Fidelity Investments
82 Devonshire Street
Boston, MA 02109-3614

February 27, 2003

Jonathan G. Katz, Secretary
U.S. Securities and Exchange Commission
450 5th Street, NW
Washington, DC 20549-0609

RE: Shareholder Reports and Quarterly Portfolio Disclosure of
Registered Management Investment Companies

Release No. 33-8164; File No. S7-51-02

Dear Mr. Katz:

Fidelity Management & Research Company, the investment advisor to the Fidelity Investments mutual funds, appreciates the opportunity to comment on the Securities and Exchange Commission's proposals to modify information provided to fund shareholders and the public on a semi-annual and quarterly basis. We strongly support the goals of the proposal, but we have concerns about how the proposal would be implemented.

I. Summary of Contents

A. Quarterly Holdings Disclosure. We support the Commission's proposal to make full mutual fund holdings available to the public on a quarterly basis, with a 60-day lag. The fund industry has made quarterly holdings disclosure work in the past, and it can work in today's environment as well. However, we are concerned about potential adverse consequences of holdings disclosure, particularly if funds choose to disclose holdings earlier than required by the proposal, and strongly encourage the Commission, in adopting its final rules, to reassess its proposal in light of these concerns.

Mutual funds are subject to the most thorough disclosure regime of any financial product, and this nation's investors have greatly benefited as a result. Regular, public disclosures of fund information -- not just holdings, but performance data, share prices, investment policies, fund expenses, portfolio manager changes, and so on -- have helped build and maintain public confidence in mutual funds and have helped them grow. In addition, the disclosure regime has fostered the development of a separate industry of fund trackers, analysts and advisors that follow funds' activities and pass that information along to their clients, adding to the mix of available information. Funds have thrived under this scrutiny, and have grown to be the investment vehicle of choice for the average investor.

Mutual funds' success has its costs, however, in the form of copycats, free-riders and front-runners who can profit from knowledge of fund shareholders' holdings. As mutual funds' assets have grown, the potential profits to be made by trading against mutual funds -- at the expense of fund shareholders -- have grown as well. The Commission plays a vital role in balancing the requests of those who want increased disclosure against the need to prevent fund shareholders from being disadvantaged when trading against others in the market. We believe it is essential for the Commission, consistent with its mission of protecting investors, to use its expertise to assure that an appropriate balance is maintained between disclosure and confidentiality.

With that in mind, we suggest that the Commission focus on the following three areas with respect to holdings disclosure policies:

  • First, the Commission should undertake a more thorough analysis of the costs to shareholders imposed by portfolio holdings disclosure in the form of sacrificed fund returns.

    Disclosing portfolio holdings to a fund's shareholders cannot be accomplished without disclosing the same information at the same time to the trading markets and all the participants in those markets. This can put mutual fund shareholders' returns at a disadvantage when trading against other institutions and individuals that are not required to disclose their positions. The costs to shareholders of front-running are difficult for any one fund firm to estimate, in part because front-runners are naturally anxious to conceal their activities from the funds they target. But the potential costs to shareholders are enormous. We estimate that the total value of the information contained in fund holdings -- i.e., the profits to be made by trading with knowledge of those holdings -- to be measurable in billions of dollars. If this information were revealed to the market on a daily basis, the result would be a massive transfer of wealth from mutual fund shareholders to institutions and professional traders. This risk will become more relevant as the industry moves toward more frequent holdings disclosure, particularly as fund advisors increasingly face pressures to disclose information more frequently than quarterly, or with less than a 60-day lag, and as technology makes it easier to put such disclosures to work in trading strategies. In the proposing release the Commission provided a theoretical discussion about the probability of front-running, but was unable to quantify any costs to fund shareholders. Given the magnitude of the dollars at risk and the interests of fund shareholders, we urge the Commission, in the strongest terms, to do more.

  • Second, the Commission should consider requiring funds to adopt a minimum time lag before holdings are disclosed to the public.

    Fund holdings data represent the intellectual property of the fund and its advisor, paid for by fund investors through the management fee. Both the advisor and fund shareholders have important interests in keeping this data confidential, counterbalanced by interests in making the information public from time to time. To set the right balance between confidentiality and disclosure, we believe a time lag is essential. Based on our experience, the 60-day lag proposed for quarterly holdings disclosure should be adequate to inhibit front-running behavior while also providing reasonably current information to the public. However, fund advisors are frequently subjected to demands for more current information from institutions, advisors and potential clients, even if that disclosure poses a risk of adversely affecting fund returns. A sufficient time lag before fund holdings are released is necessary to protect the confidentiality of fund holdings against any potential conflicts of interest that an advisor may face in these circumstances. We suggest that the Commission, in its final rule, require advisors to obtain approval of fund boards as to their policies of lagging disclosure of holdings, and that funds be required to disclose those policies to fund shareholders.

  • Finally, the Commission should address issues raised by disparate disclosure of fund holdings and provide guidance on this issue to the fund industry and fund boards.

