Monday, January 6, 2014

10 Common GIPS non-fatal mistakes

The Global Investment Performance Standards (GIPS(R)) are complex and often confusing. And while we have identified common mistakes which can lead to non-compliance (recall that GIPS is very "black-and-white," with no "materiality" for compliance: you either are or are not), there are also mistakes we often see which, though not fatal (i.e., that won't cause the firm to be non-compliant; sorry for the bit of hyperbole), they are mistakes nonetheless.

1. Using the asset-weighted version of standard deviation for dispersion. The asset-weighted standard deviation was first introduced with the AIMR-PPS(R), and falls under that broad category of things that made sense at the time (like, for example, the brake light that U.S. cars must have in the center of their rear window and the lower urinals in men's bathrooms).

I love it when Carl Bacon, CIPM and I agree on anything, and we both agree that this approach is flawed. Unlike the equal-weighted (i.e., universally standard) method to derive standard deviation, the asset-weighted result is not interpretable.

We always recommend that our verification clients replace the asset-weighted version with equal-weighted. You'll note that GIPS, unlike the AIMR-PPS, does not encourage it, and we reject it. While using it isn't non-compliant, it is not a good approach.

2. Saying too much. By this I mean having disclosures that simply aren't needed. For example, "negative disclosures." For example, while GIPS requires firms to disclose the use, extent, etc. of leverage, derivatives, and shorts, if you don't use them you don't need to say anything. Some firms may have the philosophy that the more you disclose, the less likely the prospect will read, but that's an exception, I believe. Better to drop some of these unnecessary statements and enlarge the font size.

3. Showing both firm assets and composite percent of firm assets. I'll confess that this is a subset of #2, but it warrants its own recognition. GIPS 1999 required firms to disclose composite assets, as well as firm assets and composite percent of firm: the reality is that if you have any two of these, you can get to the third, so why require all three? The folks who worked on GIPS 2005 (I was one, by the way) wisely decided to drop the requirement for all three, and only require composite assets and EITHER firm or percent of firm. Why clutter the data columns; they're crowded enough?! We recommend showing firm assets because if you show percent, while you can get to firm it's only an approximation and can be off by a fair amount.

4. Not saying enough, by not including meaningful supplemental information. First, "supplemental information" is essentially anything related to returns or risk that isn't either required or recommended; for example, attribution results. Firms can take great advantage of supplemental information to go beyond the Standards and provide meaningful information to support the firm's performance. I'll probably do a piece on this later, as there are great opportunities here.

5. Not being verified. Granted, verification remains an option and we have come out strongly opposing it ever being mandated (and we believe this will never happen). That being said, there remain some firms who have chosen not to be verified; but why not? It's an investment, just like compliance. If you're not verified you're now required to state this (in the past, you didn't have to say anything about verification). We believe it's worth the investment.

6. Not being verified frequently enough, by not having it done annually. Okay, so you've decided to be verified but periodically skip a year, what's the harm in that? What this means is that you'll be at least a full year with what might be termed a "stale" verification. You're required to indicate the period for which you've been verified, meaning if you skip a year the reader will know this and may wonder why. We recommend doing annual.

7. Being verified too frequently, by having it done quarterly. Our surveys have consistently shown that most firms get verified annually; we recognize that a few verifiers encourage quarterly, with the argument that more up-to-date verifications are better. Better for whom? We believe the verifiers themselves, as it allows them to keep their folks busy all year. The ironic thing is that we often find firms that use such verifiers being several quarters behind, meaning they're showing, for example, that they were verified through two or three quarters ago: where's the benefit in that? To us, it's a waste of money (since quarterly will usually cost more), more disruptive (since you have to respond to verifier requests four times a year, not just one), and unnecessarily time consuming. But, if you like having quarterly done, fine; we are prepared  to do them for our clients, though none (even those who switched to us from quarterly-promoting-verifiers) do.

8. Getting examinations done, without ensuring the cost is justified. I know I've "harped" on this quite a bit, but this list wouldn't be complete without me including this item. While verifications are optional and recommended, performance examinations are merely optional. We have a few clients that have them done, because they believe they have value; however, most of our clients (something "north of" 95%) do not. If you feel they're justified, great! But, at least think about it. Better, ask your verifier why THEY think it's necessary.

9. Defining composites based on client strategy requests, rather than firm marketed strategies. While it's rare, some firms create new composites whenever a client requests a variation to a standard marketing strategy (e.g., you have an emerging market equity strategy, and a client asks for the strategy ex BRIC (Brazil, Russia, India, China) based companies). While the compliant firm may want to create such a composite for marketing reasons (i.e., if they believe other prospects may be interested in such a variation to their standard strategy), to automatically create composites for every client-defined / constrained strategy will result in way too many composites than the firm will likely need. Instead, defining strategies and then establishing a policy for discretion is, in my view, a better approach.

10. Failing to properly format the firm's policies & procedures document. We recently issued a white paper on the subject of policies and procedures; this document falls on the heels of one that was released by the GIPS Executive Committee; we believe both should be referenced to ensure the firm's documents are appropriate. Organizing the document into a logical collective of policies should make its implementation and use much easier. A very simple addition is to add page numbers. Any document greater than two pages should be numbered; actually, even numbering a two page document is perfectly fine (just don't number the first page). When we do verifications and get P&P that aren't numbered, it makes referencing our comments a bit more challenging. But meeting the needs of the verifier is hardly the reason to number pages: numbering should be a standard practice because it makes communication and referencing much simpler.

11. (Bonus!) Not being selective about which recommendations to employ. GIPS recommendations are boldly referred to as "best practice," something we often take exception to, as there are some recommendations we strongly oppose (the most noteworthy is that compliant firms should provide copies of composite presentations to clients on an annual basis, something I've commented on before). There are, however, many recommendations that are excellent, such as showing the equal-weighted composite return.

I happen to believe this is a far superior metric to the required asset-weighted composite return, as it provides an non-skewed report of averages. Too often, very large accounts pull the average in their direction, thus providing something other than the experience of the average investor. Granted, the required disclosures are extensive, so we understand why adding recommended ones may seem to overdue it, but many are worth incorporating.

A word of clarification. Most men don't know that the lower urinals that are typically found in men's bathrooms are not there for children, but rather handicapped: a case of "it seemed like a good idea at the time." The center brake lights we see in the USA are there so that you can see if the car in front of the car in front of you is braking, so that you have early warning. Just try to see if you can actually see these lights. Again, it seemed like a good idea.

No comments:

Post a Comment

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