Yesterday, I introduced a case of going from a USD to a GBP return (via an FX rate conversion), then taking the underlying USD assets and converting them to GBP, deriving the return that way, and converting (via the FX converter) to USD [quite a mouthful]. Recall that the GBP and USD results were different, depending on how they were derived. Thus, the problem: what's going on?
Simple answer: Modified Dietz.
Recall that Modified Dietz (MD) is an approximation formula; that is, its result approximates the true, exact TWRR (time-weighted rate of return). We began with a USD portfolio where there was neither a gain nor a loss for the month; even though there were cash flows, our MD return equaled zero, which is what we would have gotten with the exact method. Therefore, the FX conversion worked perfectly fine.
When we converted the USD values (beginning and ending value, along with the cash flows) to GBP, however, we had a problem. This was because we were now dealing with a non-zero (approximate) return, and MD is sensitive to the size, timing, and frequency of cash flows. Plus, you may have noticed that one of the flows was more than 17% (i.e., "large"), which should result in a revaluation, at least on that date.
For our client I recalculated the GBP return, but this time revalued the portfolio daily. This graphic summarizes all the details (you can click on it to make it larger, for easier viewing):
The spreadsheet is repeated here:
And as you can see, by calculating our GBP return on GBP values, and revaluing when flows arise, we tie out to the earlier derived return of -0.0188 percent.
Well, that was fun for me ... I always enjoy solving puzzles. Hope you found it of interest. Feel free to chime in with your thoughts.
p.s., I mentioned yesterday that Sarah Ringle introduced the FX return conversion formula to me. She cautioned that this is a quick conversion methodology, and that the standard approach is to convert the underlying valuations to the alternate currency, in order to capture the correct unrealized and realized gain/loss including the derivative estimated amounts.
Friday, August 3, 2012
Thursday, August 2, 2012
A multi-currency return puzzle for you to solve (part 1)
A Spaulding Group client posed the following situation to me, which I hope you find of interest:
They calculated the return on a portfolio in US Dollars (USD). They next converted the return to Pound Sterling (GBP). Everything worked correctly; i.e., the returns make sense and are accurate.
They next converted the portfolio's assets from USD to GBP. They measured the return on the GBP-based portfolio and got a different return than they got when they simply converted the USD portfolio's return using an FX conversion formula, which I'll show you shortly. And, as expected, when they converted the GBP return they just calculated to USD, it didn't match the earlier return. The following graphic summarizes what occurred and our expectations:
Here are the details behind what is occurring, which only makes this problem a bit more challenging.
As you can see from this table, we began with $23 million, took out $4 million on the first, another $1 million on the 28th, and ended the month with $18 million. Simple arithmetic reveals that we made nothing and lost nothing. Therefore, our return of 0.00% is accurate. This activity occurred in March 2012 (a 31-day month), and the weighting reflects it. Modified Dietz was used to derive the return.
We next want to convert the return to GBP. This can be done by using a fairly simple formula:
With this in hand, along with the following FX Rates:
We find that the GBP return equals the FX return for the period (-0.0188%). This makes sense, since the USD return is zero, the only thing that would affect our return would be the conversion from USD to GBP, which is reflected in the FX rate change, from the beginning to the end of the month.
Now, we want to convert our USD values (both the starting and ending values, as well as our cash flows) to GBP. We use the FX rates that correspond to the date of each value, and get the following:
The table shows the return we get when we use Modified Dietz. And it's obviously not what we got earlier: it's materially different.
At this point it's probably not a surprise for you to learn that when we convert the GBP return back to USD (using our FX conversion formula) we get something other than 0.00%; we get 0.0626 percent, which gain is a materially different value.
What's wrong? Does it not make sense that we should have a closed system? That going from USD to GBP, then from GBP to USD, we should have returns that all tie out?
I invite you to check the math: you'll see that it's all correct. And so, what is happening? If you can figure it out, send me an email. I will reveal the solution tomorrow (at least the one I have!).
p.s., I thank Sarah Ringle of Alliance Bernstein for introducing me to the FX conversion formula more than a decade ago, when I was doing a consulting assignment for Alliance Capital.