    While most fund companies do an admirable job of disclosing fund holdings to their shareholders, many also provide access to fund holdings more rapidly, more frequently, or both to fund trackers, advisors, consultants or other firms with which an advisor has a business relationship. For example, a fund advisor may release holdings to commercial fund tracking services monthly while making them available to shareholders quarterly, or may release holdings to selected institutions shortly after quarter end and reveal them to the public at a later date. Generally, these practices are intended to add to the overall mix of information available to investors as a group. However, such practices raise serious issues involving potential conflicts of interest: they may further the advisor's interests through increased sales of fund shares to new investors, and the recipient's interests through the ability to repackage the information and sell it to the recipient's customers, but these activities may well impose a cost on fund shareholders in terms of the risk to their returns from front-running. Since it appears that practices across the industry vary widely, we request that the Commission provide guidance to funds and fund advisors in this area. We recommend that the Commission consider in its final rulemaking approaches used in other areas of potential conflicts in the fund industry - approval of policies by fund boards, and disclosure of those policies to the public - to address issues presented by these disclosure practices.

B. Mandatory Shareholder Report Mailings. The Commission also has proposed changes to the requirements for annual and semi-annual reports mailed to fund shareholders. While we support the concepts of these proposals, we have suggestions that we believe will make them more effective. The idea of reducing the amount of bulk mail required to be sent to shareholders - even if they do not want it - is a sound one, and deserves everyone's support. The Commission's proposals in this area, however, strike us as too modest, and more significant benefits could be achieved by taking advantage of today's high-tech, rapid movement of information.

We agree with the Commission that providing a summary portfolio of holdings in shareholder reports is a step toward making holdings information more useful and understandable. We also support measures designed to enhance mutual fund expense disclosure. However, by the Commission's own estimates the proposal in total could impose costs of $17 million per year on funds and the fund industry.1 This represents a substantial missed opportunity to reduce costs and increase shareholder satisfaction, by cutting down on required mass mailings while increasing the amount and usefulness of information made available to investors. Therefore, we believe the proposals should be modified as follows in order to save shareholders even more money:

  • Shareholder reports should be truly abbreviated so as to only feature the most relevant and useful fund information.

    We propose that the Commission require fund advisors to include only required performance data, management's discussion of fund performance, the new expense disclosure, the summary portfolio schedule and the Financial Highlights table in reports. Under this scenario, financial statements, notes to financial statements and other detailed data designed for expert use would be eliminated from the paper report mailed to shareholders. These details would continue to be disclosed, however, and would be available at no cost to any shareholders requesting them. We estimate these steps would save approximately $2.8 million in annual printing costs across the Fidelity funds, compared to $1.0 million under the Commission's proposal.

  • In addition, the Commission should eliminate mass-mailings of paper semi-annual reports entirely, while retaining mandatory mailing of the annual report.

    In lieu of mandating mailing of semi-annual reports, the Commission should require fund advisors to make these reports available for downloading on the advisor's web site and/or mail them to any shareholder who requests them, either on an as-needed basis or regularly if the shareholder so desires. At Fidelity, we conservatively estimate that this would save $7.7 million annually, more than seven times as much as the Commission's proposal. If this option saved seven times as much as the Commission's estimate industry-wide, aggregate cost reductions would be $39 to $49 million per year rather than the Commission's estimate of $5.6 to $7 million. Those savings could well be enough to offset the other costs associated with the proposal that have been identified by the Commission. And given the proposed disclosure of quarterly holdings across the industry, and the likely widespread dissemination of that information, it is worth questioning whether mandatory paper mailing of semi-annual reports will truly be worth $39 to $49 million a year to investors going forward.

II. Discussion

A. Frequency of Portfolio Holdings Disclosure.

1. Benefits of Disclosure. Shareholders and the advisor clearly have an interest in periodic disclosure of fund holdings, although it is not simply a case of more is better. For the advisor, disclosure aids in selling the fund and helps build public confidence in mutual funds as a group.2 For the shareholder, periodic public disclosure serves as a check on the fund's governance structure.3 In addition, full disclosure of quarterly fund holdings will enlist the support of professional investment businesses, such as investment advisors, plan consultants and fund tracking services, in publicizing fund holdings and strategies. These businesses have long asked for additional information on fund investments, for their own business purposes and to redistribute the data to their clients. We believe these businesses serve a valuable commercial purpose in informing investors who prefer not to obtain information from fund companies directly or who want an additional third party view, and that the Commission is right to enlist their aid in providing information to the public.4

2. Competing Interests Regarding Holdings Disclosure. A holdings disclosure policy must balance a set of competing interests. On the one hand, fund shareholders and the fund advisor have an interest in keeping fund holdings confidential, to protect fund returns from erosion by traders and to protect proprietary research from free-riders and copycats. On the other hand, fund shareholders and the fund advisor also have an interest in making holdings public periodically, so that shareholders can see in detail how their money is invested, so that the advisor can effectively market the fund to new investors, and so that general public confidence in mutual funds can be maintained.5

In addition, there are the interests of a variety of third parties, with differing economic interests and motivations. Some are prospective investors who want to understand how a fund's portfolio is invested before making a purchase decision. Others are fund-tracking services and consultants who make a business of analyzing fund investments and marketing that analysis to investors and/or fiduciaries. Still others are fund holdings services that sell re-packaged fund holdings information to their clients -- not to assist them in choosing a fund for investment but to help them devise strategies to gain a trading advantage over their counterparties in the secondary market (which may be the very mutual funds that publicly disclosed their positions). Finally, there are financial professionals such as investment advisors who want detailed information on funds in order to incorporate that into the advice they provide to their clients.

In our experience, the greatest pressure for detailed fund holdings data comes from third parties rather than from current or prospective shareholders -- especially third parties that provide a commercial service of analyzing or re-packaging and selling the information. Although these commercial entities serve a valuable purpose, as non-shareholders they have no direct stake in the fund and no particular economic interest or incentive for keeping holdings confidential.

