A Free GIPS Webinar!

I will host a webinar on Monday, September 17, at 11:00 AM EST, titled "The Ten Most Common Mistakes GIPS Compliant Firms Make." This is a FREE webinar, but space is limited, so please contact us soon (either by calling 732-873-5700 or emailing Patrick Fowler). GIPS(R) (Global Investment Performance Standards) has become a regular part of our lives, and we thought this would be a good session; we hope you agree!

The Spaulding Group hosts monthly webinars, and they are typically free only to our verification clients and members of the Performance Measurement Forum. Others are charged a modest fee. But this webinar will be free to all, but you must register quickly, as there are a limited number of slots.

In addition to covering these 10 most common mistakes, we will open it up to your questions, which you can submit in advance, or during the webinar. The presentation should last about an hour, and we'll have an additional hour for questions.

Many of our clients make these monthly events "lunch-and-learn" sessions for themselves, their staff, and their colleagues. You can have as many folks on the call with you as you'd like.

If you're attending the GIPS Annual Conference in Boston, please stop by The Spaulding Group's booth to learn about our services and for a chance to win an Apple iPad!

Beware of outliers: lessons from education, sports, and movies

The headline in today's Home News Tribune (a central New Jersey newspaper) is titled "Probe: Teachers enabled cheating." It points out how last year Woodbridge administrators and teachers were celebrating because New Jersey state's School Report Card reported impressive results for several township elementary schools in 2010. For example, Avenel Street Elementary, where 94 % of its third-graders scored "high marks." This was quite a feat, since only 37 % of students statewide earned such scores. As it turned out, this was a case of teachers cheating.

Last week we learned that famed bicyclist Lance Armstrong was being stripped of his medals because of cheating.

I am often reminded of Charlie Sheen's portrayal of a broker in the movie, Wall Street, who suddenly began racking up extraordinary impressive numbers. His success was largely attributed to insider trading.

These are all examples of outliers. And while we can expect to see outliers, there are times when they are just a bit too exceptional. To have a school with 94% of its students achieve "high marks" when the norm is only 37% is just a bit too good; to win the Tour de France a string of seven times is just too good without the help of something illicit.

Americans will recall a few years back when in any given baseball season we'd find several batters hitting more than 50 or 60 home runs, something that used to be quite rare had become the norm. Some observers suggested that the balls were "juiced"; it turned out that the players were. Now that steroids are the exception in baseball, the numbers have dropped to the level they had historically been.

While not all distributions are normal, we still are sensitive (or should be) to events that are just too good. In one of his books (I can't recall offhand which), Harvey Mackay speaks of things just being a bit too good; and when this happens, be careful.

Too often when exceptional events occur, we are slow to think that something improper has occurred; perhaps we are caught up in the moment, or really want to believe that someone can be that good. Who wants to be accused of being a "doubting Thomas" or a "party pooper," raining on someone else's parade?

A candidate for a "poster child" for exceptional behavior in investing would be Bernie Madoff:  his string of above average returns should have caused concerns, from his clients and the regulators, but because everyone liked Bernie, hardly anyone questioned such performance.

In the late 1960s, my older brother, who died more than 20 years ago, came home from Army basic training, and showed me a photo of the woman he had just married (a complete surprise to all of us, since we didn't even know he had a girlfriend). I knew there was a problem the moment I saw her picture: she was just too good looking for my brother. Sounds like a nasty remark for a brother to make, right? You have to understand that Bill was a bit awkard and clumsy, and wasn't particuarly good looking himself. He had a difficult time attracting girls when we were growing up. And so, for him to find someone so attractive just didn't seem right. As it turned out, this was an older woman who had been married before, to soldiers who had gone off to Vietnam (where my brother was expected to go, but didn't), never to return, and so she'd collect the life insurance money. A sad story, yes? I won't bother to share any further details about what transpired, but only to say that this was an outlier and one that I, at a fairly early age, was able to detect.

In our GIPS(R) (Global Investment Performance Standards) verifications, we look for outliers, as these are often cases of errors that crept in. And while outliers are often and perhaps usually fine, there are times when they aren't. Consequently, we should be sensitive when they occur, just in case ...

Money- versus time-weighitng: let me be very clear

It's election time here in the United Staes, so it's appropriate to cite a few of our presidents. America's 37th president (Richard Nixon) is often recalled for his "let me make this perfectly clear" line. Our 42nd president (Bill Clinton) didn't use the term "clear," though it was implied when he denied "having sexual relations with that woman." And our current president (Barack Obama) frequently uses the term "make it clear" in his speeches. Sometimes, when trying to make something "clear," we end up making it muddier; obfuscating rather than elucidating, so to speak.

