September 27, 1996
TO: INDUSTRY STATISTICS COMMITTEE No. 25-96
PENSION COMMITTEE No. 41-96
RESEARCH COMMITTEE No. 33-96
Retirement Task Force
RE: Proposal and Recommendation for Future Retirement Data
Collection
______________________________________________________________________________
The Investment Company Institute recognizes the importance of accurate, reliable and timely
statistics about the retirement market to the Industry’s and Institute’s legislative, regulatory and
marketing efforts. The Institute currently collects retirement statistics from mutual fund
companies in three surveys. Although each survey serves a different purpose, they overlap to
some degree, resulting in duplication of effort.
In March of 1996, an internal working group within the Institute, along with the help of a
member task force, began a review of the Institute’s data collection efforts regarding retirement
plan assets invested in mutual funds. The purposes of the review are:
To assure that the statistics collected are complete, accurate and consistent; that the items
collected are effective, efficient, nonduplicative; and that the collection imposes as little burden
to data providers as possible;
To ensure that the collection and estimation methodology meet the Institute’s needs in
addressing legislative and regulatory issues involving the retirement market; and
To ensure that members’ needs with respect to the retirement market are met.
This memo is intended to communicate a proposed methodology for future retirement
collection and to confirm the feasibility of this collection. In an effort to assess the feasibility of
this collection, a special feasibility survey is attached. Please respond to this survey for your
fund group by September 30. We recognize that you may have already answered a similar
survey, but this follow-up is intended to bring to conclusion several open issues. If you are not
the person at your fund group who is in the best position to respond, I would appreciate your
forwarding the questionnaire to the appropriate person. Additional comments or suggestions
are welcome. We hope that changes can be implemented for the 1996 annual survey period.
Proposal for Retirement Data Collection
Following the working group’s discussions and in-house meetings, it is recommended that the
Institute consolidate collection of retirement statistics into an annual collection and a
supplementary quarterly collection of information from a sample of respondents. The
quarterly asset data would be used to estimate net flows.
Information Collected
Information proposed for collection in the annual effort would include:
1. Assets invested in mutual funds by type of retirement plan or arrangement1, including:
Total IRAs, including SEP, SARSEP, Contributory, and Rollover IRAs
401(k) plans
403(b) arrangements
457 plans
Keoghs
Other defined contributions plans
Defined benefit plans
2. Number of accounts for total IRAs, and SEP, SARSEP, and Rollover IRAs
The aforementioned items would be collected from the mutual fund group for each fund within
that group. Data would be submitted electronically (using ICINet or other means), via diskette,
or on an ICI questionnaire, with each fund’s data identified by ICI fund number (or CUSIP
number at the providers option).
This information would be supplemented by data collected from brokers and banks to estimate
street name and other omnibus account statistics not available directly from the fund group.
Advantages of The Proposed Collection
After many months of discussion, ICI settled on this proposed collection, recognizing several
concerns, but with the understanding that it offers the following advantages.
Advantages
The proposed collection simplifies ICI collection of retirement statistics from three surveys to
one, thus reducing the burden for member companies.
The data can be combined to provide consistency in our current annual retirement data series;
Estimation of flow and asset allocation analysis would be easier with individual fund
information;
Data could be compared with other information in the Institute database collected on a
fund-by-fund basis;
Data are screened by the mutual funds and therefore relatively clean;
Estimation process is statistically sound because the reporting universe is quantifiable.
Concerns
Data collected from brokers and banks for street name and omnibus statistics may likely be
available only in the aggregate (not on a fund-by-fund basis), requiring additional procedures
for asset allocation.
Mutual fund companies may be unable to supply us with detailed data on plans or
arrangements for which they are not recordkeeper.
Many of you have been involved in the retirement review and we thank you for your
invaluable assistance. Please fax responses to the special survey and all comments to ICI
Research - Retirement at (202) 326 - 5959. Your input is important to the decision making
process.
1 For 1997 and subsequent years, the Institute will collect data relating to the new SIMPLE plan
created by the recently enacted minimum wage legislation.