3. Historical Disclosure Practices. The question of holdings disclosure is not new, and neither are the principles, incentives and conflicts involved. In 1926, a time when only a few dozen funds (or "investment trusts") existed in the US, one writer surveyed fund disclosure policies as follows:

"Most American investment trusts have at one time or another, in their brief period of existence, cultivated public confidence by revealing their investments at a particular time, or by publishing a list of "approved securities" among which their purchases are made. Practices of publicity, however, are less common among the investment trusts of Great Britain where longer public experience of the advantages of this form of investment has created considerable confidence in the ability and integrity of those prominently associated with trust organization and management. Probably less than a third of the investment trusts of Great Britain take the public fully into their confidence by printing annual investment lists.

* * * *

The principal objection voiced by investment trusts to the periodical publication of their investments is that the facts thus broadcasted often subject them to gratuitous criticism on the part of ill-informed investors, while practical use is made of these facts mainly by brokers on behalf of their clients, or by competitive trusts not so well equipped to ferret out investment opportunities for themselves. Moreover, it is felt by many trusts that they would be handicapped in realizing the best price on their holdings if the knowledge of the extent of these holdings were public property."6

The issues presented by the author, in the days before the adoption of modern securities regulation, are still at issue today: funds disclose holdings as a way of cultivating public confidence, but are concerned that holdings, if disclosed, would be used by competitors or others and could adversely affect fund trading costs.

The modern fund industry has had experience with quarterly holdings disclosure in the more recent past. Like many other fund groups, Fidelity mailed quarterly holdings reports to its fund shareholders in the 1970s and early 1980s, at a time before electronic distribution was practical. These reports included a complete fund holdings list accompanied by a brief discussion of performance over the quarter, and were typically in the form of a 12-panel folded brochure, the equivalent of approximately six pages of information in a two-column format. They were mailed out on a schedule corresponding to the mailing of annual and semi-annual shareholder reports (i.e., approximately 60 days after the end of the fiscal quarter). The quarterly mailings were discontinued in 1983 largely out of a concern that the cost of printing and mailing was not justified by the utility of the information to fund shareholders. Since the Commission's proposal wisely allows for electronic distribution of this information rather than paper mailings, these cost concerns are much reduced, and we agree that it is now practicable and worthwhile to revive the practice of quarterly disclosure.

4. Current Holdings Disclosure Practices. At Fidelity, we have developed a set of disclosure policies that attempt to balance the interests of fund shareholders, the fund advisor, and prospective investors and those who assist them. Like most fund groups, we disclose substantial information on holdings in addition to annual and semi-annual reports containing full portfolio lists. For a typical domestic equity fund, we provide the following information on holdings to the general public:

Full holdings (annual and
semi-annual reports)
As of Date
Fiscal year-end and
Fiscal half-year-end
45-60 days, depending on
printing schedules
Top 10 holdings (without
individual percent weights)
Calendar quarter-end 15 days
Industry weights as % of assets
Allocation by asset class
Calendar month-end
Calendar month-end
30 days (15 days at quarter-end)
30 days (15 days at quarter-end)

We have designed our disclosures of fund holdings in an attempt to make them fully informative to investors - but not to make it obvious which securities exactly the funds own or are trading at any given time. These policies are carefully crafted to attempt to keep this balance.7

In general, holdings of Fidelity funds are available to all investors on an equal basis - either the holdings are made available to all investors or are kept confidential from all investors (other than the fund board of trustees as representatives of the shareholders). In a small number of cases, fund holdings are made available to consultants or institutional investors on a confidential basis, subject to non-disclosure agreements that strictly limit the uses to which the information can be put and prohibit re-publication or re-dissemination of the data. In addition, Fidelity from time to time receives non-public information about other funds' holdings, typically subject to non-disclosure agreements as well.

Fidelity does not make holdings available to fund-tracking services or other entities that re-publish holdings data unless Fidelity makes those holdings available to fund shareholders on an equally timely basis. Most fund tracking services find this policy frustrating, and by adhering to it, we believe Fidelity incurs a competitive disadvantage in attracting fund assets.8

Fidelity's holdings disclosure policy is published in a handbook for internal use, and is overseen by a committee that has representatives from our portfolio management, legal, marketing, and operations groups. We have not asked the fund board of trustees to approve the policy in detail, although we have discussed the concepts with the board from time to time.

5. Risks of Disclosure. More frequent disclosure of an advised account's holdings provides advantages to investors who are not clients of the advisor. Frequent disclosure of client holdings can allow others to discern the advisor's investment and trading techniques (something the advisor naturally does not want revealed to other advisors). This knowledge can be used to help investors understand a fund and to make better investment decisions, but can also be used to facilitate free-riding and front-running behavior.

When considering the risks associated with additional disclosure of shareholders' holdings, it is important to recognize the extent of information already available to any party that wants to estimate a fund's current holdings and current trading activity. Most if not all funds currently disclose information about their holdings in addition to annual and semi-annual shareholder reports, through mandatory 13F filings9 and through disclosures of top ten or top 25 holdings, industry breakdowns, and other portfolio characteristics that are voluntary as a legal matter but mandatory from the point of view of investors.

