Net-of-fee returns: what to do with the denominator

A colleague recently brought to my attention wording that appears in the 1993 edition of the Performance Presentation Standards, published by AIMR (Association for Investment Management and Research; the former name of the CFA Institute). On page 25, under a section titled "Net-of-Fee Calculation" we find: "In a net-of-fee calculation, when fees are paid from the corpus of the fund, the payments should be included as a withdrawal of capital in F (flows) and in FW (weighted flows). In addition, performance results are reduced by deducting fees as negative income [a positive number] in the numerator." The accompanying formula (that appears on 26) has the fees removed, separate from their treatment as a flow.

What this essentially means is that the fees cancel out in the numerator (which is the same as my recommendation to ignore them). The AIMR-PPS's denominator has them as a weighted flow; I recommend not doing this. Their result is a higher NOF return (since the denominator is reduced by the weighted flow). I believe ignoring the fees entirely is correct.

As I pointed out in an article for the CFA Institute, as well as in our firm's newsletter and this blog, we should completely ignore net-of-fee payments that come from the corpus of the account; we treat them as flows for gross-of-fee returns.

Note: this is MY (i.e., Dave Spaulding's) view on this matter, but I believe that logic and the results show that it makes sense. But chime in with your thoughts, by inserting a comment below! In reality, whether you treat them as a weighted flow or not, the difference is probably de minimis.

Who to get net, who to get gross?

Okay, so yesterday we touched on the calculations for gross- and net-of-fee returns. But who should get what?

Prospective clients should get gross-of-fee returns, unless the net-of-fee returns are net of the same fee. The problem with most net-of-fee returns is that they're net of a mix of fees: how can one easily understand what the number represents? If you're going to show net-of-fee, provide helpful information so that the reader can better interpret it. Under the "old" AIMR-PPS(R), firms were required to show their weighted average fee, which could be helpful to interpret the return; but this is neither required nor recommended in GIPS(R) (Global Investment Performance Standards). In some cases, regulators (think SEC) require net-of-fee under certain circumstances, so it becomes a requirement. To me, gross-of-fee should be a requirement for marketing to prospects. Showing both is probably the ideal, even given the NOF's shortcomings.

Existing clients should get net-of-fee reporting. This represents the manager's performance after the fee they're charging has been removed. This seems best for the client.

Arguably, both prospective and existing clients should also see net-of-taxes (i.e., after-tax) returns, too. And, net-of-risk (i.e. risk-adjusted) returns.

p.s., The 2010 edition of GIPS now requires firms to indicate whether their NOF returns are net of actual or "model" fees (see ¶ I.4.A.6.b).

Deriving net rates of return

There has been a bit of a debate on one of the Linkedin groups lately, regarding the proper way to derive net-of-fee returns, when the fees come out of the corpus of the account. I happen to like this topic, and wrote an article on it for the CFA Institute some time ago.

To put it simply, fees are technically cash flows, but when it comes to calculating net- or gross-of-fee returns, we need rules on what to do with them. When we calculate gross-of-fee returns, we treat the fees as a flow; when we calculate net-of-fee return, they're ignored (i.e., they do not show up in our calculation at all).

The Linkedin question was simply whether one should always include the fee (as a cash flow) in the denominator, regardless of whether you're calculating a net or gross return. One individual suggested that since the manager controls the timing of the flows (recall that in the Modified Dietz formula, cash flows are weighted based on timing, and this weighting is done in the denominator), that they should always be included (sorry about the run-on sentence...quite a mouthful!). I failed to see the logic behind "controlling the timing" and whether to include them. My interpretation of the statement was that this "control" might be used to advantage them somehow, even though one never knows how the month will end, so why would this matter?

Anyway, I stick by my simple rule:

• Gross-of-fee: treat fees as a cash flow (in the denominator AND numerator)
• Net-of-fee: ignore everywhere.

This, of course, is for fees taken from the corpus; fees paid from outside the account are treated differently. We'll address this at another time.