They calculated the return on a portfolio in US Dollars (USD). They next converted the return to Pound Sterling (GBP). Everything worked correctly; i.e., the returns make sense and are accurate.
They next converted the portfolio's assets from USD to GBP. They measured the return on the GBP-based portfolio and got a different return than they got when they simply converted the USD portfolio's return using an FX conversion formula, which I'll show you shortly. And, as expected, when they converted the GBP return they just calculated to USD, it didn't match the earlier return. The following graphic summarizes what occurred and our expectations:
Here are the details behind what is occurring, which only makes this problem a bit more challenging.
As you can see from this table, we began with $23 million, took out $4 million on the first, another $1 million on the 28th, and ended the month with $18 million. Simple arithmetic reveals that we made nothing and lost nothing. Therefore, our return of 0.00% is accurate. This activity occurred in March 2012 (a 31-day month), and the weighting reflects it. Modified Dietz was used to derive the return.
We next want to convert the return to GBP. This can be done by using a fairly simple formula:
With this in hand, along with the following FX Rates:
We find that the GBP return equals the FX return for the period (-0.0188%). This makes sense, since the USD return is zero, the only thing that would affect our return would be the conversion from USD to GBP, which is reflected in the FX rate change, from the beginning to the end of the month.
Now, we want to convert our USD values (both the starting and ending values, as well as our cash flows) to GBP. We use the FX rates that correspond to the date of each value, and get the following:
The table shows the return we get when we use Modified Dietz. And it's obviously not what we got earlier: it's materially different.
At this point it's probably not a surprise for you to learn that when we convert the GBP return back to USD (using our FX conversion formula) we get something other than 0.00%; we get 0.0626 percent, which gain is a materially different value.
What's wrong? Does it not make sense that we should have a closed system? That going from USD to GBP, then from GBP to USD, we should have returns that all tie out?
I invite you to check the math: you'll see that it's all correct. And so, what is happening? If you can figure it out, send me an email. I will reveal the solution tomorrow (at least the one I have!).
p.s., I thank Sarah Ringle of Alliance Bernstein for introducing me to the FX conversion formula more than a decade ago, when I was doing a consulting assignment for Alliance Capital.
Wednesday, August 1, 2012
Should you have a "performance bible"?
Yesterday's WSJ had an article by Brian Costa titled "The Rookie and His Pitching Bible," about "the [New York] Met's most promising rookie," Matt Harvey, who relies on a "pitching bible" he created and maintains with much care and attention. "[Harvey] records every mechanical adjustment he makes, even if only temporary, as a reference for the future. When he pitches well, he notes what he did right. When he pitches poorly, he types in a summary of his mistakes, be they mechanical or mental."
In the world of investment performance we refer to such analysis as "attribution," identifying and classifying what worked, and what did not. And while we don't typically refer to our attribution reports as our "performance bible," perhaps such diligence, across time, wouldn't be such a bad idea.
If we look over the past month of July and evaluate what worked and what didn't, that's great. But, how does that compare to June, May, or April? Or last July? Is there a pattern? Have skills shifted? Do certain decisions work better at some times, but not at others?
A temporal evaluation, such as what Harvey does, builds upon the single period evaluation, to provide a macro view, which no doubt offers much benefit to the pitcher, and would for the performance manager and his/her team, I would think.
Your thoughts?
In the world of investment performance we refer to such analysis as "attribution," identifying and classifying what worked, and what did not. And while we don't typically refer to our attribution reports as our "performance bible," perhaps such diligence, across time, wouldn't be such a bad idea.
If we look over the past month of July and evaluate what worked and what didn't, that's great. But, how does that compare to June, May, or April? Or last July? Is there a pattern? Have skills shifted? Do certain decisions work better at some times, but not at others?
A temporal evaluation, such as what Harvey does, builds upon the single period evaluation, to provide a macro view, which no doubt offers much benefit to the pitcher, and would for the performance manager and his/her team, I would think.
Your thoughts?