Perhaps more importantly, funds provide much more information about their operations than public operating companies or other collective investment vehicles in one critical respect: funds publish their net asset values per share (NAVs) every day without any time lag. A substantial amount can be inferred about a fund's trading activity from its daily NAVs alone, and computer programs that specialize in estimating portfolio changes by analyzing daily fund returns are well-known to market participants. Releasing actual fund holdings, even after a long time lag, allows others to confirm their estimates of fund activity and improve their models for anticipating fund trades in real time. As advisors reveal more information about fund holdings, adding fiscal quarterly holdings data to the calendar quarterly data currently requested by most institutions (and mandated for 13F reporting), an exponentially greater number of "snapshots" of fund holdings enter the marketplace. We've always believed that the frequency with which fund holdings are disclosed is critical to the efforts of those who wish to trade on such information to their own benefit - and to the detriment of fund shareholders.10

Knowledge of clients' holdings or trading habits may be particularly useful in the event of a change in management at the advisor or a period of increased sales or withdrawals on the part of the advisor's clients. When a fund manager changes, traders will scrutinize the fund's most current holdings to anticipate changes in the portfolio that will result in buys or sells in the market. If a fund is growing or shrinking rapidly, traders will prepare for buys or sells of the fund's favorite holdings as the manager invests or raises cash.11 The purpose of these activities is to front-run the anticipated trades in the hopes of achieving a trading profit.12

As investors ourselves, we will take maximum advantage of other investors' public disclosures to the extent permitted by law. Traders will use every legitimate advantage to obtain a better price from the trader on the other side. It is axiomatic that if Trader A knows what is in Trader B's portfolio and Trader B does not have the same knowledge about Trader A, then Trader A will benefit at Trader B's expense. The advantage is similar to knowing some of the cards in your opponent's hand, and if the rules allow one to know that, a rational investor will try to do so. In light of our hands-on experience on both sides of this issue, we believe that the Commission should be cautious in requiring funds, as the investment vehicle of choice for tens of millions of average investors, to disclose their holdings when others (including investment vehicles for wealthy investors) are not required to do the same.13

6. Need for Improved Cost Estimates. While the trading impact costs of revealing holdings to the market are hard to estimate, the cost of imprudent or careless holdings disclosure could be enormous. As an extreme example, few investors would want to hire an investment advisor that disclosed clients' holdings to the public on a daily basis.14

The proposing release included an abstract economic discussion that appeared to be intended to discount the potential costs to fund shareholders of front-running.15 The analysis was dismissive of the probability of front-running having a negative effect on fund returns, although it expressed no definite conclusions, and comment was requested on the analysis. We know that this probability is not remote, as we see the opportunities and possibilities on our trading desks on a regular basis, using information that is in the public domain.

The theoretical analysis of potential front-running in the proposing release is disappointingly imprecise, and the chain of hedge words (it appears has been appears) undermines its force. Contrary to the assumptions that underlie the analysis, it is not necessary to have perfect knowledge about a fund's holdings in order to profit from trading against mutual fund shareholders. It is only necessary to have more information about what your counterparties own than they have about your holdings.

We suggest that, to assess empirically the costs of front-running, the Commission take advantage of expertise in its staff who supervise the activities of broker-dealers in the secondary markets. It is undisputed that holdings information has value to traders - they would not pay to have access to 13F information if it did not - although there is debate as to how much less valuable that information becomes with a long time lag. While we anticipate, based on our experience, that front-running costs to funds should be small as long as a 60-day lag is adhered to, we are not in a position to obtain information from other market participants as to the extent of their front-running activities. We urge the Commission, which has regulatory authority over many of the parties that may front-run mutual funds, to quantify the actual costs of disclosure, and the potential benefits of disclosure lags as a way of mitigating those costs. We suggest that the Commission use its authority as the regulator of the securities markets, including broker-dealers, specialists and other professional traders, to investigate the actual uses to which fund holdings have been put by traders. A practical evaluation of the actual trading profits obtained by front-runners (and of the costs incurred by mutual fund shareholders as a result) would be more pertinent than the inconclusive hypotheses contained in the proposing release. And, armed with data, the Commission's rulemaking will better serve the public interest.

7. Potential Quantification of Trading Costs. The risk of holdings disclosure from a trading perspective is essentially a matter of giving the fund's trading partners information about the fund's potential trades. While it is difficult to identify the actual impact of front-running on funds in the absence of a specific cost estimate from the Commission, the value of information can be estimated by examining commission costs in the context of institutional program trading.

Major dealer firms will execute program trades for institutional investors without necessarily knowing the names of the specific securities included in the program -- a so-called "blind principal bid" trade. When executing a large program trade for an institutional client such as a fund, a typical dealer will provide a range of commission charges depending on the amount of information provided in advance of the trade. The more information provided o the dealer, the lower the commission, and the less information provided, the higher the commission. The difference reflects the difference in cost to the dealer of executing the trade on the institution's behalf -- or, viewed another way, the difference in potential profit to the dealer if the dealer knows more about the holdings involved in the trade. Thus, the difference between the commissions when less or more information is included in the trade is one estimate of the money to be made from knowing which securities a fund is trying to buy or sell.

Commissions for program trading vary based on the types of securities involved and the size of the trade as well as the amount of information provided. For a blind principal bid for a large diversified program trade (i.e., an agreement to deliver a diversified basket of securities without knowing exactly the names of the securities in the basket) in US equities, we estimate the range of commissions to be at least 25 basis points (0.25%) between a trade with minimal information provided (such as average market cap, P/E ratio, and other general characteristics) and a trade with significantly more information (such as the name of the fund and the names of the specific stocks involved). Thus, having more information about a fund's program trade could be worth at least 0.25% of the trade amount to the trader on the other side.

