February 24, 1997

Joan Conley
Office of the Corporate Secretary
NASD Regulation, Inc.
1735 K Street, N.W.
Washington, D.C. 20006-1500

Re: Solicitation of Comment on the Use of Bond Mutual Fund Risk Ratings in Supplemental Sales Literature
(NASD Notice to Members 96-84)

Dear Ms. Conley:

The Investment Company Institute appreciates the opportunity to submit its views on the use by NASD members and their associated persons of bond mutual fund risk ratings in supplemental sales literature.

The Institute commends NASD Regulation's efforts in seeking ways to enhance information provided to investors concerning bond mutual fund risk characteristics. For the reasons detailed in our submission, we strongly support NASD Regulation's continuing to prohibit the use of risk ratings in fund sales literature. The problems inherent in fund risk ratings are of such nature that they cannot be addressed in accompanying disclosures, or by substantive requirements like those imposed by NASD Regulation on historic performance rankings of mutual funds.

Were NASD Regulation nevertheless to propose to permit the use of risk fund ratings, we believe that specific requirements relating to their use (including, for example, mandatory disclosures or substantive standards with respect to such ratings) should first be submitted to NASD members and the public for comment.

Questions or comments on the Institute's submission on the use of risk ratings in fund sales literature may be directed to the undersigned (at 202/326-5810), or to Craig Tyle (at 202/326-5815) or Amy Lancellotta (at 202/326-5824).

Sincerely,

Paul Schott Stevens
Senior Vice President
General Counsel

Enclosure

cc: Mary Schapiro
President
NASD Regulation, Inc.

R. Clark Hooper
Senior Vice President
Office of Disclosure and Investor Protection
NASD Regulation, Inc.

 

Submission of the Investment Company Institute
in Response to the NASD Regulation, Inc.
Concept Release on the Use of Bond Mutual Fund Risk Ratings
in Supplemental Sales Literature

Table of Contents

I. Introduction

II. Executive Summary

III. Background

IV. Discussion

V. Conclusion

Submission of the Investment Company Institute
in Response to the NASD Regulation's Solicitation of 
Comments on the Use of Bond Mutual Fund Risk Ratings
in Supplemental Sales Literature

I. Introduction

The Investment Company Institute (the "Institute")¹ files this submission in response to the NASD Regulation, Inc. ("NASDR") concept release (the "Release") respecting the use of bond mutual fund risk ratings ("fund risk ratings") in mutual fund supplemental sales literature.² As NASDR is aware, the Institute has long been a strong advocate of efforts to enhance risk disclosure to mutual fund investors. To that end, the Institute has conducted substantial empirical research and developed numerous policy recommendations, including, for example, specific risk disclosure approaches and a new summary disclosure document for fund investors. Thus, the Institute welcomes NASDR's initiative in soliciting comments about how to improve risk information available to investors, and in particular about the use of mutual fund risk ratings in fund sales literature. For the reasons detailed below, the Institute opposes any modification of NASDR's long-standing prohibition on the use of fund risk ratings in supplemental sales literature.³ In our judgment, such risk ratings will not help but harm mutual fund investors.

II. Executive Summary

The Institute strongly opposes any change in the rules of the NASDR to permit the market or risk rating products of commercial ratings services to be included in supplemental sales literature or advertisements of bond mutual funds.

In recent years, Standard & Poor's (S&P), Moody's Investors Services (Moody's), and Fitch Investors Services (Fitch) have developed new and as yet untested rating schemes that purport to gauge the impact that future changes in market conditions will have on bond mutual funds. Each commercial service appears to use a different proprietary method in assigning such ratings. Their schemes attempt to take account of both credit and non-credit risks and to evaluate a litany of different factors, many of them highly subjective in nature.

As the ratings services candidly acknowledge, the bond fund risk ratings they produce in this manner are predictions about a fund's future performance.

These risk ratings, therefore, are wholly unlike the rankings published by Morningstar and Lipper, which only measure a fund's past performance and are not predictive. The methods used for historical rankings are transparent—they can be and are applied equally to all mutual funds and the results of the historical analysis can be easily replicated. Fund risk ratings, on the other hand, require subjective evaluations in predicting the future performance of a fund's changing and actively managed portfolio. Risk ratings also are unlike historical fund rankings in that they are marketed as commercial products. They are not produced independently, but are assigned only to those funds that pay for them and only so long as they do.

Recognizing that fund risk ratings are inherently misleading—i.e., that they imply certainty with respect to a fund's future performance, the NASDR has prohibited their inclusion in a fund's supplemental sales literature. One major commercial rating service, however, has mounted an aggressive lobbying campaign to overturn the NASDR's prohibition and thereby widen the prospective market for its product. It contends that the fund risk ratings it wants to sell would be useful and material information for prospective bond fund investors. The Institute strongly disagrees. In our judgment, such ratings in fact would be hazardous to investors, for a variety of reasons.

Most importantly, risk ratings, by their very nature, are tempting but harmful shortcuts. Without understanding the nature and limitations of such ratings, investors will rely upon them heavily—in particular, for the predictions they make about a fund's future performance. The NASDR's rules properly regard such predictions to be misleading and objectionable per se. Nor are bond fund investors likely to understand that, even apart from their predictive value, risk ratings are fundamentally flawed in that:

• the ratings are misleading insofar as they imply any certainty about a fund's future investment performance;
• the all-in depiction of risk that such ratings purport to provide is illusory, because risk is not a unitary concept nor identical for all investors;
• the ratings are new, are largely untested, are prepared according to no standardized methodology and are inherently subjective;
• the ratings implicitly view risk to be synonymous with short-term volatility, a dubious standard for many investors;
• a fund's risk ratings may become quickly outdated, unbeknownst to those investors who relied on them; and
• the services issuing the ratings are unregulated, they disclaim any liability for the accuracy of their ratings, and in the past such ratings have proved spectacularly incorrect.

