Performance-based fees in the pension fund industry

From A14

This asymmetry may provide an incentive for the manager to add to the risk of the portfolio by “leveraging up” the underlying portfolio positions.

There is also the question around who designs the performance fee models? These models are generally structured by the asset managers themselves and are only offered when a manager is confident that the arrangement will result in higher fees than the alternative fixed-fee rate.

A more complicated issue pertains to asset manager behaving in ways that are not optimal for the investor. First, the manager is motivated to take excessive risk when “behind” part-way through the year, having no chance to earn a performance fee except by making risky investments.

Second, the manager may be tempted to “shut down” and invest in riskless strategies when “ahead” part-way through the year, so he can lock in the performance fee.

If performance fees are calculated over a period, which is too short, then fees could be levied on performance that was not a result of the manager’s efforts, but from random short-term market movements. Furthermore, the investor runs the risk of a “heads I win, tails you lose” situation when performance fees are collected immediately after outperformance, but not given back in periods of underperformance.

Still further, fees calculated over “rolling” periods have the drawback that recent investors could be charged fees for returns that accrued to the portfolio in the past, but not to their investment.

Of course, the list would not be complete without mentioning the real “joker” in the game – and that is, performance fees can still be charged even if portfolios deliver negative performance.

For instance, if a manager uses the Zimbabwe Stock Exchange All Share Index as its benchmark and the index falls 20%, the portfolio will still have beaten the benchmark and can still charge performance fees if it only falls 19%.

This means that investors can be paying performance fees, even though they are actually worse off. That brings into question the assertion from investment managers that performance fees ensure that investors only pay “when they can afford to”.

Performance fees, where they have been more popular, have also recently become topical among regulators and industry bodies with some regulators highlighting such fees as an area for attention. This is largely as a result, in most instances, of them being poorly designed and too complex for clients to understand.

Limited disclosure and a lack of transparency have also been highlighted among the major pitfalls of these fee models.

The solution is in the design

We will be the first ones to admit, fair and effective implementation of performancebased fees is quite challenging. But, we should not despair and throw the baby away with the bath water.

There are ways to fix the bulk of the issues highlighted here.

By co-designing the performancebased fee model with their asset managers, trustees, guided by their investment consultants, can ensure that they settle for a fair and equitable performance-based fee structure.

The starting point is establishing what the base fee should be. This must be much lower than the fixed fee the manager would have charged on a flat-fee model.

That then is immediately followed with a focus on what the manager’s proposed participation rate is once an agreed upon hurdle rate has been beaten.

The simple rule of thumb is that the base fee and the applicable participation rate should be such that there is a 50:50 chance that the manager’s aggregate fees from a performance-based fee charging model will be above or below a given, and acceptable, fixed fee.

With that agreed on, the real craftsmanship of performance-based fee designing begins. This, of course, assumes that an appropriate benchmark has already been agreed on too.

Among the critical issues of this design work is addressing the thorny issue of the asymmetry of fees. The fee structure should be such that the manager participates both in the upside as well as in the downside – that is, not only earns a performance bonus when he outperforms, but also gives back to the portfolio an equitable fee in the year he underperforms.

Mechanisms to address this would include providing a clear position on when any performance bonuses accrued are actually crystallised.

Added to that, a high watermark provision ensures that the manager does not get remunerated twice for the same outperformance as they recover from the bottom following a period of losses.

Trustees will also need to ask how frequently the performance is calculated. Investing is a long-term commitment and investment skill, as opposed to luck, can only be properly determined over fairly long periods of time – three years being considered an acceptable minimum.

Performance, annualised over such a period or longer ensures outperformance is tempered and trustees are less likely to reward lucky outperformance or too easily forgive under-performance.

Other focus areas include establishing whether the performance fee is calculated before or after the base fee has been deducted, and whether a cap is applicable.

A follow up question would be, if a cap is applicable, is it applied to the performance fee in isolation or whether it is applicable to the base fee and the performance fee in aggregate.

A cap serves to ensure that an asset manager will not simply pursue performance fees “at all costs”. They are also a good way of avoiding the rewarding of lucky oversized out-performance.

Where a manager proposes a performance-based fee, it is good practice for trustees to ask if there is a flat-fee alternative. Ifnot from the same manager, then from other managers offering comparable portfolios.

Trustees need to determine a breakeven point, or “sweet spot”, at which they would be indifferent with either a flat fixed fee or a performance-based fee.

Once they have determined this, it then becomes a question of how exactly they would want to ensure that the asset manager “sweats” for his fees.

Co-designing the fee structure with their asset managers automatically addresses the issue of transparency too as trustees would have been a party to designing the model based on which they then remunerate the manager.

This too will ensure that trustees only settle for a model that is simple and understandable to both themselves as well as their members.

Conclusion

Performance-based fee structures remain one of the most powerful tools to comprehensively address investor-manager agency problems.

They seek to compensate an active investment manager only when he has genuinely added value to an investor’s portfolio. That the practice has been flawed does not make the principle wrong.

It is the practice instead that simply needs to change — and all that it requires is for those on either side of the capital-allocation fence to come together and design models that are fair and equitable to both the investment managers and the pension funds.

—