Tuesday, July 31, 2012
How best to handle orphan income
A common challenge with asset managers is how to deal with what might be termed "orphan income": income that arrives after the asset that produced it has departed the portfolio. This often happens when we sell a stock position after it has gone ex (i.e., any purchasers of the stock after this date will not be entitled to the dividend, but the sellers will be, even though they sold their shares) but before the dividend has been received. It is not typical with fixed income securities, because when a bond is sold, the purchaser typically pays the seller any accrued income.
Case in point: you own stock that goes ex dividend on the 5th of the month. You sell it on the 6th, and the pay date is the 15th. When the money arrives, there is no asset to tie it to. If you measure performance on that stock, you find yourself going from a zero market value to the value of the dividend, which can be a bit messy.
Option #1: shift the date you received the money (the 15th) back to the date you sold it (the 6th). That makes everything work out fine, right? This would not be appropriate, since you shouldn't adjust the pay date, as this is altering the true accounting that occurred. This can be especially problematic at year-end, as you would receive income in the year prior to its actual receipt.
Option #2: carry an accrual until pay date. THIS makes sense, does it not? Ideally you're carrying an accrual, anyway, once the stock went "ex." But if you didn't, create an accrual in cases like this. This way, you will derive the correct return and not mess with the accounting.
Have different thoughts? Chime in!
Case in point: you own stock that goes ex dividend on the 5th of the month. You sell it on the 6th, and the pay date is the 15th. When the money arrives, there is no asset to tie it to. If you measure performance on that stock, you find yourself going from a zero market value to the value of the dividend, which can be a bit messy.
Option #1: shift the date you received the money (the 15th) back to the date you sold it (the 6th). That makes everything work out fine, right? This would not be appropriate, since you shouldn't adjust the pay date, as this is altering the true accounting that occurred. This can be especially problematic at year-end, as you would receive income in the year prior to its actual receipt.
Option #2: carry an accrual until pay date. THIS makes sense, does it not? Ideally you're carrying an accrual, anyway, once the stock went "ex." But if you didn't, create an accrual in cases like this. This way, you will derive the correct return and not mess with the accounting.
Have different thoughts? Chime in!
Monday, July 30, 2012
Why it is good to look at outliers
In the past two weeks I've conducted two software certifications for Spaulding Group clients, and in both cases they had controls to test portfolio returns against specified benchmark tolerances. And, as you might expect, the fact that a portfolio blows past a tolerance level doesn't mean it's wrong; merely that it's worthy of review. After all, we expect to see outliers appear.
The London Olympic games already have a case of someone going past a benchmark, but really not a benchmark we'd typically expect to see. The London Mail had an article about 16-year old Chinese female swimmer Ye Shiwen's performance that not only beat the female record, but also surpassed American swimmer Ryan Lochte's time in the last 50 meters. "Ye Shiwen posted such a good time that her final 50m was in fact faster than Lochte's performance in the men's event, at just 28.93 seconds...Her achievement was so unprecedented that it even led some broadcasters to question whether Ye had benefited from underhand practices."
Perhaps Ye did not benefit from any banned substance. But such an extraordinary performance is worthy of review, and this is what will occur. Controls such as this in performance measurement should almost be routine, as a way to catch errors.
The London Olympic games already have a case of someone going past a benchmark, but really not a benchmark we'd typically expect to see. The London Mail had an article about 16-year old Chinese female swimmer Ye Shiwen's performance that not only beat the female record, but also surpassed American swimmer Ryan Lochte's time in the last 50 meters. "Ye Shiwen posted such a good time that her final 50m was in fact faster than Lochte's performance in the men's event, at just 28.93 seconds...Her achievement was so unprecedented that it even led some broadcasters to question whether Ye had benefited from underhand practices."
Perhaps Ye did not benefit from any banned substance. But such an extraordinary performance is worthy of review, and this is what will occur. Controls such as this in performance measurement should almost be routine, as a way to catch errors.
Thursday, July 26, 2012
Results not typical (too bad there isn't a GIPS-type standard for all advertising)
While exercising this morning, I caught an advertisement on the television for a supplement to help men guard against prostate cancer. I had the sound off, so don't know what was said, but at one point noticed the following:
Results not typical.
Difficult to read? It says "Results not typical." Imagine trying to read it from across the room! It was in rather small white print, against a light blue background.
