Now that we've dispensed with the conceptual (touchy-feely?) case for performance fees, you're probably asking yourself: How the heck would these things work?

Here's how we envision it: The fund would pair a modest base fee--maybe 0.20%--with a very significant performance-based adjustment, say one of every two basis points of gross outperformance outside of a specified corridor (i.e., 150 basis points above or below the benchmark).

So, if the manager beats the benchmark by three percentage points before the base fee, it earns a 75 basis point performance adjustment (i.e., collects 95 basis points in total management fees; investors get the other 205 basis points of gross outperformance).  If it underperforms by that same amount, the manager is on the hook for a 75 basis point performance adjustment payable to the fund (i.e., owes 55 basis points to the fund, investors incur the other 2.45% in gross losses).

Are performance fees always the answer?

This question is especially timely given the trend towards outcome products, which turn the notion of ‘outperformance' on its head. The answer is --no, of course they're not. Nor are they guaranteed to light a fire under fund sales right from the get-go. But in circumstances where funds are sold under the pretense that outperformance is possible, it's very hard to argue that they're inferior to the traditional fee structure.

Will these fees encourage managers to recklessly swing for the fences?

There will always be a small contingent of managers that, fee structure notwithstanding, take imprudent risks. As I've pointed out above, the traditional fee structure disproportionately penalizes the investor-who shoulders every basis point of underperformance when a manager's call blows up royally-not the manager. So, it's very hard to argue that a fee structure that puts the manager on the hook is a bad thing.

That's not to say it couldn't encourage previously sober-minded managers to take bigger chances than before. To that I'd say - good! The idea is to foster risk-taking and discourage benchmark-hugging masquerading as active management. We want managers to be bold because that strategy is consistent with success. Of course, those bets must pay off in greater proportion to any excess risk the manager is taking. But remember that the performance adjustment is symmetrical (excluding the base management fee, that is). If the manager blazes for glory only to blow the fund up, then it will share in that misfortune with investors. That reality--the outsized potential gains and the heightened risk of paying a price for underperformance--should impose its own form of discipline, forcing managers to strike the appropriate balance between caution and daring.

Is the backward-looking nature of performance-based fees a fatal flaw?

And what of the argument that performance-based fee structures are fatally flawed in the way they're applied-prospectively based on historical performance. That is, you charge investors for something that's taken place in the past, thereby potentially charging investors extra for performance they never enjoyed (or, conversely, cut them a discount for underperformance they never suffered). Investors' tendency to chase performance-buy at peaks, sell at troughs-accentuates the concern.

This seems like a surmountable challenge. For instance, a firm could create two classes of shares-one that levies a traditional, flat management fee only, the other that employs a fee structure similar to the one I've described--a modest base fee along with a performance-based adjustment. When investors initially invest in a fund, the fund would issue them the shares that levied the flat management fee. Those shares would then convert automatically after a certain minimum time period-1 year, 3 years, etc.-had elapsed. (This process isn't much different than the B- to A-share conversion that takes place with broker-sold funds.) In that way, you avoid them disconnect between investor experience and the performance-based fee that's levied.

How do you keep investors from trying to trade around performance adjustments?

So, what if an investor bought into a fund that utilized a one-year performance measurement period, the fund blew past its benchmark over the course of those 12 months, and the investor sold just prior to the one-year anniversary, when the shares would have converted to the performance-based fee class? In such a scenario, the investor could sidestep the performance adjustment, yet participate in the fund's run-up.

The best way to combat that problem is by adopting a formulaic redemption fee structure. Under this approach, investors who bailed out would have to pay the higher of a fund's traditional flat management fee or a performance-adjusted fee.