The problems inherent in such risk ratings are, in our judgment, not of a nature that can be addressed through accompanying disclosures or through the development of substantive standards, such as those that NASDR has adopted in the case of historical performance rankings. At a minimum, however, were the NASDR to determine to authorize the use of such ratings, extensive disclosure requirements and detailed substantive standards would be vitally important. These would require a formal rulemaking process subject to substantial further factual development regarding the ratings services and their rating processes as well as an opportunity for wide public comment.

Finally, it is important to note that, while the Institute opposes the use of fund risk ratings in sales literature, we have strongly advocated alternative means to improve investor understanding of bond fund risk. Indeed, we have been at the forefront of efforts to inform and educate fund investors and to develop more effective disclosure based solidly on the needs and preferences of mutual fund investors themselves. In our judgment, opening up fund sales literature and possibly advertisements to the competing schemes of commercial ratings services will mislead and confuse investors, not enlighten them.

III. Background

For many years, rating services have analyzed the creditworthiness of corporate and government debt instruments for the purpose of evaluating the likelihood that they would be paid in accordance with their terms. Since 1981, the SEC has permitted disclosure of credit ratings in registration statements, prospectuses and other offering documents.

In the past several years, commercial rating services have attempted to develop analytical models that purport to evaluate the credit and non-credit risks of bond fund portfolios and to market a new bond fund risk rating product. In essence, these fund risk ratings claim to measure the relative sensitivity of bond fund portfolios to short-term interest rate volatility and other market risks such as liquidity risks, concentration risks, call and options risks, currency risks, and risks arising from the use of leverage, hedging and derivative instruments. S&P began offering its Bond Fund Risk Rating System in January 1994. Fitch introduced its Mutual Fund Volatility Ratings in the same approximate time frame. In January 1996, Moody's joined S&P and Fitch in assigning risk ratings to bond funds.

Currently, such risk ratings do not appear in fund prospectuses because the commercial rating services, though professing great confidence in their expertise, adamantly refuse to accept expert liability under Section 11 of the Securities Act of 1933. In addition, the NASDR has not permitted fund risk ratings in supplemental sales literature for fund shares on the grounds that such ratings imply certainty about a fund's future performance, are predictive in nature and therefore are per se misleading. A collateral issue—concerning whether risk ratings could be included in fund advertisements—has not before been seriously considered, because of the requirement in Rule 482 that fund ads may contain only information the "substance of which" appears in a fund's registration statement. Because the rating services would not accept normal expert liability, their ratings have been excluded from fund registration statements and as a result also excluded from Rule 482 advertisements. The recently enacted National Securities Markets Improvement Act of 1996, however, includes a provision that directs the SEC to permit fund advertisements to include information "the substance of which" does not appear in a fund's registration statement. Thus, if the NASDR were to lift its current prohibition, it appears the use of risk ratings would not be limited to only supplemental sales literature, which is accompanied or preceded by a fund's prospectus, but could and likely would be included in newspaper ads, radio and television ads and any other mass media used to market funds. Absent any limitation by NASDR on the use of such risk ratings, they are likely to become ubiquitous and, as such, exert a powerful—and, in our judgment, harmful—influence over retail investors' investment decisions.

S&P has been the leading advocate for permitting fund risk ratings to be included in a fund's supplemental sales literature and has provided the most public details about its system. For this reason, we focus our discussion herein primarily on S&P's system. However, our opposition to the inclusion of fund risk ratings in fund sales literature extends to all such systems and is by no means limited to S&P's system.

A. The Methodology of Assigning Risk Ratings is Novel and Untested, Raising Serious Concerns About the Quality of the Ratings

The proprietary methodologies employed by S&P and other services in assigning bond fund risk ratings are of recent invention and novel at best. Indeed, S&P boasts that its fund risk ratings are "new," "innovative," and "represent different cutting edge technology."⁴ To our knowledge these methodologies have not been fully disclosed to the SEC, NASDR, the industry or the public, and have not been subject to any significant testing. Because of the lack of marketplace experience with fund risk ratings, there is no basis for confidence about the quality of the ratings. As even one ratings service has observed of these new products, "[N]on-credit designations, such as volatility or market risk ratings, are untested, and it is not clear that they will be sufficient to protect investors when conditions change."⁵

The very limited information available on such ratings, including the abbreviated description of S&P's evaluation process that it has made publicly available, only underscores the basis for serious questions about the quality and reliability of such ratings. According to S&P,⁶ upon a fund's request, it will be assigned a risk rating based on an analysis of the fund's portfolio holdings, historical performance, and management. This rating will reflect S&P's assessment of a fund's potential rate of return variability and its analysis of key risks, including:

• interest rate risk
• credit risk
• liquidity risk
• concentration risk
• call and option risk
• currency risk
• the effects of various portfolio strategies such as the use of leverage, hedging and derivative instruments.

In essence, S&P's system purports to provide information regarding a fund's short-term sensitivity to changing market conditions—predominately, but not exclusively, market interest rates—and whether over the near future fund returns can be expected to be relatively stable or relatively volatile.

S&P's fund risk rating system purports to employ a three-pronged approach to analyzing market risk. First, S&P performs an historic risk analysis that measures the variability of past portfolio returns using standard deviation calculations and other measurement techniques. Second, S&P examines current portfolio holdings and evaluates the factors that can create variability in returns (e.g., interest rate risk, prepayment risk, etc.). Finally, S&P performs a management analysis which takes into account management's experience, operating policies, risk preferences, credibility and commitment to policies, and thoroughness of internal controls.

This final step, the analysis of fund management, is particularly important where there are significant differences between the fund's historic risk analysis and its current risk profile. In conducting its analysis of management, S&P meets with fund management and discusses a number of factors that may impact the fund's return, including investment policies and strategies, asset selection, and risk monitoring. According to S&P, the purpose of these discussions is to evaluate "the effectiveness of fund management in implementing a portfolio strategy consistent with its stated investment goals."