Yes, it was there; but clearly (no pun intended) not meant to be read by everyone, save for those who were sitting quite close to the screen. And yet their advertisement had testimonials which were apparently "not typical."
To present performance of outliers would, of course, be an example of non-typical results. And while a distribution has outliers on both the positive and negative ends, I doubt that we will see an advertisement touting the results of those outliers that are on the left side of the curve.
GIPS(R) (Global Investment Performance Standards), of course, guards against this by requiring the inclusion of all accounts that meet the criteria the strategy being marketed; not individually, but collectively, with a dispersion method (if there are six or more accounts present for the full period) to provide even greater insights.
Many men over 50 will no doubt pay close attention to advertisements offering aids to guard against prostate cancer, and many (most?) will overlook the fine print, but will willingly pay for a supplement whose results may, in reality, not do much at all. Likewise, a manager who sells their money management services by highlighting the performance on the far right of the distribution will also be bringing on clients who will not normally achieve such results.
What this of course also means is that any time information is shared with a prospect that comes from a subset of the entire composite (e.g., a representative account's attribution results), sufficient details must be shown so that what it truly represents is understood.
Results not typical.
Difficult to read? It says "Results not typical." Imagine trying to read it from across the room! It was in rather small white print, against a light blue background.
Yes, it was there; but clearly (no pun intended) not meant to be read by everyone, save for those who were sitting quite close to the screen. And yet their advertisement had testimonials which were apparently "not typical."
To present performance of outliers would, of course, be an example of non-typical results. And while a distribution has outliers on both the positive and negative ends, I doubt that we will see an advertisement touting the results of those outliers that are on the left side of the curve.
GIPS(R) (Global Investment Performance Standards), of course, guards against this by requiring the inclusion of all accounts that meet the criteria the strategy being marketed; not individually, but collectively, with a dispersion method (if there are six or more accounts present for the full period) to provide even greater insights.
Many men over 50 will no doubt pay close attention to advertisements offering aids to guard against prostate cancer, and many (most?) will overlook the fine print, but will willingly pay for a supplement whose results may, in reality, not do much at all. Likewise, a manager who sells their money management services by highlighting the performance on the far right of the distribution will also be bringing on clients who will not normally achieve such results.
What this of course also means is that any time information is shared with a prospect that comes from a subset of the entire composite (e.g., a representative account's attribution results), sufficient details must be shown so that what it truly represents is understood.
Wednesday, July 25, 2012
What do you need to do when you change your return methodology?
Last week I began to conduct a software certification for a Spaulding Group software vendor client. My primary contact mentioned that they changed their policy for timing of cash flows, so that inflows default to start-of-day events, while outflows default to end-of-day (though these can be overridden, if necessary). She asked if anything needs to be disclosed, and also, what happens if they recalculate history because of an as-of trade?
First, we would not expect anyone to recalculate history, just because a change is made to the formula, with only one exception: if the method was flawed (we've found flawed formulas used at a few firms, who developed their own system, using their own in-house designed formula). Okay, maybe a second example: if an asset manager wants to comply with GIPS(R) (Global Investment Performance Standards), and the one(s) they had been using historically don't comply with GIPS, so in order to comply, they have to recalculate past performance).
As for disclosures, I'd say that they need to indicate the timing of changes and what formula(s) is(are) used at which time.
This disclosure can also include a statement that in the event of as-of trading history is recalculated, the newest employed formula at that point in time will be used.
Make sense? Have a different view? Chime in!
First, we would not expect anyone to recalculate history, just because a change is made to the formula, with only one exception: if the method was flawed (we've found flawed formulas used at a few firms, who developed their own system, using their own in-house designed formula). Okay, maybe a second example: if an asset manager wants to comply with GIPS(R) (Global Investment Performance Standards), and the one(s) they had been using historically don't comply with GIPS, so in order to comply, they have to recalculate past performance).
As for disclosures, I'd say that they need to indicate the timing of changes and what formula(s) is(are) used at which time.
This disclosure can also include a statement that in the event of as-of trading history is recalculated, the newest employed formula at that point in time will be used.
Make sense? Have a different view? Chime in!
Subscribe to:
Posts (Atom)


.png)










