Composite examinations are like audits: they are done to validate the accuracy of a given compliant presentation. The firm can opt to have none done, have them done for selected composites, or have them done for all composites. As you will see below, we generally recommend not having them done unless the firm has a good marketing reason to do so, and then only for those composites that they feel will benefit from having them done.

During GIPS verifications, the verifier checks that the firm’s policies are designed to present performance in compliance with the GIPS standards, which includes checking that calculation methodologies are appropriate and applied consistently. This is done using sampling, but sampling may not cover everycalculation, statistic or disclosure for everycompliant presentation. Composite examinations provide an extra level of assurance for a given compliant presentation, and all of the information in it. Thus, deeper sampling specific to the composite is done for an examination. An emphasis is also placed on tying the underlying accounting and valuation information to external records (i.e., those of an independent third party, such as a custodian).

This means that while the return formula may be right, if the input data is wrong, or has been manipulated, or if the calculations weren’t even performed but rather numbers were simply inserted into the report to make the firm look good, it is possible it won’t be discovered.

This is why the claim of compliance for firms that do not undergo composite examinations must include the following statement: “Verification does not ensure the accuracy of any specific composite presentation.”

In general, we discourage clients from having composite examinations performed.

First, unlike verification, the Standards do not recommend having them done.

Second, while prospects seeking a manager will often ask whether the firm claims compliance with the Standards and whether they’ve undergone verification, it is rare that they ask if composites have been examined.

And so, why spend the money?

We believe firms comply for marketing reasons: that in order to attract new business, compliance is often a de facto requirement. Likewise, verifications are done for the same reason. Granted there are other reasons to comply and be verified, but these are the chief
ones. And so, the money spent to comply and be verified are investments. However because composite examinations are neither Page 6 of 15 recommended nor often asked about, the money spent to have them done are costs.

Yes, there are exceptions. Often, asset owners who comply want them done, because they want to be able to state that an independent party affirmed the accuracy of their returns. And, there are times when asset managers want them done for marketing purposes, which is also fine.

Our research has shown that in the United States, 60-70% of managers have composite examinations done. Contrast this to TSG’s clients: less than 10% have them done.

Why the disparity?

Recall that the Big-8 did not do “Level I” verifications under the AIMR-PPS. What they would do is the Level II: i.e., the equivalent of composite examinations. As a result, those managers who wanted to be verified by a large CPA firm would have to accept composite examinations.

Once GIPS was introduced, with the removal of the declaration that the verifier had found that the firm was, in fact, compliant, these accounting firms then began to offer to do verifications. They saw this as “good news” for their clients. The good news for the verifiers was that these clients, perhaps at the urging of the verifier, continued to have composite examinations done, too!

We have taken over verifications from dozens of firms who previously had composite examinations done: in the majority of cases, we were able to persuade the firm to stop them, or at least to reduce the number they were having done.

This may seem like an odd thing to do, since we charge for composite examinations, so why not continue to?

The simple reason is that we don’t believe they have value, in most cases. Therefore, why encourage a client to spend the money when they don’t have to, even though the result is less revenue for ourselves. We believe this is the right, ethical thing to do.4

An interesting point to make: elsewhere in the world, composite examinations are rarely done. These managers were not claiming compliance with the AIMR-PPS, so did not have the history of having Level II verifications done. We believe this fact adds credibility to our belief that them being done today stems from the Level II verifications under the old standards.

The main objective with an examination is to validate the returns and other statistics that are shown on composite presentations.

And so, verifiers will ask to see both the data that they used to derive the reported returns, as well as the data from the “official” sources (e.g., custodians).

The verifier will validate that the data that was used is correct (that it matches5 what’s on the official records). They will also ensure that the calculations were done properly (e.g., for large cash flows, that the portfolio was revalued).

As with verifications, the verifier may do sampling. They may, for example, select only certain months to test, as well as a random sample of the accounts.

4    We took over the verification for one New York City-based client about ten years ago, who had
previously had examinations done. For years, they were done for all of their composites, even those they
didn’t market. In our view, it was unethical for the verifier to do this, knowing that these examinations had
little value. When we took over, one of the first questions we asked was “why do you have composite
examinations done?” They seemed surprised by the question, but acknowledged they didn’t have a reason,
other than they had always had them done. We suggested they stop, and they did. They have never had them
done again. Granted, that’s a loss of quite a lot of revenue, but it was the right thing for them to do. And, for
us to do, too.

5 There are times when the data won’t match 100 percent. For example, there may be differences in
prices for non-market priced securities (e.g., bonds), the custodian may not accrue for interest, the
custodian’s records may be settlement date, not trade date. And so, some adjustments or review may be
necessary, when differences arise. In other words, differences may not mean errors..