Following the review of a fund's management, an S&P employee makes a presentation about the fund to S&P's rating committee. After that presentation, the committee members discuss their opinions concerning the fund's rating. At the conclusion of that discussion, the committee, by vote, determines the final rating to be granted to a fund.

The rating services express their overall opinions as to a fund's market volatility in the form of some alphanumeric or descriptive grade. For example, S&P uses a seven-step scale ("aaa," "aa," "a," "bbb," "bb," "b," and "ccc").⁷ This is basically the same scale that S&P uses when grading the credit risk of individual bonds except that, for its fund risk ratings, lower case letters are used. Funds that have "extremely low sensitivity to changing market conditions" may receive an "aaa" rating. Funds that are "more sensitive to changing marketing conditions than higher rated funds," may receive a "bbb" rating. Funds with a "ccc" rating are "highly speculative with currently identifiable risks." The remaining grades represent various gradations between these ratings.

After issuing a rating, S&P monitors the fund at least monthly "to ensure that changes in portfolio and fund management operating policies do not alter the fund's credit risk or market risk and thereby the accompanying fund rating[s]." S&P also conducts a yearly management update to review any changes that may have occurred.

B. Fund Risk Ratings are Predictive of Fund Performance

NASDR has prohibited fund risk ratings in fund sales literature because such ratings are predictive in nature. These ratings are designed to predict the extent to which market conditions will impact the fund's rate of return. Thus, for example, when S&P rates a fund "aaa", it is predicting that, even if market conditions change, the fund will have "extremely low sensitivity" to those changes and, thus, offer the greatest stability of returns.⁸

Indeed, S&P, in its own words, frankly has admitted that its fund risk ratings are predictions of future investment returns. For example, in the January 17, 1994 edition of Creditweek, an S&P publication, there appears an article entitled "S&P's NEW BOND FUND RISK RATINGS EXPLAINED." The article states:

While most other mutual fund ratings systems derive a rating based solely on past performance, S&P's bond fund risk ratings are forward looking.⁹

In this same vein, the S&P article emphasizes that "[t]he rating will reflect S&P's assessment of a fund's potential rate of return variability and its analysis of key risks." The article continues that "[t]he goals of S&P's analysis are to uncover risk sources in a fund's portfolio holdings and investment strategies and assess the potential impact they may have on a fund's rate of return and net asset value (NAV) variability." The article adds that "[I]nvestors require additional information about the potential risks and returns these securities offer."

Moody's also has characterized its fund risk ratings as predictions of future performance. Thus, in one of Moody's "Special Comments," Moody's states that its fund risk ratings "are designed to provide investors worldwide with a simple to use, forward looking tool for evaluating net asset value (NAV) volatility." Moreover, Moody's adds that its "MR [market risk] ratings are prospective in nature."¹⁰

Not surprisingly, those financial commentators who have examined S&P's and Moody's fund risk ratings also regard them as predictions of future performance. One prominent commentator on financial matters in the popular press has specifically characterized fund risk ratings as efforts by rating services to "predict performance":

Most of the mutual-fund risk ratings now available are calculated from the fund's past performance, according to [S&P's] Managing Director Sanford Bragg. The new approach looks at historical data but also tries to predict performance by studying bond funds' portfolios and putting their investments through a computer model.¹¹

Another commentator has reached a similar conclusion:

The main concern, says Clark Hooper, senior vice-president for disclosure and investor protection at NASD Regulation, is the ‘predictive nature' of such ratings and what inferences investors might make about fund performance from them. A legit complaint? Of course those new ratings are predictive.¹²

Based on the methods used by the commercial risk rating services, the statements that those services have made about their own products, and the observations of independent financial commentators, there is no question that fund risk ratings, pure and simple, are devised and held out to the public as predictions of a fund's future performance.

C. Fund Risk Ratings are Wholly Unlike Historical Performance Rankings

For a number of reasons, fund risk ratings are entirely unlike historical performance rankings compiled by organizations such as Morningstar and Lipper. First, the predictive nature of risk ratings sharply differentiates them from historical performance rankings. Performance rankings purport only to measure how well a fund has performed in the past compared to other funds. Morningstar's publication highlights this focus:

The star rating is neither a predictive measure nor a "buy/sell" recommendation. It is a purely descriptive representation of how well a fund has balanced risk and return in the past.¹³

As Don Philips, the President of Morningstar, has stated, "we can't predict the future. Let's do the best we can to analyze the past."¹⁴

Moreover, performance rankings are strictly based on mathematical formulae that are available in the public domain. As a result, performance rankings are easily constructed for a universe of funds and the results are readily replicated.¹⁵ In contrast, fund risk ratings are proprietary or "black box" evaluations, which take into account numerous subjective and other factors, as detailed above. They cannot be independently constructed on the broad range of funds, nor independently replicated as to any specific fund.

Finally, and very significantly, ranking entities like Morningstar and Lipper are entirely independent of the funds they rank. Their rankings are not sought out by the funds they rank nor do they receive any compensation from those funds. In contrast, fund risk ratings will be assigned by commercial enterprises which have been paid by the funds, and the various rating services will compete with each other for the business of the funds they rate. S&P rates a fund only if the fund's adviser requests a rating, selects S&P rather than a competing commercial rating service, agrees to compensate S&P, and cooperates with it in providing information for analysis. S&P is compensated by the fund not only for initially rating the fund but also for continually monitoring it.¹⁶

This distinction is significant. Under the NASDR rules governing the use of historical performance rankings, funds are specifically precluded from using a performance ranking if they have paid for that ranking.¹⁷ Thus, even if fund risk ratings in all other respects were identical to historical performance rankings, they still could not be used in fund supplemental sales literature.