Over the recent past I have attempted to make the differences between time- and money-weighting clear, and to some degree have met with success, but not to the extent I would like.

I recently received the following note from a long time friend and colleague:
 
"I am working with some financial institutions and they use XIRR to measure the return of even stock and mutual funds. I am trying to convince them to adopt GIPS standard and use Modified Dietz and Time Weighted Rate of Return to calculate rates of return. Since I have never used XIRR, I wanted to see if there was anything you might have or be able to share to help me convince this client - ideally a white paper comparing the two methods and which one is most appropriate."

First, the XIRR is a form of the internal rate of return, and therefore it's a money-weighted method. Second, and more important, we have to understand the reason behind their use of performance / returns. Is it (a) to tell clients how THEY did, or (b) tell them how their MANAGERS did? This is critical in order to decide the way to proceed.

If the answer is "to tell our clients how they performed," XIRR is perfectly fine.

If, however, it's "to tell our clients how those who manage their money have done," then TWRR is probably the right formula, unless their managers control the cash flows, in which case XIRR is the way to proceed.

More and more folks are seeing the wisdom behind this. And while I may sound like a broken record, I (along with several of my colleagues who feel the same way) believe it's important to keep pressing this point. Disagree or, have other thoughts? Please chime  in!

Just because the return or risk formula seems to make sense doesn't mean it's valid

The Spaulding Group offers Operational Reviews and Software Certifications, both of which expose us to formulas that firms put into use. Sometimes, they are variations of formulas that have been around for years; but occasionally, they're brand new; ones the clients developed themselves. Over the years we have encountered a variety of "home grown" methods, which are usually invalid.

An Abbott & Costello routine serves as a great example of just because you can make it look like it makes sense, doesn't mean it does.

The scene: they're in the Navy, and Costello is a baker; he explains that he made 28 donuts for the officers; there are seven officers and he has just enough so that each will get 13 donuts (a "baker's dozen").

You read it right: 28 donuts will be enough so that each of the seven officers will get 13 donuts. How can this be? Well let's see (and I'll do my best to explain, though the video is better):

1) Division: 28 ÷ 7 = 13. How? Seven cannot go into two, so we put that aside. Seven goes into eight once right? And so we divide the seven into eight and have one left over. We now bring over the two, giving us 21. Seven goes into 21 three times, so our answer is 13.

2) Multiplication: 7 x 13 = 28. How? Multiply seven times 3 and we get 21. Seven times one is seven. Add seven to 21 and we get 28!

 
3) Addition: 13 + 13 + 13 + 13 + 13 + 13 + 13 = 28. Begin by adding the 3s: 3, 6, 9, 12, 15, 18, 21; and then add the ones (22, 23, 24, 25, 26, 27, 28).

Clearly, this creative arithmetic is wrong; however, there is no doubt that many who, when presented with it, would be inclined to think that it makes sense and that Costello must be correct.

If you think this is silly, you should see some of the methods that have been given to us to measure returns!

Hopefully you'll agree that this was a "fun" way to end the week. Care to see "the boys" in action? You can, on YouTube! While perhaps not as famous as their "who's on first" routine, it's still enjoyable.

p.s., Ma & Pa Kettle do a similar trick showing how five times 14 equals 28! This adds even more credibility to this creative math.

"Technology Training Can Be A Boon To Productivity"



This post's headline comes from a book by the late Chet Holmes. It's doubtful, very doubtful, that anyone would disagree with such a statement. Our firm provides several training courses, but nothing that deals specifically with technology ... well, this isn't completely true.

We did host a webinar on "Excel Tips & Tricks," which also made its way into both The Spaulding Group's Performance Measurement Forum and Performance Measurement, Attribution & Risk (PMAR) conference. No one walked away from these sessions saying "I knew all that" or "how can I benefit from this?"

And now we're hosting a webinar on the web, titled "Internet Tips & Tricks" (see a pattern?). The host is a professor at Pace University: Dr. Vasant Bhat. I had Professor Bhat for a doctoral course, and found him to be quite gifted when it came to navigating the web and using tools that most folks are unaware of. The webinar will be this coming Monday, August 27, 2012 at 11:00 am (EST).

I invited him to host this session a few months back, and he was kind enough to agree. Our verification clients and forum members can participate at no cost; for all others, a nominal fee is charged. If you want to learn more, please contact Patrick Fowler or just call our office (732-873-5700).

This program is a great way to train your team; the timing might work as a "lunch & learn" session. If you sign up and decide it wasn't beneficial, we'll refund your investment. We have a limited number of slots, so register soon!

