Tuesday, October 25, 2011

A new GIPS rule has been introduced

The GIPS(R) (Global Investment Performance Standards) Q&A desk appears to have introduced a new rule:

Firms may not add new accounts within a month*

While I have no problem with this change, it would have been nice if it had been put forward for public comment, since it doesn't appear to be based on any language that is actually in the standards themselves; i.e., no paragraph is referenced to justify its introduction. This is arguably a change in the standards, which should (I believe) have been open to public comment. [Recall that GIPS-compliant firms must comply with Q&As and Guidance Statements, in addition to the standards themselves. Therefore, this Q&A applies to you.]

In this Q&A post we find the following: “When calculating composite returns for a specific period, only portfolios that are included in the composite for the entire performance measurement period are included in the calculation.” [emphasis added] We also find “When calculating a monthly composite return and the performance measurement period is defined as a month, a firm must not include in the composite calculation portfolios that were not managed for the full month and therefore do not have a full month of performance.” [again, emphasis added]

Consider the expression "performance measurement period." In the early days of the AIMR-PPS(R), we saw both this term and "performance reporting period" used.  One might report quarterly but measure monthly.

While this new "rule" seems to prohibit monthly additions (or removals) of accounts to (from) a composite, it appears to be based solely on this concept of this undefined term: "performance measurement period." If I calculate the return for January, but revalue the portfolio for large cash flows, aren't I measuring performance across multiple performance measurement periods in order to arrive at my January return? Firms that do daily performance might say that their performance measurement periods are days, might they not? Would this therefore allow them the opportunity to include an account within a month (since the month isn't the performance measurement period; the subperiods of the month are)?

Let's return to the first sentence from this rule: "When calculating composite returns for a specific period, only portfolios that are included in the composite for the entire performance measurement period are included in the calculation.” I've now emphasized two terms: specific period and performance measurement period. Are they interchangeable? It's a bit confusing, I think.

If the intent of this rule was simply to say "you can't add or remove accounts within a month," why doesn't it say that? I firmly believe that this is what is intended, but again, from the standards perspective, I fail to see the justification for it; sorry.

I will confess that I was aware that this change was "coming down," and I am perhaps responsible for it, given my challenge of the aggregate method (many, though not all, of my examples of flaws were based on cases where accounts were added within the month). A colleague told me that firms cannot add accounts within the month. It probably won't surprise you to learn that I asked "what paragraph in the standards is this interpretation based on?," but I got no response. I think this Q&A is the response.

Again, I am fine with this change, but feel that it is based on nothing in the standards. There is a process and protocol to introduce changes to the standards; this wasn't followed here (unless someone can point out the chapter and verse that the Q&A's response is based on). The Q&A desk's role is to interpret questions relative to existing rules; no rule was cited as the basis for their response. In my view this constitutes a new rule, not an interpretation of an existing rule.

In addition, given that there ARE managers who (a) claim compliance with the standards and (b) HAVE been doing this in some cases for many years, does this mean that they aren't compliant, or that they have to recalculate their performance? More is needed, I believe. In my view, this is a new rule that should have an effective date, so as to avoid confusion (and again, should have been open for public review). There are composite software packages that permit their customers to add or remove accounts from composites within a month; I suspect that they should consider restricting it going forward, so as to avoid putting their customers at risk in violating the standards. This issue will not apply to the vast majority of compliant firms, who add accounts at month-end; however, it's still important to consider the process and how the change was introduced. Can we expect other changes to be introduced in this manner?

* Note: I don't want to confuse anyone; this isn't a quote from the Q&A; it's my interpretation of what the new rule is.


  1. My suggestion is that the GIPS EC have an effective date associated with this Q&A. I would actually suggest that it become effective January 1, 2013 at this point, because we are so late in 2011 that I have to wonder how feasible it would be for firms to make a change, if necessary.

    This Q&A potentially draws a line in the sand (between the compliant and those who only thought they were) where there is a lack of provisions already in effect to back it up. While I am fine with the interpretation, I am unaware of any existing provisions or guidance to support the new interpretation. (Perhaps I'm wrong - if there is a provision or guidance to support the Q&A hopefully someone will point that out).

  2. Thanks, John. You're right that 1/1/13 makes sense for an effective date. Perhaps you can raise this as an issue at the conference this week.

  3. Dave, I'm glad you posted this. I read the Q&A last week as well and was not clear on what can and cannot be done. I would think that if the aggregate method of calculating performance is allowed, the addition (or subtraction) of a portfolio would simply be treated as a cash flow and would follow the rules for calculating portfolios (either monthly instituting a large cash flow policy, or daily). In either case, it should result in a time weighted return that would reflect the strategy.
    I hope there will be more clarity on this subject and would be interested to see if this is a prospective change or if some firms have to restate their numbers.

  4. Jed, it is somewhat ironic, is it not? I criticize the aggregate method and the only place that intra-month flows really work IS with this approach; and can be justified as being permitted. And yet it's specifically addressed and not allowed! I can see banning it for the asset-weighted approaches, and have commented on this in this blog. Perhaps you can seek clarity at the GIPS conference.


Note: Only a member of this blog may post a comment.