To illustrate, let's suppose the following:

• Performance measurement period: Three years
• Investor's holding period: One year
• ‘Traditional' Share Class Fees: flat 0.20% fee
• ‘Performance' Share Class Fees: 0.20% base fee, one basis point adjustment for every two basis points of out/underperformance vs. S&P outside of 1.5% corridor
• Example 1: Fund returns 8% vs. 5% for S&P during investor's holding period
• Earns 75 basis point upward performance adjustment (3% outperformance less 1.50% corridor = 150 basis points excess; one basis point for every two basis points of excess equates to 75 basis point adjustment)
• Example 2: Fund returns 6.5% vs. 5% for S&P during investor's holding period
• Earns no performance adjustment (1.5% less 1.5% corridor leaves no excess)
• Example 3: Fund returns 2% vs. 5% for S&P during investor's holding period
• Incurs 75 basis point downward performance adjustment (3% underperformance less 1.50% corridor = 150 basis point deficit; one basis points for every two basis points of deficit equates to 75 basis point adjustment)

How would the redemption fee work in these situations?

• Example 1: Investor pays 75 basis point penalty equal to performance adjustment
• Holding period composite fee = 0.95%
• Example 2: Investor pays no penalty
• Holding period composite fee = 0.20%
• Example 3: Investor pays no penalty, but doesn't get the 75 basis point fee reduction.
• Holding period composite fee = 0.20%

(If need be, the manager could also levy an additional redemption fee, say 0.40%, in which case the investor would pay the higher of (a) the sum of the base fee and redemption fee or (b) the sum of the base fee, redemption fee, and performance-based adjustment.)

In example one, the manager would share the redemption fee penalty proceeds 50/50 with remaining shareholders. In that way, we discourage the manager from gunning for short-term gains to the possible detriment of the strategy's long-term success.

Example three is trickier. Technically, there are no redemption fee proceeds to divvy up. After all, the fund is lagging the benchmark by a significant amount and, thus, the manager would be on the hook to reduce the base management fee should the underperformance continue through the full three year measurement period. But at year one, the redeeming investor forgoes the chance to receive a fee reduction.

What makes it tricky is this: If a manager was lagging the benchmark by a significant amount and wished to avoid having to pay a performance penalty to the fund, it could try to drive off shareholders by intentionally sabotaging the fund's performance. In such a scenario, because investors were fleeing before the measurement period had fully elapsed, they wouldn't receive the fee reduction and the manager would be off the hook.

As a practical matter, however, we doubt this would arise frequently. Any cost savings a manager might achieve in such a situation would pale in comparison to the damage it would wreak on its reputation.

How do you ensure that the manager won't liquidate a fund to avoid making a performance fee payout?

But there's another risk-that a manager who is underperforming the benchmark by a large margin will try to liquidate or merge the fund away so as to circumvent the performance-based adjustment. However, I don't see this as a huge risk, either. Fund investors already court the risk of a manager defaulting on a promise to cover/repay fees. For instance, if a manager agrees to waive certain fees per an expense cap, it's not uncommon for the fund to carry a receivable on its books for a time. Should the manager go belly up, those amounts-in theory at least-might not be repaid.

True, we're potentially talking about large dollar amounts when it comes to performance-based adjustments. Further, it's a receivable that could extend over multiple years, depending on the length of the performance measurement period. Moreover, it's difficult for a manager to estimate what it might owe in advance, arguably complicating budgeting and capital planning and, in the process, setting the manager up for a situation in which it's difficult, or impossible, to pay what's owed the fund.

There is no doubt this is a risk, and perhaps steps would have to be taken to ensure that assets are set aside to cover accrued performance-based adjustments payable to the fund. Yet, there are a few mitigating factors to bear in mind. First, each performance-adjustment payable is dependent on when an investor buys into the fund and the performance over that unique time period. Since investors buy into the fund at various times, this creates numerous measurement periods. Second, performance tends to ebb and flow above and below the benchmark's return in the best of circumstances. This is especially true of shorter time periods. Taken together, these factors-numerous distinct measurement periods coupled with inevitable performance swings-should ensure that a manager's total liability isn't too onerous, as there's diversification of performance across many different periods.