Incentive Lab: modeling success
Corporate Secretary spoke with Jack Zwingli, CEO of Incentive Lab, the day after the data and analytics firm was acquired by ISS in mid-October. Incentive Lab’s in-depth compensation data and pay plan modeling capabilities are expected to be powerful tools for the proxy advisory firm’s institutional investor clients and the corporate clients in its consulting practice. As Incentive Lab’s founder, chairman and chief scientist Dr Carr Bettis says in a press release: ‘This transaction underscores the importance for both corporations and their shareholders in getting compensation right.’
What would be the best way for ISS to use Incentive Lab’s tools for its corporate clients?
On average, performance-based compensation is now more than half of executive compensation for the CEO and most named executive officers. Being able to use Incentive Lab’s tools to model these awards, value these awards and understand what the impact is of changes in plan design, that by far is the best case as this is looked at as something ISS can provide to companies or compensation consultants to help them better understand the changes they’re making to executive compensation.
You have cited companies’ increased use of modifier metrics, which has added to pay plans’ complexity. What’s the desired effect of modifier metrics for the payout structure of awards?
With a modifier metric, a company may say, ‘The payout is $1 million at target, but we’re going to adjust that depending on what the total shareholder return was over the performance period.’ It’s a way of trying to ensure pay and performance are well aligned. If you had a pretty high level of confidence that your award was going to pay out and that pay and performance would be well aligned, you wouldn’t need to add these additional exotic metrics to keep a cap on pay or try to modify pay based on how the shareholder return – the stock price – moved over the performance period.
What tools will Incentive Lab offer as part of ISS’ suite of services to help companies better understand their awards?
One area we’re focusing on is better benchmarking. We have more in-depth data. If a company is looking to compare itself with its peers, is it using similar metrics? Is it setting goals at similar levels to make sure its plan is pretty well aligned with those of its peer group? Peer benchmarking is one area where we can provide better data.
The other area is analytics – the ability to put a value on these awards and test different award designs. If you add this modifier metric into the mix, what impact is that likely to have on the probability of achieving the award?
What is it about your data that makes your benchmarking more in-depth?
We’re diving into the proxy and pulling out a lot of the detailed information that’s in text form, so it’s not easy to collect. It’s not only what metrics the company is using but how it weights those metrics, how those metrics are related. Do you have to hit both metrics or either metric to get paid? Modifier metrics would be a good example. You don’t capture that data unless you read the proxy statement and manually enter this information. We’re the only data provider that’s been doing that.
And the analytics follow from that?
They do. You can’t value these awards unless you’re able to test how your model works. We’ve collected this data over 15 years, so we can go back historically, model awards, test to see how they actually paid out and make sure our model is effective. No approach is perfect in valuing performance-based awards because they’re based on future performance, but there are much better ways than just taking them at face value or grant date value. While we’ve been collecting that data since 1998, 2006 was the year when disclosure got noticeably better. It’s really over the last eight or nine years that we’ve been able to collect data that’s truly useful in understanding how these awards work.
We’re looking at performance-based awards, we’re valuing those awards and determining what the probability is of achieving them, and to do that you have to make some assumptions about how the performance metrics will do over the three-year or longer performance period. So we’ve been able to go back historically, test our model and validate it. No one else has the data to be able to do that.
The validation process entails looking at what the company actually paid out in performance awards compared with the target amount. Then we compare that with our models. If the target value was $1 million and the award actually paid out at $2 million, what did our model say? If our model said somewhere around $2 million, we’re doing a good job of predicting the real value of the award, not the grant-date value.
Usually compensation awards have a threshold, a target and a maximum level of payout. If you hit threshold performance, you get 50 percent of your bonus opportunity. If you hit the target level, you get 100 percent of your bonus opportunity, and if you hit maximum you get 200 percent. That’s a standard performance award structure. So we looked back and asked how many companies made their target goal. In the most recent year it was about 60 percent of them. If you’re likely to hit the target about 60 percent of the time, it’s not worth what’s in the proxy, what the grant-date value is stated at. It may be worth more or less. What we’re looking to do is help companies get much better clarity on what that award is really worth.
Companies will always want the plan to include the threshold and maximum levels. That’s what they’re trying to encourage – some minimal level of performance that the firm has to hit for executives to get any payment on the bonus. That’s the whole point around performance-based awards. You can get zero, you can get some partial payment or you can even get above payment, depending on how your performance goes. That’s not going to change.
What companies will be able to do is have a much higher level of confidence that the goals they’re setting actually make sense and that they line up with the expectations of the board, the compensation committee in particular.
And presumably, this will help them make a much stronger case in their compensation discussion and analysis?
Absolutely. An article in Agenda recently said goal setting is the number one issue for boards. They’re coming under greater pressure to make sure they can defend how they set up these performance awards, how they pick the metrics. What do they expect the awards to actually pay out at? From a fiduciary standpoint, the board’s endorsement of pay plans is the biggest issue. You’re now paying more than half of CEO remuneration in performance-based awards, so the metrics you pick and the goals you set are extremely important.
Incentive Lab recently did an in-depth study on goal setting. What did you find?
Our report was based on 223 companies, which is a very large sample size, and we found there’s a lot of variance in goal setting. We felt at least 50 percent of the companies we looked at had goals that were either too easy or too hard, which is really high. Companies are clearly struggling with setting goals at the right level, and over time they’re going to have to answer questions about that. We believe we’ve come up with a much better approach to help companies set goals.