explicit about it to avoid downstream surprises.

Our experience is that if compensation committee adjustments to reported GAAP results are unclear, auditors are liable to conclude one of two things. Either it triggers modification accounting, or it precludes specification of a grant date, therefore causing mark-to-market accounting. Neither outcome is good.

The solution is to spell out what adjustments you’ll make and how you’ll make them. The grant agreement itself should have enough detail for clarity, with the rest going into an administrative procedures document. That way, there’ll be no misunderstandings if an adjustment turns out to harm recipients (e.g., backing out a foreign exchange gain).

6. Two accounting standard changes are coming down the pike: revenue recognition and lease accounting.

The problem with this is that companies are setting incentive metrics before accounting standard changes go into effect. But then they measure performance after the changes are active. If the updated standard buoys (reduces) earnings, executives are artificially benefited (harmed).

Changes to accounting standards should not be the reason for high or low payouts.

There are a couple of ways to manage this dilemma. The first is to specify that final performance results will be adjusted upward or downward to remove the effect of the accounting standard change. The second is to specify that the starting goal will be adjusted upward or downward based on what it would have been had the new accounting standard always been in place. The difficulty is that most companies will not have two parallel measures of earnings that show the with-and-without effect of the accounting standard change. So take whichever approach your systems and processes can accommodate best.

7. If used in a long-term incentive design, non-GAAP metrics require careful handling in the proxy.

Why? Because the SEC’s May 2016 CDI covers all general press releases and disclosures of financial metrics. Since the proxy references business performance, it’s not off the hook.

General uses of non-GAAP metrics in the proxy are subject to regulation G and item 10(e) of regulation S-K (which requires a direct reconciliation of the non-GAAP metric to its corresponding GAAP version). In question 108.01 of its May 2016 CDI, the SEC states that in these contexts companies can provide that reconciliation in an annex to the proxy, so long as a prominent cross-reference is included to call out the GAAP metric.

Even more specifically, the SEC commented that disclosure of target performance levels based on a non-GAAP metric must include a reconciliation of how that number connects with its equivalent GAAP metric.

8. More companies are including stock compensation in any non-GAAP metrics they disclose.

Alphabet and Facebook are two recent, high-profile examples.

Optimists might conclude this is happening because stock compensation is a normal and recurring business expense. If these companies didn’t grant equity compensation, their cash compensation charges would be considerably higher. That makes sense, especially in light of the SEC’s May 2016 CDI.

But there’s another, less optimistic interpretation. It’s possible that companies want to secure the tax windfall benefits of ASU 2016-09 in their non-GAAP earnings disclosures. A prerequisite for doing that is to include stock compensation in general.

In any event, plan for this trend to continue. You’ll see it especially among mature companies where equity compensation looks more like cash compensation (as opposed to startups where equity compensation is more like a lottery ticket).

This specific topic is far from resolved, and it’s one we intend to track as better disclosure surrounding ASU 2016-09 becomes available. Quite simply, there are potentially offsetting effects, and we are concerned with the potential ambiguity they inject. Effect A is the initial recording of compensation expense under ASC 718, which reduces earnings. Effect B is the eventual tax benefit or deficiency upon settlement, which could increase or decrease earnings. Effect B is also largely driven by the stock price, meaning executives could hit or miss their LTI targets in part due to stock price fluctuations.

9. Meanwhile, companies are reducing their reliance on non-GAAP metrics in their financial reporting. This will make non-GAAP incentive metrics stand out even more.

In the Reuters story, author Noel Randewich notes that among the S&P 500, the average wedge between GAAP and non-GAAP earnings is now only 10%. That’s down from 33% in 2015. This suggests that companies really are responding to the SEC’s May 2016 CDI. They’re being more conservative about the adjustments they make to arrive at a non-GAAP earnings disclosure.

For incentive design, this creates a corresponding pressure to be more conservative about selecting metrics and the non-GAAP adjustments behind them. An EPS target with adjustments not explained to investors? Get ready for criticism: “You stopped excluding metrics A, B, and C because they could mislead shareholders. Yet you exclude the same metrics when it comes to calculating your CEO’s compensation?”

So be careful about paring back your non-GAAP earnings disclosures. If you’ve stopped excluding financial components, think hard about whether it still makes sense to exclude those same components from your long-term incentive awards.

10. On top of that, ISS will probably change their quantitative pay-for-performance tests to include financial metrics.

Later this year, we expect to see ISS add a non-TSR metric to its quantitative pay-for-performance test. It’s not that ISS dislikes TSR. They just recognize it’s not the only way to think about organizational performance.

We predict they’ll add a return on invested capital (ROIC) metric, though we’ll all find out for sure in early November. They might pair it with the current relative TSR assessment.

Companies tend to pick metrics that match the way ISS assesses performance. That might not be the most prudent thing to do (incentive design should match business strategy, not ISS’ preferences), but it clearly happens.

ISS uses only reported GAAP results when they look at financial measures like ROIC. This creates a problem for companies that set incentive metrics using non-GAAP measures that substantially diverge from their corresponding GAAP measure. They may run into a situation where their award pays out lucratively but ISS grades their performance poorly. This in turn creates pressure to jettison certain non-GAAP adjustments in favor of tethering award payouts closer to how ISS measures financial performance.

Looking Ahead

There certainly are legitimate reasons to use non-GAAP metrics for incentive design purposes. But as I told Reuters, “The SEC was upset about how non-GAAP was being used, and I think people have tried to listen to that. Nobody wants an SEC comment letter.” Be sure to keep these 10 points in mind so you can satisfy regulators, investors, and their advocates while rewarding executives for a job well done.

Don’t miss another topic! Get insights about HR advisory, financial reporting, and valuation directly via email: