From Steve's Blog

Subscribe to Steve's Blog

Categories

Archive


Dealing With Data


Posted By: Steven D. McDonnell | Posted On: Sep 26, 2014

Much of the past two weeks I’ve been working on a number of projects, among them updating my spreadsheet of data related to variable insurance volatility-managed portfolios, and I intend to post a second quarter update on Monday. For a number of reasons, this time around I found the compiling exercise a rather tiring and perplexing one.

Firstly, during the quarter a large number of existing variable funds – in particular from AXA Financial – were reworked to add vol management, which to my mind raised a few questions: should I show assets of these portfolios for periods before the change? Should I list the actual commencement date of the fund or only note when it became vol managed?  I think I figured out how to deal with those issues, which I’ll share with you in the commentary I’ll post on Monday.

Then there’s the fact that the addition of these revised funds greatly increased the total assets in my data set; thus I can’t say recent asset growth was entirely for organic reasons (i.e., through net flows). So when I cite the data I will probably put in some sort of disclaimer.

And I second-guess myself regarding how to categorize the various management layers for these funds. There are quite a few in the set that invest in just one other fund managed by a third-party, combined with a vol-managed overlay. Some are structured as master-feeder funds. So the fund in question will have an advisor, often affiliated with the insurer, but technically no sub-advisor, although the name of the asset manager working on the underlying fund will appear in the vol-control fund’s name.

I hope to have this pretty well covered in my spreadsheet. To avoid confusion, in the section under sub-advisor I have put either reference to the master-feeder arrangement, or the words: “underlying fund managed by…” followed by the name of the manager.

I sometimes dither about what to do with funds that have ceased operations for one reason or another.  In most cases, for historical purposes I keep them in the asset survey, however in the “active” funds section I omit them. For portfolios that were merged out of existence – rather than closed and liquidated – since the assets moved to a destination fund, I remove mention of the merged fund from the active list and asset tally. Is this a seamless approach? I’m not positive. If I were to calculate net flows (which  I haven’t tried yet), these situations would be tricky to address, I think.

An anomaly I’ll mention is that a few variable trusts report on a fiscal year basis – thus the fund assets for Q2 won’t be through June 30 but April 30. For now I am simply making a note of this in the sheet; I have not attempted to create an estimated number for those portfolios to link up with the rest that report based on a calendar year. Not yet, anyway.

So it’s an inexact science. I am tinkering with the spreadsheet before I post it on Monday. Thus far there are 187 portfolios in the 2Q14 set. Two AXA funds I had in the coverage earlier were reworked and renamed and I dropped them because I no longer see vol management in their investment objectives. Things like that will come up in the future too.

Also, I wish I could publish the data quicker, but since I gather the numbers from SEC filings, via quarterly or semi-annual reports, I have to wait until those filings come out, and often there’s a pretty long wait. For example quite a few of this year’s semi-annual reports began appearing on EDGAR in September so there can be a bit of a wait before I can start digging into the numbers.

So those are a few thoughts as to the quandaries that can arise when compiling industry data – and similar difficulties come up in other areas of the business, certainly. Despite these issues, I intend to keep putting out quarterly updates on vol-control funds, as they are of much interest to the industry. I will look to refine my coverage as I go.