News
FOCUS - summer 2007
A Standard for Simulated Returns
Industry comment by Remco van Eeuwijk,
Global Head of Investment Risk & Performance,
ABN AMRO Asset Management
GIPS is great. It provides an excellent minimum standard to ensure fairness and comparability in performance reporting to prospective clients. Most asset managers comply with GIPS and most of us are able to produce excellent GIPS compliant reports, using systems such as PerformaGlobal®.
But what do prospects do? They rarely ask for our GIPS-compliant composite performance reports. It’s not that they are not interested in reviewing the performance of portfolios that we have already managed. Instead, prospects want to get an impression of what the performance for their particular mandate would have looked like. And who can blame them? They don’t want the standard product. Just like when I buy a new car or couch, they want a different shade of blue and they want to see what that shade looks like before buying it.
In the world of portfolio management, these different shades of blue may range from the requirement of a different benchmark, active risk level, or active risk profile (e.g. by hedging out certain exposures). So what do you do in those cases? Of course, you produce the performance information by doing a simulation of what performance would have been, using some combination of the prospect’s mandate specifications, historical market data, and the historical returns of your own portfolios and/or composites. After all, it is a fair question to which the prospect deserves an answer.
In many cases, the simulation is not rocket science. For example, if your outperformance versus one benchmark has been 2%, then why wouldn’t it be 2% versus some other benchmark as well? Well, perhaps. Perhaps applying the same active strategy to a different benchmark means that you would have had short exposures to certain securities. That is no problem if you are allowed to take short positions and if you actually take them if the strategy requires it, but one or both of these may not be appropriate. And if that is the case, then how fair is it to say that you will outperform the prospect’s benchmark by 2%? If difficult questions arise even in a simple example like this, then it is obvious that simulations for different risk levels and profiles can lead to even more difficult questions that may be answered in different ways. All of which may lead to fair presentations, but presentations that are no longer comparable; let alone that we have already determined which disclosures to put around such simulated performance information.
You may expect that GIPS answers these difficult questions so as to ensure that the performance information we present to prospects is fair, comparable, and surrounded by proper disclosures. Wrong. In fact, on this point GIPS buries its head in the sand, stating that simulated returns are not allowed, while at the same time allowing us to present any returns, including simulated returns, using the catch-all cop-out: supplemental information.
Given the frequency with which simulated returns are presented to prospects in our industry and given the scope that exists for less than fair information that is not comparable between asset managers and lacks appropriate disclosures, I believe it is high time for standards on simulating returns. Ideally, these would be developed within the GIPS framework; and we had better do so quickly. As we have seen with UCITS III implementations by European regulators; if we don’t act as an industry, they will. And the outcome, I promise, will be worse; not only for us as performance professionals, but for our employers and our clients as well.