D. Fund Risk Ratings are Very Different From Credit Risk Ratings and CMO Volatility Ratings

Fund risk ratings are different from credit risk ratings in a number of significant ways. First, assessing the market risks of an entire portfolio involves many more factors than are involved in assessing credit risks. As discussed above in paragraph A, S&P's fund risk ratings purport to take into account a large number of factors apart from creditworthiness. Moreover, the consideration of all of these varied factors necessarily entails significantly more subjectivity than the process for assigning credit risk ratings. Also, credit ratings have been used for many years and, presumably, the credit rating process has achieved a significantly higher degree of standardization among rating services.¹² Furthermore, as discussed in Section IV.A.7. below, the services that issue credit risk ratings must be reviewed and approved by the SEC. No review or approval process currently exists for the commercial services that compile fund risk ratings.

Credit risk ratings and fund risk ratings differ in another significant manner: who decides what constitutes success. A credit rating attempts to "quantify the likelihood that an issuer will be able to comply with the terms of a particular obligation...In the case of a debt securities, including convertible debt securities, a rating is an evaluation of the likelihood that an issuer will be able to make timely interest payments and will be able to repay principal. Similarly, a preferred stock rating assesses the relative security of dividend payments."¹⁹ Thus, in credit risk ratings there is an independently defined concept of success: the likelihood of timely payment of interest and return of principal.

In contrast, the risk associated with fund shares is not related to the failure to fulfill any predetermined obligations, such as to make defined interest or dividend payments, or to repay principal. Instead, it will be the commercial rating services that both define what constitutes "success" (e.g., an "acceptable level" of volatility) and evaluate the likelihood that such "success" will be achieved.

Fund risk ratings are also unlike volatility ratings of CMOs. Those ratings seek to measure the impact changes in interest rates and mortgage prepayment rates have upon a CMO. CMO volatility ratings are derived from a mechanical formula based on price and cash flow volatility and accordingly do not present the same concerns raised by the subjectivity and non-reproducibility of risk ratings. Moreover, unlike a mutual fund, the composition of CMOs are not actively managed and thus there is much less risk that volatility ratings will become "stale." Significantly, volatility ratings of CMOs are not designed to be used by individual investors, but by sophisticated, institutional investors who are in a position to understand the basis of, and limitations inherent in, such ratings.

IV. Discussion

A. NASDR Should Continue to Prohibit the Use of Fund Risk Ratings Because They Would Be a Harmful Shortcut for Investors
Most fundamentally, the Institute opposes any modification of the NASDR's current prohibition on the use of fund risk ratings because we believe that such ratings will become a harmful shortcut for those many bond fund investors who—as the rating services clearly invite them to do—will rely heavily on a fund's assigned rating. As discussed below, this reliance will short-circuit efforts to ensure that retail investors undertake a meaningful examination of a fund's risk characteristics in light of their own investment objectives and time horizon. Moreover, retail investors are likely to have no understanding of the real nature and limitations of a fund's risk rating, including the following: (1) that the ratings are misleading insofar as they imply any certainty about a fund's future investment performance; (2) that the all-in depiction of risk that such ratings purport to provide is illusory, because risk is not a unitary concept nor identical for all investors; (3) that the ratings are new, are largely untested, are prepared according to no standardized methodology and are inherently subjective; (4) that the ratings implicitly view risk to be synonymous with short-term volatility, a dubious standard for many investors; (5) that a fund's risk ratings may become quickly outdated, unbeknownst to those investors who relied on them; and (6) that the services issuing the ratings are unregulated, that they disclaim any liability for the accuracy of their ratings, and that in the past such ratings have proved spectacularly incorrect.

The sheer number and variety of the problems with fund risk ratings demonstrate that they cannot be addressed effectively through supplementary disclosures or other ancillary requirements. For these reasons, we would urge the NASDR, in the strongest possible terms, to continue to prohibit the use of such risk ratings in funds' supplemental sales literature.

1. Investors Will Unduly Rely on Risk Ratings to the Exclusion of Other Risk Information

The Institute is particularly concerned that investors are likely to place undue reliance on fund risk ratings if they are permitted in supplemental sales literature. Investors are likely to regard the rating as the most relevant measure of a fund's risk, eliminating any incentive to read the discussion of the fund's investment risks, which might include information far more relevant to an investment decision. As such, the use of fund risk ratings will frustrate the ongoing efforts of the SEC and the industry to equip investors with understandable and meaningful information on risk.

The Institute's concerns in this regard are well-founded. The propensity of investors to rely on ratings has been demonstrated by the almost blind faith that many investors place in performance rankings, such as those compiled by Morningstar and Lipper. The statistics are enlightening: According to a study prepared by Financial Research Corp., in 1995, 87% of all new money that was invested in stock funds went into funds that received five stars by Morningstar, despite the fact that lower ranked mutual funds would be more appropriate for many investors.²⁰ Even Morningstar itself recognizes that many investors place undue reliance on their rankings. According to Catherine Voss Sanders, an editor at Morningstar, "no matter how many times we say it, some people insist on buying five-star funds without knowing what five stars really means."²¹ As at least one commentator has noted, "For a fund company, getting the top five-star rating from the oracles at Morningstar is better than being blessed by the Pope."²²

Concerns have been expressed in the past about the possibility of even the most sophisticated investors placing too much reliance on ratings. For instance, in 1991, the SEC adopted amendments to its rule governing money market mutual funds, a rule that relies heavily on the credit rating of money market instruments to determine their eligibility for money fund portfolios. The rule nonetheless requires that funds perform their own creditworthiness determination and, in this connection, the SEC noted:

Possession of a certain rating is not a "safe harbor." Where the security is rated, having the requisite NRSRO rating is a necessary but not sufficient condition for investing in the security and cannot be the sole factor considered in determining whether a security has minimal credit risks.²³

From time to time, as problems have arisen unexpectedly with various issuers, the SEC staff has had occasion to reiterate the need for funds to look behind such credit ratings. If there are serious grounds for concern that even sophisticated investors may place undue reliance on ratings, it is a foregone conclusion that individual retail investors will do so.

