“Well I think I've been in the top 5% of my age cohort all my life in understanding the power of incentives, and all my life I've underestimated it.”
Spoken by Charlie Munger, Warren Buffett’s business partner.
Pre-tax returns to investors in a fund will amount to returns from investments generated by the fund minus the fees charged by fund managers. Yet the ramifications of management fee structure extend beyond this mathematical impact. Fee structure also impacts fund performance as a result of the incentives offered to management.
Management fees create a perverse incentive to acquire new funds, which actually harms the interests of existing fund members. However they are designed, management fees do not adequately penalise mediocre or poor performance.
Performance fees can create an incentive to make higher-risk investments. However, this is mitigated if the fund manager has a significant portion of their own wealth invested in the fund. A performance fee, without a management fee, is apt to appropriately align manager and member interests.
Categories of Fees
For the purpose of analysis, fund management fees are placed into three categories: management fees, performance fees and extra fees. Management fees are fees calculated on funds under management, regardless of performance. Performance fees are fees calculated on performance relative to a certain benchmark. Extra fees are fees that do not fit under either of these categories. Such fees include “transaction fees,” “withdrawal fees,” “membership fees” “establishment fees” and various other fees that reduce total returns.
Some money managers simply add an additional layer of fees. For example, members of Colonial FirstChoice Superannuation who choose the Fidelity Australian Equities Fund pay a 1.54% management fee, $60 a year investment fee and 0.2% transaction fees. In comparison, the Fidelity Australian Equities Fund charges investors a 0.85% management fee and 0.25% transaction fees. Colonial therefore takes a 0.69% management fee for doing very little.
Extra fees make it more difficult for investors to discern the true costs of investing in a fund. Arguably, a transaction fee is appropriate to discourage unnecessary transactions that add to the frictional costs of the fund. However, frictional costs can also be limited by developing a strong relationship between investors and management or by requiring advanced notice for withdrawals. Extra fees complicate fee structure without creating useful incentives for managers.
Management fees are charged as a percentage of funds under management (FUM). Therefore, management fees increase as FUM increase. This incentivises managers to grow FUM. There are two ways to do this. The first is to make successful investments. The second is to attract new funds through marketing.
The second path is often the easier one. Such funds usually report monthly (but sometimes quarterly). Each month, investors may be tempted to withdraw funds. The priority, therefore, is to avoid temporary underperformance that might lead to a loss of members or stymie attempts to attract new members. Alignment with the index reduces the chance of outperformance and underperformance alike, but preserves the ability to market the fund. As FUM grow, underperformance becomes more likely, resulting in a stronger incentive to align the fund with the index.
Managers receiving management fees often claim that their interests are aligned with members because they own equity in the company. But managers must own a large proportion of the fund for the effect of greater or lesser returns to outweigh the incentive to profit from fees. The incentive to prioritise acquiring new funds grows stronger as new funds dilute management.
These perverse incentives can be mitigated by either closing funds to new investors or capping management fees at a certain level. However, in the absence of a performance fee such a structure still fails to reward outperformance or adequately penalise underperformance.
Performance fees either replace or supplement management fees. This section considers the impact of performance fees in the absence of management fees.
Performance is measured relative to a benchmark. Above that benchmark, managers share in the returns achieved at the “incentive rate.” The incentives created by the performance fee depend on both the benchmark and the incentive rate.
The most important feature of any performance fee is whether underperformance is carried forward. In most cases, underperformance (that is, failing to achieve the benchmark) will result in a corresponding reduction of future performance fees. This is sometimes referred to as a high water mark. Without such a mechanism, a performance fee merely rewards volatile returns and is completely inappropriate.
Even when underperformance is carried forward, a fund manager could close the fund after a bad year and therefore avoid making up for underperformance. The best way to defeat this risk is for the fund manager to have a large portion of their own net worth in the fund. The threat of reputational damage and personal ties between manager and member also reduce this perverse incentive.
The virtue of performance fee remuneration is that it incentivises high returns on investment rather than the acquisition of new FUM. A performance fee will have no mathematical impact on pre-tax returns when the fund underperforms the benchmark. However, when the fund consistently beats the benchmark, the mathematical impact on returns may well be greater under a performance fee structure than a management fee structure.
