Thursday, May 30, 2013

Performance Holidays

A subject that has come up occasionally, but that has gotten little attention (and none formal), is the idea of "performance holidays."

These arise at times when a client makes a significant contribution to their portfolio, knowing that because of the nature of the strategy, asset class or market, it may take time to get the money invested, and the portfolio back to the point that it's representative of the strategy. Even though time-weighted returns eliminate (or reduce) the impact of cash flows, the cash that's sitting around waiting to be invested can still influence the portfolio's return. The "Performance Holiday" is a way to eliminate this problem.

Questions arise as to what the firm should do, and even perhaps, what is "best practice" to handle them?

First, I think it's important that the firm have a policy in place as to how they wish to handle performance holidays. Second, it's important (make that, necessary) that the client agree with the process. Third, make sure you document whenever performance holidays occur (when, what was done, who authorized it, etc.). And fourth, as with many things we do, consistency is important, so as to avoid the appearance of gaming situations to their advantage.

Performance Holidays & the Global Investment Performance Standards (GIPS(R))

Whatever holiday arrangement a manager has with his/her client has no effect on what they do with the portfolio's respective GIPS composite. If the firm has a significant cash flow policy and the portfolio's flow applies, it gets pulled for the predefined length of time. If not, it stays.


As for the portfolio itself, there are a few options firms can consider.
The benchmark as a surrogate
First, use a benchmark return as the surrogate for the period. If the firm and  its client agree on such a benchmark return, that's fine; it's up to the firm and client (but again, this return won't be what the firm uses for the composite; the firm must use the portfolio's return).
Creating a "gap" (break) in performance
Second, the holiday may create a "break" or "gap" in the performance track record. This means you'd have a return UP TO the point of the flow, a break, and then a return AFTER the money has been invested. You could not link across this gap.
Temporary Account
Third, the firm could use a "temporary account" for the flow: that is, you'd move the cash into this temporary account and move the investments across as the money is invested (securities purchased).
This is probably the ideal approach, though it means there isn't a holiday, other than for the cash, which is sitting in the temporary account. Temporary accounts can be challenging to implement, so it's important that the accounting folks are "on board with this, too.

Use a return of zero
Fourth, you could set the account's return to zero percent. The problem with this approach is that zero is a return value. There are times when it would be higher that the manager would have obtained had there not been a contribution (in which case the firm is unfairly rewarded) or lower than what would have been experienced (in which case the firm is unfairly penalized).
Therefore, I'd  recommend against this option. It would be better to use the benchmark's return.
Bridge the gap!
Fifth, you could have a gap in performance, but agree to link across it (i.e., the gap disappears!). The problem here is that's equivalent to setting the portfolio's return to zero, so I'd argue against it.

Best practice?
I see three options: using an agreed upon benchmark, having a break, or using a temporary account. I'm sure there are other approaches, but you and your client need to be comfortable with whatever you choose.

Some “best practices” are probably in order. I've already identified a few, but further discussion and review is in order. To summarize:
1) Establish a written policy
2) Get the client's approval as to how their holiday(s) will work
3) Document performance holidays
4) Be consistent in implementing performance holidays
5) Ideally use a temporary account for the cash flow. If this can't be done, the second best approach is probably the use of the benchmark's return during the period.
Where to next?
First, if you have suggestions, ideas, insights you want to share, please send me a note.
Second, I may expand upon this further in our newsletter.
Third, we may get input from the Performance Measurement Forum on this topic; if we do, I'll pass it along.



    Some of the "confusion" in this area comes from working with an unclear question, so I suggest that we first clarify what we are measuring: the return of the manager's fund or the return of the client's holding in the manager's fund. Two different things and two different calculations, and also two different people using the information - the fund manager and the client. If the client wants to know the manager's return relative to the benchmark for some sort of market-relative evaluation, then the client can simply use the composite's time weighted return (and if it's a mutual fund or other commingled vehicle, then it's even easier: use the fund return.) If the client want's to know the return of his investment in the fund then the client must use the money-weighted return. The difference between the TWR and the MWR is the result of cash flow activity. Good insights.

    This is pretty obvious stuff. The "confusion" or "controversy" comes from trying to see the world through a GIPS lens, when in reality GIPS is simply a specific application that supports marketing of individual investment products. It has little to do with evaluating the results of investments on an ongoing basis, and it has nothing to do with evaluating either an actual investment in an individual product or that product within an overall portfolio invested across many asset classes.

    That's the difference between marketing and the real world of investments.

  2. Steve, great point about using the composite's return as the proxy during the holiday!


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