Using this estimate, the total value of the information contained in fund portfolios is staggeringly large: with over $2 trillion invested in equity funds, if equity funds incurred an additional trading cost of 0.25% on each trade, assuming a 50% average annual turnover rate the cost for all equity funds would be over $2.5 billion per year in lost returns to mutual fund shareholders. This estimate approximates the value of having current information about the holdings in a fund's portfolio at the time a trade is placed, and almost certainly overestimates the value of having holdings information that is lagged 60 days. However, even if lagging the disclosure of holdings information reduced its value by a factor of 100 times, the cost to funds would still be over $25 million per year. This illustrates, at the very least, the importance of having a policy of lagging disclosure of fund holdings.

We believe this potential risk is too large to be dismissed solely on a series of theoretical assumptions, and believe the Commission should perform additional fact-based analysis and advise funds of the results.

8. Portfolio Holdings: Role of Fund Boards. The investment positions of a fund clearly can be seen as intellectual property of the fund and its advisor and each have a financial interest in fund holdings that should be protected.

Because a portfolio holdings disclosure policy should strike a balance that advances the twin interests of fund shareholders -- providing sufficient information to form investment decisions but limiting information to protect their fund against third parties with competing economic interests -- it would seem appropriate that a fund's board of directors play a role in approving the policy and overseeing its implementation. Involving a fund's board seems especially appropriate when one considers the interests of the fund's advisor. As we noted earlier, a fund's shareholders and advisor generally will have a common interest in striking the right balance regarding holdings disclosure, but circumstances can, and do, arise when the interests of a fund's advisor and shareholders are not fully congruent. Simply put, an adviser, to attract assets to a fund under its management, can have an incentive to provide holdings disclosure on a more frequent basis or on a disparate basis, especially to be responsive to consultants and investment advisors who are in a position to influence the flow of retirement and other investment assets.

Involving a fund's board to oversee a fund's holdings disclosure policy, under these circumstances, is analogous to the role that a board may perform in overseeing the use by a fund's adviser of the fund's shareholder list, as exemplified by the no-action relief granted by the staff in Dreyfus Liquid Assets, Inc. (pub. avail. Sep 5, 1976). That letter discussed a situation where the Dreyfus Corporation wanted to mail an offer to shareholders of the Fund inviting them to purchase a financial book at a reduced price. Dreyfus would use the shareholder list of the fund in making its mailing, but did not propose to give a lease payment or other compensation to the fund for the use of the shareholder list.

Dreyfus stated that (i) the fund's board had approved the use of the fund's shareholder list and (ii) the list, which was developed through Dreyfus's own efforts, could be considered to be joint property of Dreyfus and the Fund. Dreyfus did not claim a unilateral right to use the shareholder list, but rather sought and obtained the consent of the fund's board. The staff granted no-action relief under Section 17, provided "that the [f]und's board of directors, including a majority of the directors who are not interested persons of the [f]und, determine in good faith that benefits to the [f]und and its shareholders resulting from the transaction are sufficient to justify the transaction."

While use of a shareholder list and shaping a portfolio holdings disclosure policy clearly have their differences, they have similarities as well. A fund's shareholder list has a cash value because others are willing to pay for it for direct mail purposes and other marketing efforts. By the same token, if we were to make Fidelity fund holdings available for a fee, willing buyers would come forward. Shareholder lists are developed through the efforts of the advisor (and/or distributor) rather than the fund, but shareholders have an interest in keeping them reasonably confidential. The same is true of a fund's portfolio holdings. Finally, like the use of a shareholder list for marketing purposes, release of fund holdings for marketing purposes may benefit both the advisor and fund shareholders, though the relative benefits to each party may vary. The approach endorsed by the staff in the Dreyfus no-action letter - invoking the oversight of the fund's board - seems appropriate as well in the context of portfolio holdings disclosure policies.

9. Time Lag for Disclosing Fund Holdings. We believe a time lag is essential for funds to achieve a proper balance between the need for confidentiality and the need for disclosure. Just as the Commission has set a maximum lag time for annual and semi-annual reports of 60 days, we believe it would be appropriate for the Commission to require funds to adopt and disclose a minimum time lag for disclosure of holdings information. The maximum time lag of 60 days protects the interest of fund shareholders in knowing what is held in their fund on a timely basis; a minimum time lag would protect the interest of fund shareholders in preventing anyone else from knowing what the fund is trading currently.

As a matter of policy, Fidelity does not release any fund holdings information until at least 15 days after period end.16 We believe a strong case could be made for the Commission adopting, by rule, a minimum holdings disclosure lag for mutual funds to accompany the 60-day maximum time lag. However, we recognize that holdings disclosure may have less adverse impact for some funds than others (for example, disclosure of the top ten holdings of an index fund has little value to opposing traders), and that the Commission may wish to encourage experimental funds like the OpenFund (now liquidated) that promote full holdings disclosure as a selling point.

We recommend that funds be required to establish minimum disclosure lag policies, and that, consistent with the reasoning of the Dreyfus letter discussed above, funds' disclosure policies be subject to approval by the fund's board and by the advisor. We also recommend that funds be required to disclose their minimum lag policies to the public. Given the magnitude of the potential costs to fund shareholders from front-running of their funds' trades, we believe that the minimum lag time for holdings disclosure deserves as much attention as the maximum lag time.

