Friday, January 21, 2011

Performance Fees - Good or Bad?

Our second "guest blogger" for the year is Carl Bacon, CIPM. Carl is the chairman of Statpro, an internationally recognized authority on performance measurement and risk, an author, a consultant, and currently a member of the GIPS Executive Committee and chair of the GIPS Verifier/Practitioner subcommittee. 

After a brief lull during the credit crisis performance (or incentive) fees are again becoming increasingly high profile.  An excellent time therefore to pose two questions:  Are they a good thing? And if used, how should they be structured?

Supporters of performance fees would suggest that they are desirable because they align the interests of the asset manager with the investor. If the manager performs well, the investor's assets rise in value.   For the most part investors appear happy to pay performance fees for good performance. If the manager performs poorly only a smaller base fee and no incentive fee is paid. For more see the following from the CFA Institute:

Those not in favor claim that rather than aligning interests, performance fees create a conflict of interest, and for the most part are biased in favor of the asset manager. Certainly during my 25 year plus career in asset management performance fees have had a beneficial impact on revenues.  In the first 10 years the pressure on fees trended down.  For the last 15 years average fees have increased significantly across the entire industry. Hedge funds demonstrated to the entire industry that clients, including institutions (and the hedge fund industry has become more institutionalized during this period), are prepared to accept high management fees if they are dressed up in the form of a performance fee.  Hedge fund clients first accepted 2 % annual fees plus a 20% incentive fee, and just before the credit crisis 3% annual fees and 30% incentive fees(or even higher) were not uncommon.  Recently fees may have dropped back to 1.5% and 15% but are already showing signs of reverting back to 2% and 20%.

Of course the answer really depends on how the performance fees are structured. There are many flavors, but essentially there are two main types; asymmetric and symmetric. Asymmetric fees shown in the diagram below are by far the most common and are very seductive to investors.

Typically asset managers will present these types of fees to clients at a base fee rate considerably lower than their normal management fee with an incentive fee only for performance greater than a hurdle rate.  Clearly there is considerable variation in the hurdle rate, angle of the participation rate, time period, excess or absolute return, caps and collars and perhaps a high water mark but the basic structure is the same.  This is attractive to investors; they are paying a lower fee and will only pay higher fees if the manager performs well.

A rare alternative would be a symmetric structure. In this type of structure the base fee would probably not be lower and may even be higher than the normal management fee. Incentive fees are paid for out-performance, but crucially fees are rebated for under-performance to the extent   that the asset manger may ultimately make a net payment to the client if performance is sufficiently bad. Though unpopular with asset managers, these types of fee genuinely align the interest of both parties.

Some critics of performance fees make the claim that asymmetric fees in particular encourage managers to take more risk if performance is poor because they have little further downside and considerable upside. This is not my personal experience.  Often when in a hole, I've observed that managers in effect stop digging and take less risk when performing badly.  On the flip side I have observed managers "lock in" out-performance when they have achieved the cap and take much less risk. This obviously reduces the business risk of the asset manager but presumably the client would like the manager to continue talking risk if they are obtaining good rewards. Risk-adjusted returns such as M2 would help alleviate this problem.

Sadly performance fee arrangements often end in acrimony. For long term agreements the original authors on both sides may have moved on, and if badly written at the point when both partners should be celebrating good performance, relationships can be damaged by a dispute about the size of fees to be paid. Performance fee agreements should be clear, concise and as simple as possible. I would always recommend including a few worked examples to ensure both parties fully understand the agreement. There is a tendency to complicate performance fees with high water marks and variable hurdle rates; frankly the more complex the agreement the more likely interests will be misaligned.  Keep it simple.  Investor interests are best protected by structuring performance fees over longer time periods, three or even five years. 

It always surprises me that endowment or pension funds that are selecting a portfolio of managers to diversify their manager selection risk require performance fees as a matter of principle. I can't prove it, but my observations over many years measuring the performance of managers with and without performance fees would suggest the implementation of a performance fee does not actually improve performance.

A portfolio manager is either skilled or unskilled. Those extra hours staying behind in the office will not necessarily add to performance.  Sufficient constraints and incentives already exist such that managers are always encouraged to deliver good performance with or without performance fees.  In order to attract new clients managers must demonstrate a superior, consistent track record, they often have their own money invested and if they have two clients with the same strategy, one with a performance fee and one not, legally they are simply not able to direct their favorable trades through the performance fee client.

Logically the investor is choosing the asset manager.  Why if they are skilled in their choice of manager, and choose good performing managers,  should they wish to penalize themselves with a higher performance fee? And, if they choose unwisely, reward themselves with a lower base fee for underperforming managers?  Worse still, if they are in the majority and merely average, the performance of the good performing managers will offset the underperformance of the poor performing managers leading to average performance overall.  But since performance fees are asymmetric in nature the performance fees will more than exceed the advantage gained from the lower base fees of poor performing managers and investors ultimately pay above average fees for average performance. Investors should treat performance fees as a necessarily evil, only accepting performance fees to access demonstrably superior managers where a simple base fee is not available.

There is a place for performance fees but if used they should be:
  1. Unambiguous and fair to both parties
  2. Symmetrical
  3. Risk-adjusted
  4. Simple

Unfortunately, many performance fee structures are unfair and ambiguous, asymmetrical not symmetrical, rarely risk-adjusted and frequently complex.


  1. Totally agree - the way the investment management industry will earn a living will different going forward. Whether the move will be to one simple base fee that reflects economies of scale or a symmetrical fee model will be interesting to watch.

  2. Thanks for your comment. Glad you like Carl's contribution.


Note: Only a member of this blog may post a comment.