Performance fees are not unrelated to the amount of FUM, but increasing FUM alone is not sufficient for the managers to receive their fees; they must still beat the benchmark. Indeed, increases in FUM make outperformance more difficult. The incentive to acquire new funds is likely to be weaker than under a management fee structure, but will depend partly on the benchmark and incentive rate.
A fee structure with a low benchmark and a low incentive rate emulates a management fee and will create a stronger incentive to acquire new funds. The higher the benchmark, the more important it will be for managers to achieve strong performance. However, a benchmark that is very difficult to attain will encourage risk and increase the risk of a manager quitting after severe underperformance.
The most common types of benchmark are:
a) An absolute return, and;
b) A market index.
Comparison of Absolute Return and Market Index Benchmarks
|Type of Performance Fee||Advantages||Disadvantages|
|Absolute Return||Incentivises managers not to lose money under any circumstances. Managers are more likely to avoid holding shares if they think the market is overvalued. Capital preservation is more important to the managers, leading to a more conservatively managed fund.||May pay outsized performance fees, even if returns are largely a result of a buoyant market.|
|Market Index||The fund manager cannot benefit from a rising market alone, but must outperform the market.||In a year where the market produces substantial negative returns, the fund manager can be paid a performance fee, despite the investors losing money.|
These benchmarks are inspired by alternatives available to investors, namely, cash and low cost index funds. An index based benchmark creates more of a disincentive to hold cash in a rising market than does an absolute return benchmark. Both types have their merits; one solution could be to design a hybrid benchmark derived from both an index and a flat rate.
Prior to taking over Berkshire Hathaway, Warren Buffett had various partnerships with various fee structures. Notably, none of these structures included a management fee.
Buffett used the following combinations of benchmark and incentive rate: 
a) 6% benchmark, 33.33% incentive rate
b) 4% benchmark, 25.00% incentive rate
c) 0% benchmark, 16.67% incentive rate
Buffett subsequently consolidated these arrangements into a single partnership with the following fee structure:
d) 0% management fees, 25% incentive rate above a 6% benchmark, and any deficiencies in earnings below the 6% carried forward against future earnings, but not be carried back.
Although it is difficult to compare performance fee structures, it is clear than an inverse relationship should exist between the benchmark and incentive rate.
An absolute benchmark may be more appropriate if members wish their managers to consider cash holdings to be a viable option. It may also be appropriate to adjust an absolute benchmark with reference to prevailing interest rates. For example, it would not make sense to have a 6% benchmark if term deposit rates were at 8%.
The Best Fee Structure
There are many fine fund managers who do not use the ideal fee structure, who do achieve fantastic results for their clients, and who do have high levels of integrity. However, the power of incentives means that as a group, fund managers who incorporate some element of the best fee structure in their practice, should outperform fund managers who do not.
Warren Buffett has described fee structures that combine management fees and performance fees as a “grotesque arrangement” that is guaranteed to make investors (as a group) poorer. Such a structure permits the negative qualities of management fees and performance fees but negates the positive qualities of the latter. Under such a structure, when strong performance is achieved the fund manager makes immense profits, but when the performance is poor the manager still profits handsomely. A hybrid fee structure has the potential to be superior to a management fee structure, although this will not necessarily be the case.
This article contends that management fees are never appropriate, as they create the perverse incentive to focus on acquiring new funds, to the detriment of existing members. Management fees impact performance by encouraging managers to stick closely to the index. As a result of the mathematical impact of fees, the average fund will produce unsatisfactory returns for investors with this fee structure.
Performance fees do not perfectly align management and member interests. However, the main shortcomings of this structure are remedied if the manager invests a significant proportion of their own wealth into the fund. A fee based on performance alone, with a reasonable benchmark, paid to a manager that has a significant equity interest in the fund, is the optimum remuneration structure for fund managers.
 Colonial First State, FirstChoice Employer Super Product Disclosure Statement, 11 June 2013.
 Fidelity Worldwide Investment, Fidelity Australian Equities Fund Product Disclosure Statement, 11 June 2013.
 Warren E Buffet, To My Partners, 22 July 1961
 Berkshire Hathaway, 2006 Letter to Shareholders, 28 February 2007