2. Risk Ratings are Predictive and are Therefore Per Se Misleading

The Institute's concerns that investors will unduly rely on risk ratings are compounded by the fact that they are predictive. NASDR has long recognized the dangers that predictive fund performance information can pose to mutual fund investors. Thus, NASDR rules flatly prohibit the use of any predictive fund performance disclosures. Paragraph (d)(2)(N) of NASD Rule 2210 squarely states that "[i]nvestment results cannot be predicted or projected."²⁴ Market risk ratings are also prohibited because they improperly predict a level of stability in a bond fund's level of return. Rule 2210(d)(1)(B) of the NASD Rules proscribes misleading information in sales literature, specifically, and expressly cautions against statements that suggest certainty with respect to rates of return.²⁵

Fund risk ratings violate NASDR rules because, as NASDR has properly determined, they involve representations about future investment performance and seek to predict or project investment results. It is, of course, true that a risk rating for a particular bond fund does not predict that the fund will achieve a specific rate of return in any particular future year. But far less than that kind of specificity may violate the NASDR rules. The ratings most certainly do purport to predict the extent to which changes in market conditions will affect a rate of return. Thus, for example, when S&P rates a fund "aaa," it is predicting that, even if market conditions change, the fund will have "extremely low sensitivity" to those changes, implying that the fund will offer the greatest stability of returns. An investor in an "aaa" fund certainly will be led to believe that if the fund earned a 10% return last year, it will earn a return of about 10% the following year because, whether or not market conditions change, returns will be "stable." That is precisely what S&P is prognosticating with its "aaa" rating; that is precisely what investors will take the "aaa" to mean; and that is precisely what is disallowed under NASDR rules.²⁶

In light of S&P's own candid admission that its risk ratings are intended to make representations to investors with information about a fund's "potential rate of return," and the fact that others construe S&P's ratings as "tr[ying] to predict performance,"²⁷ it is very hard to understand how the NASDR Investment Companies Committee ("Committee") could have reached any conclusion other than that reflected in its May 26, 1995 letter to S&P. That letter pointed out that S&P's market risk ratings would, "by implication, be predictive of [fund] performance," and, therefore in violation of NASD Rule 2210(d)(2)(N). As reported in that letter, the Committee concluded that S&P's market risk ratings "implie[d] a degree of certainty regarding returns which contradicts a fundamental premise of Article III, Section 2210 of the [NASDR] Rules: that communications with the public address the uncertainty of returns inherent in investment securities." Because these ratings have not changed significantly since the Investment Companies Committee's 1995 decision, we are hard-pressed to understand how a different conclusion could be reached today.

In the section entitled "Arguments in Favor of the Use of Bond Mutual Fund Risk Ratings," the NASDR Release references the safe harbors for forward-looking statements found in SEC Rule 175 under the Securities Act and established by the Private Securities Litigation Reform Act, as well as the Management Discussion and Analysis ("MD&A") section of SEC Regulation S-K. The purpose of these references is to suggest that the NASDR's prohibition is an outdated vestige in conflict with some recent trend of increased disclosure of forward-looking information.

As relates to mutual funds, there is no such trend. To the contrary, in recognition of the unique dangers of predictive statements concerning mutual funds, both the SEC and Congress recently have prohibited investment companies from relying on the safe harbors available to other issuers for predictive statements. Investment companies can not rely on the safe harbor for projections found in Rule 175 under the Securities Act. As the SEC has stated, investment companies are "not covered by rule 175, the safe harbor for projections, because forecasts of the securities markets may pose a significant risk of misleading investors and can quickly become stale."²⁸ Still more recently, Congress has reached the same conclusion. In the Private Securities Litigation Reform Act of 1995, after much deliberation, Congress created a safe harbor for certain forward-looking statements. Notably, however, Congress expressly precluded investment companies from relying on the safe harbor created by that act.²⁹

Finally, in recognition of dangers associated with predictive statements concerning mutual funds, the SEC specifically rejected the notion that mutual funds should make any forward looking statements in their MD&A disclosures. Instead, MD&A disclosures for mutual funds are limited to discussions of past performance. Item 5A to the instructions to Form N-1A "requires funds to explain what happened during the previous fiscal year and why it happened."

3. Market Risk Ratings are Fundamentally Flawed Because Risk Cannot Be Captured in a Single Measure

Market risk ratings are fundamentally flawed because they attempt to capture "risk" in a single measure. S&P characterizes its system as one designed to "evaluate all of the significant risks associated with bond fund investing."³⁰ We believe that no such system of numbers, letters, symbols or stock phrases can adequately gauge or convey the "riskiness" of a bond fund (or any mutual fund) either relatively or absolutely. This conclusion follows from the fact that there simply is no fixed, comprehensive, and universally applicable concept of "risk" to be measured in such a system. As the Director of the SEC's Division of Investment Management, Barry Barbash, observed recently, the staff had reviewed the responses of 3,600 individual investors to the SEC's March 1995 concept release and "[g]iven the variety of risk, we concluded that coming up with a standardized risk measurement would not work."³¹

The SEC staff's conclusion is supported by the results of a ground breaking survey of recent fund investors undertaken by the Institute.³² See Investment Company Institute, "Shareholder Assessment of Risk Disclosure Methods," (Spring 1996) ("ICI Survey"). The ICI Survey makes clear that there are wide differences among investors about what, for them, constitutes "risk" in a fund investment and that they perceive "risk" as a multifaceted, not a unitary, concept.³³ Indeed, only 16% of participants identified one risk concept in their definitions of risk—29% selected two concepts, 25% three and, 30% four or more.³⁴ Moreover, what constitutes risk normally will shift over a person's lifetime and can vary widely depending on the investor's time horizon, goals, financial situation, other portfolio investments, and basic attitude. In the case of an investor concerned about potential short-term loss, money market funds may rightly be characterized as low risk and long-term bond funds as higher risk. By the same token, for an investor defining risk as the potential long-term loss of purchasing power (inflation), aggressive growth funds could be considered low risk and money market funds as high risk.