10. Disparate Holdings Disclosure. Disparate disclosure of fund holdings presents a particularly difficult public policy issue. Disparate disclosure may have financial benefits to the advisor (through increased fund sales) and the recipients (through the ability to repackage or analyze the information and sell the result to their customers, or other uses of the information for their financial gain), but comes at a potential cost to the fund in terms of increased risk of front-running. And in the case of disparate disclosure, there is no clear benefit to fund shareholders as a group (except to the extent that all shareholders benefit from maintaining and growing fund assets), since the information is not made available to all shareholders on an equal basis.

Because of the potential for harm to shareholders with little clear benefit, we believe disparate disclosure of holdings information is rarely justified. In general, we have sought to manage this issue at Fidelity carefully, with particular attention to keeping holdings confidential, and we believe our peers have been sensitive to this issue as well. Nevertheless, we do not believe that fund shareholders, fund trackers, investment advisors or the fund industry benefit from regulatory silence on this issue. We believe it would be in furtherance of the interests of fund shareholders if - after study - the Commission provided guidance on how to achieve fair dissemination of mutual fund portfolio holdings. At a minimum, we suggest that funds' disparate disclosure policies (if any), like minimum time lag policies, be subject to review by fund boards and disclosed to the public.

B. Mandatory Shareholder Report Mailings.

We believe the Commission's proposals regarding the contents of paper shareholder reports are too limited. We urge the Commission to more fully embrace the idea of turning mandatory shareholder mailings into brief, investor-friendly summaries, while making full and enhanced disclosure available on request.

Abbreviating mandatory mailings is a worthy goal, especially in light of the natural tendency for mailings to get a little bit longer every year. Shareholder reports today are much longer now than they were twenty years ago, when funds typically mailed quarterly reports to shareholders. But it is not clear that the longer reports are better at communicating to investors. In our view, investors would be better served by brief, helpful disclosure with the opportunity for further technical research available at no cost for any who are interested.

We believe that investors share this point of view: several of the public comments on the proposal stressed a desire for plain English and expressed frustration at the tendency for disclosure to be highly technical. The interests of investors would be better served if the most technical parts of shareholder reports -- for example, the statement of operations and the statement of changes in net assets -- were no longer required to be mailed to all shareholders, but were still made available to expert investors and commercial data users (or anyone else) upon request. We urge the Commission to take bolder steps to reduce the burden of required paper mailings, while retaining full disclosure for investors who want the details. We strongly believe that such a policy represents the disclosure policy of the future.

Fidelity supports the effort to abbreviate the holdings provided in reports mailed to shareholders, though we note that listing top 10 or top 25 holdings seems to be a more standard industry practice (for fund groups and fund tracking services alike) than listing the top 50. For example, trimming the holdings list of Fidelity Trend Fund to 50 names would probably reduce the number of pages required for that part of the shareholder report from 12 to 3, which would be a good start.17 But Trend Fund's most recent annual report mailing also included 11 pages of financial statements, footnotes and schedules, and 9 pages of biographies of fund trustees and officers, none of which would be abbreviated under the proposal. Thus, we believe the Commission's proposal does not go far enough, and still requires too much paper to be bulk-mailed to investors when on-demand disclosure would be far more cost-effective and better received by investors.

The idea of reducing the bulk of fund shareholder reports is not new: Fidelity submitted a proposal to the Commission in 1995 recommending that the holdings listing in shareholder reports be reduced to the top 50 holdings with full holdings available on request, similar to the proposal presented now by the Commission. Our proposal, however, also recommended that all other mandatory data be dropped, except for a brief interview with the fund manager and a discussion of fund performance. We continue to believe that bolder steps than merely reducing the holdings list are needed to make mandatory shareholder reports truly useful, and to deliver true cost savings. We recognize that expense disclosure serves an important public policy objective, so there is an urgent basis for adding a page of expense disclosure.18 The proposed expense disclosure would have more impact, however, if it were part of an eight-page report than if it were one more page in a 30- or 40-page report.

Regarding financial statements, if the Commission believes it is necessary to provide a summary of the fund's financial position we recommend that the Financial Highlights -- which is already designed to be a summary -- be the only financial statements required to be included. We find it hard to conclude that investors would be better served if several pages of holdings are dropped from the bulk mailings but several pages of esoteric financial data, with confounding titles like "Statement of Changes in Net Assets," are retained. Let the investors who are skilled in understanding that kind of detail request free copies of the full report. But do not insist that funds bear the cost of mailing that detail to every shareholder in the fund.

As long as the need to disclose information is linked to a requirement to mail it to all shareholders and fill their mailboxes with paper, additional disclosure will translate directly to additional costs. We encourage the Commission to break this link completely, and eliminate the requirement to mail semi-annual fund reports to shareholders. This is not a proposal to reduce disclosure: we recommend that funds be required to deliver a copy of the full semi-annual report to any shareholder who requests it, free of charge. It is a proposal to reduce bulk mailings, and lower the costs of all mutual funds that are ultimately borne by fund shareholders.

Eliminating the requirement to mail semi-annual reports would save far more cost than merely shortening the reports, because it would eliminate not only printing costs but postage and mailing costs as well.19 We believe that the overwhelming majority of investors would not ask to be mailed semi-annual reports if delivery were changed to an on-request basis.20 The effect would be the same if the Commission eliminated the requirement to mail semi-annual reports entirely or if the Commission merely allowed funds to stop mailing reports to shareholders who did not respond to an invitation to request reports going forward (a so-called "negative response" election). The greatest cost is associated with the large proportion of investors who do not respond at all, and will neither request that reports be sent to them nor respond to a letter asking them to voluntarily stop receiving reports in paper form.

