Monday, November 5, 2012

When short is too short

I have a confession to make: I used to be 6' 1/2" tall (I liked the 1/2", because it put me ABOVE 6'). My father was 6'6", my older brother roughly the same, and my younger brother is, I think, around 6'2". My two sons are both taller than I am. And so, I can be thought of as the (a) runt of the litter and (b) still the shortest guy in the family. (Okay, it's true that I am taller than my grandsons, but they aren't fully grown, yet).

Sadly, I am no longer 6' 1/2" tall. Gravity has set in (so they tell me) and I am now 5' 11"; i.e., LESS THAN 6'. Yes, I've needed therapy to deal with this, but I'm managing quite well, thank you.

And yes, I know that I am still taller than average, so have no complaints.

But in this post, I'm speaking about a different kind of short. I'm speaking of shortness in time.

John Simpson, CIPM recently conducted a webinar on the mathematics of multicurrency: a session that was very well attended. One participant raised a question regarding the possibility of instantaneous measures of change. If you have a mathematics or physics background, you can probably relate to this. I recall in calculus, where we're introduced to the idea of shortening the measurement period, so as to find the instantaneous changes.

The question: should we concern ourselves with instantaneous changes in returns, or perhaps more correctly, valuations?

My reaction is that this is of little value, and I would discourage doing or pursuing it. I have heard some folks ask for intraday returns (which isn't quite as short as instantaneous, but it's moving in that direction), and am concerned that this could be something we'd adopt.

Investing is supposed to be long term, is it not? Looking at short periods introduces noise and causes frustration.

This past summer we replaced the robot (vacuum) for our swimming pool. It was very different from the ones we previously had. Someone from the company urged me NOT to watch it as it worked, as it would cause frustration, as I'd see it miss things here or there, or fail to see it move into certain parts of the pool, and might wonder if it's working properly. Instead, I was advised to step away and look back after it's been given several hours to work.

There are, I believe, only four reasons to calculate performance within a month (i.e., to calculate daily performance; more correctly, value the portfolio so that we have a point in time to measure from):
  1. Because someone has asked "what's my month-to-date performance" or "MTD returns are included on a report (portfolio managers often want to see MTD results, and this is fine)
  2. To improve accuracy, by revaluing the portfolio for large or, better yet, all cash flows
  3. Because an account terminated, and we wish to calculate the return up to the point where management ceased
  4. It's a new account, and we want to start the valuation on the date it began being managed.
Disagree? Have other thoughts? Please chime in!

I should add: there is nothing to be gained from an accuracy perspective to move to instantaneous or daily valuations/calculations; cash flows are the primary delimiter for valuations (and period ends, of course). In addition, daily valuation is fine and becoming standard, and I support this.

No comments:

Post a Comment

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