To further complicate matters, there are also innumerable kinds of risks, unrelated to the investor's own circumstances, that can affect the performance of a bond fund. These include, for example, credit risks; interest rate risks; liquidity risks; currency risks (for foreign bonds); political risks; risks from call or pre-payment provisions; risks from the use of leverage, options and derivatives; risks arising from over concentration (lack of diversification); and operational risks such as unanticipated changes in advisory personnel, changes in sales and redemption patterns that affect portfolio decisions, competitive pressures, and policy changes in the SEC, NASD, Treasury Department, Department of Labor, or various state and local governments.

Fund risk ratings obviously cannot take into account all of these types of risk in a single measure, useful to and useable by all investors. For this reason, it would appear that the commercially marketed risk ratings predominantly measure only short-term volatility. Most fund investors, however, will not be in a position to appreciate that fact or its implications for them. Indeed, we believe it would be extraordinarily difficult for most bond fund purchasers to have anything but the vaguest idea of what the risk ratings actually purport to measure. Consider, for example, how S&P construes its "aa" and "bbb" ratings:

Funds with an "aa" rating have "low sensitivity to changing market conditions"

Funds with a "bbb" rating are "more sensitive to changing market conditions than higher rated funds"

It is unrealistic to believe that investors will comprehend what jumble of considerations is encompassed within the phrase "sensitivity to changing market conditions" or what the rating scale is meant to capture, much less understand why an "aa" fund may be better for their investment portfolio than a "bbb" fund, or vice-versa.

4. Fund Risk Ratings Lack Uniformity; Their Subjectivity and the Absence of Common Methodologies Will Only Confuse Investors

Fund risk ratings will confuse investors because they will lack uniformity. Rating services do not employ common methodologies or follow common guidelines in preparing them; instead they are characterized by the high degree of subjectivity that their varying methodologies appear to reflect. Moreover, the differences among the rating systems themselves will generate great confusion. For example, as noted, S&P uses a seven-step scale ranging from "aaa" to "ccc." Fitch, on the other hand, uses a ten-step scale ranging from "V-1" to "V-10." Other rating services can, and presumably will, use entirely different rating symbols to designate different degrees of market risk.

The lack of common methodology and terminology is not surprising. Fund risk ratings are based on proprietary models and methodologies that involve extensive amounts of subjective judgment and analysis. There are no guidelines either imposed or agreed upon that apply to the process by which rating services assign fund risk ratings. There are no common risk designations to express the ratings of the different services in a uniform way. Accordingly, these ratings can vary markedly, even as to funds that have essentially the same portfolios.

The potential for investor confusion is obvious, especially when investors try to compare the ostensible risks of one bond fund with another fund rated by a different commercial rating service. An investor may know, for example, that S&P has given ABC Bond Fund a "bb" rating and that Fitch has given XYZ Bond Fund a "V-5" rating. But how is the investor to know whether these are equivalent ratings or whether these ratings purport to measure the same risks? The lack of common methodology and the importance of subjective judgment in formulating risk ratings allows investors to do little more than make guesses. Confusion will only increase with the proliferation of competing services offering to rate bond funds.

The potential for wide variances in fund risk ratings is far greater than in credit risk ratings in light of the fact that market risk ratings are novel and are entirely untested. By contrast, credit ratings have been used for many years and the credit rating process has had an opportunity to achieve a significantly higher degree of standardization among the rating services. Even so, despite the greater standardization of credit risk ratings, staff members of the Federal Reserve Bank of New York have observed broad differences in the credit ratings of various services and reported that more than 50% of the time ratings of corporate bonds did not agree.³⁵

Assessing the market risks of an entire portfolio involves many more factors than are involved in assessing the credit risks of a single bond. As noted above, S&P's fund risk ratings take into account numerous quantitative and qualitative elements.³⁶ Evaluating all of these disparate risk factors, singly and in combination, involves many subjective judgments. For these reasons, even if different rating services consider all of the same risk factors (which itself is far from clear), it is hard to see how they are going to reach the same rating judgments—and harder still to see how bond fund investors can help being confused by the risk ratings of multiple services rating potentially hundreds of different bond funds. Fund investors will be at a total loss to know what is being rated, what the ratings mean, or how they compare with the ratings that different services give different funds.

Given the potential for wide variations in the ratings assigned by multiple competing rating services, the likelihood of rate shopping is great. Ratings are paid for by the funds that are rated. If ratings are permitted, competitive pressures may prompt funds to buy them. Understandably, funds will buy the most favorable rating offered by the competing rating services. Accordingly, competitive pressures on the rating services will lead to the kind of "rate shopping" that will be so destructive to the integrity of the ratings process. Indeed, as one rating service has candidly acknowledged, this is precisely what has happened in structured market transactions, particularly in the mortgage-backed securities market.³⁷ In its letter to the SEC, Duff & Phelps, a leading rating service, describes two kinds of rate shopping in the structured finance market:

Standard rate shopping. This occurs when an issuer approaches a number of rating services to determine which will provide the most favorable requirements for the rating.

Abusive rate shopping. This involves a rating service making a material departure from its standards in order to increase or protect its market share.

Moody's also has complained that rate shopping has been encouraged by the SEC's regulatory reliance on Nationally Recognized Statistical Ratings Organizations ("NRSROs").³⁸ Thus, in a letter to the SEC, Moody's states that "[b]y giving regulatory benefits to certain securities with the requisite NRSRO ratings and designating multiple NRSROs, the SEC has unintentionally encouraged issuers to shop among them for the most favorable rating."³⁹ This, according to Moody's, "places economic pressure on rating agencies to erode analytic standards in order to gain, or to avoid losing, market share."