At the least, eliminating the requirement to mail semi-annual reports would provide an opportunity to gather data on the possibilities of on-demand disclosure. The Commission could evaluate the impact of changing from paper mailings to disclosure on demand, and could observe the cost savings and any adverse consequences that might occur, always reserving the right to return to paper mailings in the future. The proposing release admits that it is difficult to quantify the benefits of quarterly holdings disclosure: how much more difficult it must be to quantify the benefit of mailing fund shareholders paper disclosure documents they did not request.


Fidelity strongly supports the proposal's goals of improving disclosure to fund shareholders while reducing disclosure costs. One need only look at the history of the fund industry to see how mutual funds and fund shareholders have prospered though a tradition of accurate and helpful disclosure, which the proposal would help to maintain. We urge the Commission, however, to consider modifications to the rules before adoption, to enhance investor protection and to make the proposals truly cost-effective to the benefit of all shareholders.

* * * * *

We appreciate the opportunity to comment on the Commission's proposals to improve disclosure to shareholders. Should you wish to discuss our comments, please contact the undersigned: Eric D. Roiter or David B. Jones at 617-563-7000.


/S/ Eric D. Roiter

Eric D. Roiter
Senior Vice President and General Counsel
Fidelity Management & Research Company

  /S/ David B. Jones

David B. Jones
Vice President
Fidelity Management & Research Company

cc: The Honorable William H. Donaldson

The Honorable Paul S. Atkins
The Honorable Roel C. Campos
The Honorable Cynthia A. Glassman
The Honorable Harvey J. Goldschmid
Paul F. Roye, Director
Susan Nash, Associate Director
Paul G. Cellupica, Assistant Director
Division of Investment Management

1 The cost/benefit analysis in the proposing release estimated aggregate gross cost reductions industry-wide of $5.6 to $7.0 million per year from the use of summary portfolio schedules, mostly from savings on printing and postage. However, the Commission's analysis also projected additional costs of $22.6 million per year from the proposal, made up of $13.58 million in internal costs for new quarterly reporting requirements, $4.96 million in costs of preparing and filing Form N-CSR, and $4.07 million in costs of adding the new required summary information to shareholder reports. These cost estimates did not include any costs associated with front-running, which we believe could be substantial.
2Fidelity firmly believes one reason that mutual funds enjoy greater public confidence than hedge funds is the fund industry's public disclosure record.
3 Holdings disclosure has occasionally been proposed as a way to detect prospectus violations or other wrongdoing by funds. The potential for periodic holdings disclosure - as of a few points in time each year - to serve this purpose in an effective manner is exaggerated in our view, and reflects a misperception of the governance structure of mutual funds. Since fund holdings are confidential at most times, during most of the year it falls to fund boards of directors or trustees, as shareholders' elected representatives, to oversee fund advisors' compliance procedures. At most, periodic holdings disclosures provide an occasional check as to whether this governance structure is working properly. They are no substitute for daily compliance monitoring.
4 We believe that typical individual investors will not make significant use of full holdings information themselves, but will instead rely on financial professionals - including fund trackers, investment advisors, the financial press, and fund companies themselves - to digest and summarize the information.
5 Holdings disclosure policy is sometimes discussed in terms of a shareholder's right to know what is in their portfolio, as if disclosure could be limited to existing shareholders only. Unfortunately, given the nature of modern mutual funds, it is impractical to limit disclosure of fund holdings to existing shareholders only, if only because any trader could simply buy a few shares and thus become a shareholder. In other words, from a shareholder's perspective disclosure is not just a question of your right to know what is in your fund, but of your neighbor's right to know what is in your fund. Consequently, we believe it is reasonable for shareholders, as owners of the fund, to want to keep their fund's holdings confidential at most times. The owners of the fund have an interest in being informed of the contents of their portfolios periodically. However, the owners of the fund also have an interest in keeping the contents of their portfolios confidential, even secret, from non-owners - even at the cost of being unable to access portfolio holdings themselves.
6 Leland Rex Robinson, Investment Trust Organization and Management. New York: The Ronald Press Company, 1926: 87-89. The author goes on to note that "[...] From this it must not be assumed that publication of investments is to be condemned. On the contrary, this practice has much to commend it, and is followed by some of the largest and most conservative of the British investment trusts. A number of the latter, however, would probably gladly abandon publicity of their holdings if the contrary precedent had not been established, and the writer has never observed that public confidence in British investment trusts is affected one iota by any decision which may have been reached concerning publication or non-publication of holdings."
7 We note that summary information about a portfolio can do at least as good a job at informing investors as full holdings data. In particular, industry weightings, which we release more frequently than full holdings and with a shorter lag, provide substantial information to investors who wish to remain informed of how their funds' portfolios over shorter time periods. For example, when well-known fund manager George Vanderheiden retired from his longstanding position as manager of Fidelity Advisor Growth Opportunities Fund on February 1, 2000, we published industry weighting data at the end of each calendar quarter showing how the fund's technology position was increasing, as follows:

Holdings as of: 12/31/99 3/31/00 6/30/00
Technology percent: 15.6% 45.8% 44.2%
Date published: 1/19/00 4/24/00 7/24/00

Thus, even though full holdings were not released until July 19 (for the fund's fiscal half-year ended May 31), the industry data published for the fund showed shareholders that significant changes in the portfolio were underway, and showed it promptly after what turned out to be the turning point in the equity markets. Of course, when a well-known manager retires a fund advisor ordinarily takes pains to discuss the change with investors, and Fidelity also supplied investors with extensive text materials explaining the new manager's investing style promptly when the change was announced. In addition, the manager change attracted substantial interest from fund trackers, including an April 28, 2000 "Morningstar Take" that stated:

"By March 31, manager Bettina Doulton had completely revamped Fidelity Advisor Growth Opportunities Fund. Whereas Vanderheiden, who left at the end of January, had underweighted technology stocks on valuation concerns, Doulton hiked the fund's tech stake to 46% of assets." [,3606,36953-1100107062,00.html, visited February 23, 2003]

Since we provide full holdings to Morningstar on the same schedule that we provide them to shareholders, this article was written without the benefit of a full holdings list. Thus, our policies with respect to holdings disclosure neither concealed the extent of the style change undertaken by the fund from shareholders nor prevented fund trackers from reporting promptly to their customers that the fund had been "completely revamped."

