The subject of net-of-fee returns, as they pertain to the Global Investment Performance Standards (GIPS(R)) came up twice this week already, so I’ll share with you the questions and my responses.

First, I was asked about firms using “model fees” to derive their net-of-fee returns. Is this permitted? I’m unclear as to what a “model” fee is, but would suggest that it probably is NOT permitted. Rather, firms can use the highest management fee “incurred by portfolios in the composite” as opposed to the highest stated fee. It may be easier to use the highest stated fee, rather than try to figure out what the highest fee was. Or, firms may use actual fees, which leads us to the second question.

“Thanks for continuing to keep in touch with us. We appreciate the work that David did for us in the end of 2008, it has helped us tremendously over the past year. We have created our composites and wondered if you or David could give us some guidance. We have calculated both Gross of fee performance and net of fee performance of the composites. We have a dilemma in that the management fees applied to the composite are different for each account. We would like to report ‘Net of Fee’ performance, but feel it’s not entirely accurate to show a ‘net of fee’ composite that has a blended fee lower than the actual fee we will be charging the vast majority of our clients (1.50%). Is it standard practice to accrue a standard fee (i.e. 1.50%) on top of a gross of fee composite calculation on a monthly basis? We want to make sure we are reporting this in accordance with best practices. Can you offer any suggestions in this matter?”

There are actually a few questions here, but I’ll summarize. First, I am not a fan of net-of-fee returns to begin with. And, I’m not alone. One only needs to turn to their trustee GIPS Handbook to discover the following, from the Interpretive Guidance on Fees Provisions points out, “the most universal point of comparison is the Gross-Of-Fees Return less the Investment Management Fee that the prospective client expects to pay.” If one shows gross-of-fee returns and tells the prospect the fee they would pay, all they need do is back out the fee to approximate what the net-of-fee equivalent is. A return derived from a blend of fees has little value. As I recall, the AIMR(R) standards once required firms to show an asset-weighted fee when they reported a net-of-fee return, so that they could better gauge what the return meant. Sadly, this requirement was dropped long ago and never made it into GIPS. As for accruing fees, this is recommended, though it can be a challenge for many. My recommendation: only show net-of-fee if you have to as gross-of-fee is better. If you have to show net-of-fee returns, then I recommend using the highest actual fee rather then to use a blend. I would ALWAYS show gross-of-fee, even if you show net.

AMENDMENT!

I had a discussion today (1/27/10) with a colleague who mentioned a Q&A on the GIPS website:

We want to report the net-of-fees performance numbers of a composite. Should we state the actual net-of-fees performance returns for all portfolios which would reflect the deduction of the actual fee paid by each portfolio, or can we use a model fee?

The net-of-fees return is defined as the gross-of-fees return reduced by the investment management fee incurred. Firms are permitted to use either the actual investment management fee incurred by each portfolio in the composite, or the highest investment management fee incurred by portfolios in the composite to reduce the gross-of-fees return to calculate the net-of-fees return.

Funny…the writer uses the term “model fee” but it isn’t addressed in the response. And so what IS a model fee? The GIPS Handbook offers no answer. My colleague said it’s anything that isn’t an actual fee. I would suggest that the ONLY other fee you can use is either the highest used or the highest on your fee schedule, which are both a heck of a lot better then a blended fee. Amen!