Michael Oak
Peer comparisons may be helpful to the investment and compensation committees of institutional asset owners—including pension funds, endowments and foundations—but relying too heavily on them can be perilous.
These institutions frequently look to comparable investment offices when evaluating performance and designing compensation. Peer data give an impression of reality, defensibility and objectivity. However, there is often too much deference given to peer data by compensation committees. The information intended for context turns into governance by analogy.
Peer data are often used in three distinct—and often conflated—ways in pay design: performance measurement, pay-level benchmarking and comparing incentive-plan design. Each serves a different purpose, requires different criteria and carries different risks. Treating them as interchangeable is where most compensation committees go wrong.
Measuring Performance: Useful, but With Limitations
A comparison of investment performance with peers may prove useful to the compensation committee, provided that the selection of peers has been done correctly and that the chosen time horizon is long enough, with proper performance attribution. When constructing a performance peer group, institutional asset owners should be chosen based on size, investment strategy, asset allocation, risk profile and overall investment complexity. The time horizon for evaluating relative peer performance should range from five to 10 years, or even longer.
More importantly, a critical question that is rarely raised is: What caused the difference in performance? It is important to conduct performance attribution. Did the difference result from the staff’s activities, the asset allocation policy set by the board, the manager’s recommendations provided by a consultant or market action? In general, most performance differences are due to asset allocation, which is most often made at the board level.
Peer data are messy. Returns are self-reported. Costs have inconsistent treatment across different institutions. Delays from private markets complicate things. There are differing opportunity sets. Moreover, since peer data are backward-looking, staff often lack contemporaneous insight into their own performance, relative to the group with which they will eventually be compared.
There is one more dimension worth noting. Most mission-driven investors are willing to share and learn from each other. But when peer rankings become a competition, that spirit of collaboration begins to erode. It is a subtle cost of overreliance on peer comparisons that rarely appears on any scorecard.
Used carefully, long-term peer performance can offer context. It should never substitute for a direct assessment of whether pay is aligned with an institution’s mission, goals and investment strategy.
Pay-Level Benchmarking: Match the Labor Market, Not the Industry
Compensation peer groups should reflect the actual labor market for talent, the organizations from which you recruit and the ones to which you risk losing staff. For some institutional asset owners, that means other institutional investors. For others, it includes OCIOs, family offices, private equity firms or traditional asset managers.
Many institutions benchmark against organizations they identify with culturally, rather than those with which they actually compete for talent. Institutional identity is not the same thing as labor-market reality.
The trap here is that institutional identity can be considered a valid ground for comparison. This is risky. While two universities or hospital systems may view each other as operational peers, it is the makeup of the investment portfolio that matters more. If you don’t actually hire from or lose talent to those peers, there is no competition for your talent, and they should not be included in compensation peer groups.
A misfit between the compensation peer group and the labor market results in risks of overpaying or underpaying. This situation creates churn for high-performing employees on the one hand, while fostering mediocre performance in the organization on the other. Both have costly consequences for the portfolio.
A simple test is to assess whether your peer group aligns with the organizations you actually compete with for talent:
- Where have your hires come from?
- How long did it take to recruit them?
- How many of your departures were regrettable? and
- What has been the quality of the talent you brought in?
If those answers do not track with your peer set, the peer set is wrong. Institutional asset owners that use compensation data well are competitive enough to attract talent without overpaying and disciplined enough to manage expenses without underpaying.
Problems Designing Incentive Plans
This is where peer data are the most problematic. While a particular design works for one institutional asset owner, it often does not necessarily work when copied elsewhere. An incentive plan designed for a particular organization’s governance model, culture and mission loses its nuances in surveys and informal peer discussions. You may know what common peer practices are, but you will never learn how and why they get results in their respective organizations.
Each incentive plan is grounded in assumptions about the quality of governance, decisionmaking capabilities, risk appetite and organizational culture. The assumptions are hardly replicable across institutions.
Copying a peer’s incentive plan design is plain governance negligence. Not only does it fail, but it fails slowly and is costly to fix.
It happens all the time: Something works at one place, but the same thing bombs at another. The plan might sound sensible in writing, but it ultimately does nothing more than damage credibility and waste resources for many years to come.
Raising the Bar
Directors and compensation committees need to hold themselves to a higher standard:
- Philosophy first. Align your pay system with the institution’s philosophy and people strategy;
- Keep peer groups separate. Performance and pay peers are two separate things. Mixing the two leads to false conclusions;
- Triangulate. Consider a wide variety of information from many perspectives. Avoid confirming your biases;
- Use judgment. Carefully evaluate peer data. Be certain that you consider only those factors over which your employees have control; and
- Adapt, do not copy. Learn from your peers, but customize their systems to meet your needs.
The Questions Peer Data Cannot Answer
Effective governance goes beyond simply comparing compensation to what your perceived peers are paying. It involves judgment, responsibility and an understanding of what kind of behavior the organization wants to promote.
Good governance means asking the right questions: What did the team deliver? How did they deliver it? How did it serve the mission?
Peer data can inform that conversation. It cannot answer it.
It is not about being like your peers. It is about designing a compensation structure that helps your organization achieve its objectives.
Michael Oak is the founder and managing director of Michael Oak Advisors.
This feature is to provide general information only, does not constitute legal or tax advice, and cannot be used or substituted for legal or tax advice. Any opinions of the author do not necessarily reflect the stance of ISS STOXX or its affiliates.
Tags: Benchmark, compensation, Governance, investment performance, peer rankings