8At the same time, even while accepting preferential holdings information from other fund groups, some fund services express discomfort with the inequality of disclosure that the practice implies. One such entity stated, in comments on this proposal, "we deeply appreciate the willingness of fund families to help us get timely information to our customers, but we believe shareholders should also be able to obtain that information directly from their funds."
9 We submit that 13F reporting is more of a harm than a benefit to mutual fund shareholders. We support the suggestion of the Investment Company Institute that 13F filing requirements be changed from quarterly to semi-annual, and that the lag for 13F reporting be increased from 45 to 60 days.
10 The saying "put enough snapshots together and you have a movie" expresses the risk involved in frequent disclosure, even with a time lag, as well as the benefit of such disclosure to those who want to fully understand a fund's trading techniques.
11 Aware of this practice, most large fund groups restrict access to information about fund asset changes and fund cash flows.
12 Note, however, that these techniques have little value unless the funds involved are large, either in absolute terms or relative to the size of the stocks they own. A small fund's or advisor's trades are unlikely to move stock prices significantly unless the advisor is a poor or careless trader. A large advisor's trades, however, may have a more significant impact on the prices of the stocks they trade, especially if the trades are the result of activities over which the advisor has little control, such as significant cash flows from investors. Thus, there is more opportunity to profit from knowledge of a large fund's or large advisor's holdings than from knowledge of a small funds or advisor's client holdings. This explains the fact that larger advisors tend to be the most protective of the confidentiality of client holdings, while smaller funds and advisors may be more forthcoming.
13 In the case of hedge funds, in which most retail investors cannot legally invest and where public disclosure is not mandated, funds of all sizes practice a high degree of secrecy with respect to their holdings and their trades. This secrecy is motivated by many factors, which may include not only the desire to create a cachet or aura of exclusivity around the products, but also a desire not to reveal trading techniques to competitive advisors or to other traders in the market.
14Funds that disclose holdings daily have appeared on the scene from time to time, but have rarely if ever attracted significant assets.
15The release hypothesized as follows:

" order for "front-running" to significantly decrease investment returns under the proposed quarterly reporting requirements, it appears that the following conditions may have to be present:

  1. To accomplish the goals of the trade that might be front run, the fund manager has limited discretion over the timing of the trade.

  2. The trade occurs during a quarter at the end of which the fund otherwise would not have had to report its portfolio holdings. 

  3. The order is so large that it cannot be reasonably completed within the disclosure window. 

  4. The market is sufficiently illiquid so that large orders may be reasonably expected to have a substantial impact on price. 

  5. When the fund's portfolio is revealed, the size of the remaining order is sufficiently large that it is worth front-running. 

  6. Other traders recognize the front-running opportunity. 

  7. Other traders are willing to assume the risks of trading on the front-running opportunity. 

  8. The fund manager cannot delay the trade without a significant effect on performance.

It appears that front running may be a profitable strategy if all of these conditions hold simultaneously. It also appears that these conditions may rarely be met. If this is correct, the resulting costs of front-running under our proposals should be minimal.

Furthermore, there are two additional potential costs that may be associated with these proposals. First, it has been argued that, given public data about aggregate flows of new cash to funds, more frequent disclosure of portfolio holdings would allow traders to effectively identify the securities in which the fund(s) will transact to accommodate this flow. This may, in turn, provide a potentially profitable front-running strategy to these traders. Second, a requirement for more frequent disclosure may disrupt trades that are made for potential tax-timing advantages. These potential costs are particular cases of front-running, and it appears that they may require the same conditions to hold as those described above, along with additional conditions specific to these strategies."

16 Appropriate exceptions are made for specific purposes, such as disclosure to trading counterparties (for example, disclosing whether a sell order represents a short sale) or to the Commission or other regulators.
17 Though the abbreviated holdings list would save nine pages, the new expense disclosure requirements would add a new page. Consequently, total page count would drop by only eight pages.
18To make this disclosure more useful to investors, we support the recommendation of the Investment Company Institute that the number of expense scenarios proposed to be presented in shareholder reports be reduced from two to one.
19 Eliminating the requirement to mail the annual report would save a smaller amount, since fund groups typically mail prospectuses to existing shareholders each year around the time of the annual report mailing, and often combine the two mailings to save postage.
20 Based on our experience with retail fund shareholders, when investors are asked to elect to receive fund reports electronically, approximately 4% of shareholders elect to do so and 96% do not respond. On the other hand, based on our experience in distributing information to ERISA plan participants, when investors are told that they will receive reports on request only, approximately 4% request reports and 96% do not respond. Of the over 90% of shareholders who do not respond in either case, we believe that most would be better off if their funds were not incurring a significant printing and mailing cost to deliver them information that they have not requested.