Smoothing and geometric attribution

Carl Bacon, CIPM, when asked to contrast geometric and arithmetic attribution, will no doubt point out the chief advantages geometric offers:

• Proportional: the active return is a ratio, not a difference, as we find with arithmetic.
• Convertible: the active return is independent of the base currency; the geometric active return will be the same whether it's expressed in dollars, euros, pounds, yen, etc.
• Compoundable: the geometric active return multiplies across time; no "smoothing factor" is required, as it is with arithmetic.

Carl is, of course, correct in all of these points; however, he fails to explain that while a smoothing factor isn't needed "across time," as we "link" our single period effects, one is needed "within time." That is, we need a smoothing factor for every single time period.

In his '05 FAJ article* on geometric attribution, Jose Menchero, PhD, CFA discusses the need for a smoothing factor. He points out how one can arrive at a "pure" geometric equivalent of an arithmetic model, but that this will not result in an approach that will fully reconcile the effects to the excess return: a smoothing factor of some sort is necessary.

Jose offers a rather healthy formula to "smooth out" the residual. Contrast this with Carl, who assigns the residual entirely to the selection effect. And so while Jose ensures that the residual is assigned across all effects in some appropriate or proportionate fashion, Carl is content with it residing entirely with selection. It is not my purpose here to justify one over the other: Carl's is clearly much simpler to work with than Jose's, and perhaps the differences are immaterial.

My point is merely to explain, in as clear a fashion as possible, that just like arithmetic, smoothing is needed. It's just that arithmetic attribution has no residual for the single period, but will across periods; while geometric has a residual for single periods, but once they're smoothed out, won't have one across time. It's the attribution equivalent of you can pay me now, or you can pay me later, but surely you will pay (i.e., you must smooth out a residual at some point).

Space does not permit the elaboration necessary to give this topic the attention it needs, so please consider this merely an introduction to what will follow in this month's newsletter.

*Menchero, Jose, 2005, Optimized Geometric Attribution, Financial Analysts Journal 61.

Personal rates of return ... what are they, really?

I am home this week, working on my doctoral dissertation proposal, and need references to cite for a "personal rate of return."

And so, like any good researcher, I began with a "Google search," and found the "Finance guy's" blog, which has a post on this subject.  He references an earlier one, where he advocates using the Original Dietz, across the full year, rather than the XIRR. While it's true that it can provide roughly the same result, the industry has pretty much abandoned this mid-point method. Okay, if your readers are truly unsophisticated investors, with limited math skills, perhaps this is okay, but I would couch such a formula with an explanation that its accuracy is not very good.

In the more recent post he has the following:

The personal rate of return you get from a financial service provider like Fidelity or Schwab is usually a Time Weighted Rate of Return. If you want a Dollar Weighted Rate of Return, you will have to do it yourself.

A "time-weighted rate of return" as a "personal" rate of return? What is personal about a time-weighted return? Ten people are invested in the same fund, but contribute different amounts during the year. What will their "personal rate of return" be under time-weighting? The same as the fund's performance, because by definition, we've eliminated the impact of the cash flows.

He cites another website, dailyVest, whose post on this subject states

Time-weighted rates of return can be calculated on a daily basis (one method known as Daily Valuation) or on a slightly less accurate monthly basis (known as Modified Dietz) where inflows/outflows are averaged for the month. This time-weighted methodology used for calculation of personal rate of return provides a truer measurement of how investments are performing.

No, no, no! A "personal rate of return" has to be "personal," does it not? And how do we get this? By taking the flows into consideration.

Perhaps we have a bit of Clinton-like speech here (as in, "it depends on what you mean by the word, is"), at least in dailyVest's case, because one might ask what "This" means, in "This time-weighted methodology." If they're referring to Modified Dietz, then they get partial credit. However, by linking the intraperiod Modified Dietz-derived returns, we achieve an approximation to a time-weighted return, which is not a money-weighted method.

The site "All Financial Matters" gets it right, because this author cites the XIRR as the formula to use. In his post he provides an example of an investor putting money into the Vanguard S&P 500 Index Fund (VFNIX), and states:

Some simple math will tell us that VFNIX returned 6.8% (not including dividends) from 1/31 – 12/31 ((111.64 – 104.54) ÷ 104.54). However, the real question is: how did the portfolio perform for you? Or, what was your personal rate of return?

Given that the investor made contributions during the year, he correctly references the XIRR (i.e., a money-weighted method) to obtain the personal rate of return.

Speaking of Vanguard, they have a brief video that describes the personal rate of return.

Time-weighting offers nothing personal to the investor.We need a money-weighted formula for that, be it (an unlinked) Modified Dietz, as a proxy for the true, exact, money-weighted return, or the internal rate of return. If there are no flows, then the TWRR equals the MWRR.

There is nothing "personal" about time-weighted methods. If you want a "personal rate of return," use an MWRR formula.