The rating services themselves candidly acknowledge these abuses, which continue to plague the relatively established process for credit ratings of individual debt instruments. There is no reason to believe that the same rate-shopping abuses will not happen with commercially marketed bond fund risk ratings—and undermine the integrity of fund risk ratings if they were allowed to be included in fund sales literature. The problem, of course, would become even more serious as the number of rating services increases.

5. Fund Risk Ratings Falsely Imply That "Risk" And Short-Term Volatility Are Equivalent Concepts

A basic premise underlying fund risk ratings is that "bond fund investors have a strong interest in the sensitivity of the current values of their investments to changing market conditions."⁴⁰ S&P defines "market risk" as "a fund's sensitivity to changing market conditions."⁴¹ In fact, it would appear that the S&P rating method principally measures short-term market risk—i.e., whether investment returns will be more or less volatile in the short term, depending upon various market conditions including, most previously, rising and falling interest rates. S&P predicts, for example, that a fund receiving an "aaa" rating from S&P will have highly stable future returns whereas a fund receiving, say, a "bb" rating will experience significant short-term volatility.

For most investors, however, such short-term volatility neither is, nor should be, the most important risk factor in deciding which bond fund to purchase. Based on the ICI Survey, the vast majority of investors tend to view risk from a more long-term perspective. When those bond fund investors participating in the ICI Survey were asked to describe what "risk" means to them, only 7% cited the concept of investment volatility. Only 1% of bond fund investors said their primary investment strategy was to avoid short-term risk.⁴² The ICI Survey likewise confirms that investors tend to buy funds for the long-term, not as short-term trading vehicles. According to the survey, over 60% of bond fund investors surveyed had a risk time horizon of 6 years or more and only 4% had a risk time horizon of less than one year.⁴³ The estimated median horizon for all investors was 8 years.44

For these long-term investors—constituting by far the majority of fund investors—the rational concern is whether the fund will provide an adequate rate of return over the investor's time horizon, not whether, in the process, short-term returns will fluctuate around some postulated "mean" (e.g., the fund's own historical rate of return or some market index). An investor with a 6 or 8 year time horizon is obviously better off with a fund that has an annualized return of 7%, although experiencing a fair amount of short-term volatility, than a fund that returns 3% per year with little deviation from that figure. Because fund risk ratings are constructed principally to gauge short-term volatility, however, the second of the two funds in the foregoing example presumably would receive the better (i.e., more favorable) risk rating.

Particularly those less sophisticated investors are likely to be misled, and to take a fund risk rating as a depiction of the risk most significant to them, when such in fact is not the case. It is readily predictable that bond fund investors will assume, from their experience in other contexts, that an "aaa" risk rating means "superior," and make their investment decisions accordingly. Indeed, in the context of credit ratings, a triple-A rating really does mean "superior." It would only be natural, therefore, for investors to draw the same conclusion with respect to fund risk ratings.

In selecting a bond fund portfolio investment, however, an "aaa" fund risk rating is by no means necessarily the superior choice. An "aaa" rating, in all likelihood, will go to those funds that invest in short-term bonds and pay the lowest yields. While such funds may be the best choice for an investor with a short time horizon who does not wish to risk selling fund shares in a down market, this is often precisely the wrong choice for many investors, particularly those who are saving for retirement, college education, or other long-term goals.⁴⁵ Moreover, by encouraging investors to purchase funds with high fund risk ratings, the very real possibility exists that investors will fail to diversify the categories of mutual funds in their investment portfolios. The lack of such diversification will expose these investors to greater risks.

6. Fund Risk Ratings May Become Quickly Outdated But Investors Are Unlikely To Become Aware of Changes In Such Ratings

Fund risk ratings are constructed in light of a fund's portfolio holdings. Bond funds, however, are actively managed to respond to constantly changing market conditions and the fund's sales and redemption levels, which may vary greatly day to day. Investment methods, portfolio composition, and risk control techniques obviously can and do change. Because portfolio managers can easily alter the risk exposures of a portfolio, fund risk ratings can quickly fall out of date, perhaps even before promotional materials containing the ratings have been distributed to investors. In order to overcome the problem of "staleness," rating services may seek to update their ratings periodically.⁴⁶ Nonetheless, the nature and effectiveness of the rating services' ongoing surveillance of a fund, and the frequency and speed with which they conduct reviews and update assigned ratings, is at best unclear. Presumably, the monitoring and updating practices of different rating services will vary widely. Even if adequate updating procedures are in place, it is doubtful that rating changes will come to the attention of investors who relied on the ratings when purchasing their fund shares but do not follow the financial media sufficiently to learn of any updates or changes that may have been made to the ratings.

7. Fund Risk Ratings May Be Harmful Because Of The Lack Of Regulation Or Accountability Of The Services That Issue Them

The Institute's opposition to the use of fund risk ratings also stems from serious concerns about the commercial services that issue those ratings. These concerns flow from the fact that (1) there are no barriers to the entry of new services that may provide fund risk ratings and no standards for such entities; (2) rating services are subject to minimal government regulation; and (3) rating services are subject to limited legal liability for their ratings.

a. There Are No Regulatory Barriers To Prevent New Commercial Rating Services From Issuing Fund Risk Ratings
Although only S&P, Moody's, & Fitch currently prepare fund risk ratings, there are no regulatory barriers to prevent other services from issuing those ratings. Indeed, unlike services that provide credit ratings, there is not even a requirement that the entities that provide fund risk ratings be reviewed by the SEC or any other regulator. Prominent commercial rating services such as S&P, Moody's and Fitch, have been designated by the SEC as "nationally recognized statistical rating organizations" ("NRSROs") for their issuance of credit ratings. The NRSRO designation is used in the federal securities laws, and rules and regulations thereunder, as a basis for distinguishing between high quality securities and those that are unrated or rated below investment grade.

Currently, any rating service that wishes to be designated as an NRSRO for purposes of the federal securities laws must request the staff of the SEC to issue no-action relief to that effect. If the organization's structure and debt rating process, among other things, satisfy criteria the staff finds generally necessary for recognition of the entity as an NRSRO, the staff will issue a no-action letter stating that it will not recommend enforcement action to the SEC if broker-dealers consider the particular rating service an NRSRO for purposes of the SEC's net capital rule.⁴⁷

The criteria for designating a rating service as an NRSRO were developed solely in the context of credit ratings for debt securities. Permitting NRSROs to issue fund risk ratings for use in fund sales literature should, at the very least, require the NRSROs to demonstrate any qualifications they have for issuing market risk ratings to the SEC staff. Given the experimental nature of such ratings and their enormous complexity compared to credit ratings, we are at a loss to know how the SEC staff could meaningfully evaluate the qualifications of a NRSRO to issue fund risk ratings.

Nevertheless, should NASDR reverse its current prohibition against the use of risk ratings in supplemental sales literature, then the SEC will be required to adopt an entirely new scheme of regulation. Otherwise, there would be no barriers to the entry of new, and possibly bogus, rating services into the business of providing risk ratings.

b. Rating Services are Subject to Little or No Regulation or Disclosure Requirements
As former Commissioners Mary Schapiro and Richard Roberts correctly observed while at the SEC, there is an "absence of any meaningful authority of the Commission to regulate these [rating] entities" and "compared with the role that rating agencies play in the capital markets, the Commission receives relatively little information about these entities and their operations."⁴⁸ Although NRSROs are required to register with the SEC as investment advisers, "the investment adviser registration system applies awkwardly at best" to them.⁴⁹ These and similar considerations caused Commissioners Schapiro and Roberts to recommend legislation that would require rating services to register with the SEC and provide the SEC with authority to promulgate rules governing these entities. Such legislation, however, has been vigorously opposed by the rating services and has never been enacted.

c. Rating Services are Subject to Very Limited Liability for the Ratings They Issue
In addition to opposing anything but minimal government regulation, the rating services also strenuously resist any legal accountability for their ratings. As noted above, fund risk ratings are not used in prospectuses because: (1) such ratings are expert opinions for purposes of Section 7 of the Securities Act; (2) expert opinions can be used in registration statements and prospectuses only if the expert provides a "consent" to the use of its opinion; and (3) rating services refuse to provide such consents because it would expose them to liability under Section 11 of the Securities Act.⁵⁰ Moreover, rating services maintain that they are members of the "media" who are providing their "opinions," and thus claim that they are not even liable to investors who may have relied on their negligent misrepresentations. Instead, they assert they can only be liable if their conduct can be said to have been "reckless."⁵¹

In short, the commercial rating services want to have their cake and eat it too. They hold themselves out as having great expertise in assessing fund risk ratings, and they seek to exploit such expertise in the securities marketplace among retail investors. Nonetheless, they reject the legal accountability normally demanded of such experts under Section 11. They seek to have their ratings freely used in supplemental sales literature, and presumably fund advertising as well, with all the attendant risks for retail investors, but they eschew any meaningful SEC oversight and even disclaim any requirement to exercise ordinary care in the ratings process.

We believe that the NASDR should not allow the rating services to have it both ways. Conscious policy decisions have been made that preclude the dissemination of market risk ratings in mutual fund prospectuses so long as the services creating those ratings are unwilling to consent to their use and assume the risk of potential Section 11 liability. Those decisions would be subverted if the rating services were able to market their ratings to investors via supplemental sales materials or fund advertising, when they have refused to accept the accountability that would give them front door access through prospectus disclosure.

d. The Rating Services' Assessments of the Orange County Investment Pool Highlights the Dangers of Fund Risk Ratings
The recent experience with the Orange County Investment Pool ("OCIP") illustrates many of the Institute's concerns about the difficulty of reliably rating an entire portfolio of securities for market risk or predicting its future performance. The OCIP pooled the County's monies with monies of other municipal government agencies for appropriate investment management. Rather than managing the OCIP conservatively, as was his statutory duty, Orange County Treasurer Robert Citron leveraged the proceeds of short-term borrowings and other monies and invested in longer term securities as well as complex, highly volatile derivatives. As a result, when interest rates rose, the value of the OCIP tumbled. Orange County lost $1.8 billion, and, in December 1994, filed a Chapter 9 bankruptcy petition, the largest municipal bankruptcy in history.

Throughout 1993 and 1994, both S&P and Moody's gave Orange County notes their highest short-term rating. They did so only after an extensive review of the OCIP. Indeed, as Diane Brosen, an S&P director, stated: "We probably have put the Orange County Pool under more scrutiny than any other investment pool."⁵² Yet this review apparently failed to address the sensitivity of the OCIP to increases in interest rates. Based upon their self-described exhaustive analysis, S&P and Moody's gave their highest ratings (SP-1+ and MIG-1, respectively) to new debt offerings that S&P and Moody's knew would be invested in the OCIP. The SP-1 rating signifies: "very strong or strong capacity to pay principal and interest." The addition of a "+" to the SP-1 rating signifies "issues determined to possess overwhelming safety characteristics." Moreover, on April 19, 1994, the Los Angeles Times reported S&P as stating: "We don't have any major concerns with the Orange County portfolio."⁵³ Similarly, on June 2, 1994, the Moody's Municipal Credit Report described the OCIP's financial condition as follows: "The pool is well-monitored and managed to provide the timely availability of liquidity in the event of market fluctuations."

As late as September 1994, when both S&P and Moody's undertook a rigorous review of the OCIP in connection with the issuance of taxable pension notes for which the OCIP acted as a liquidity source, the rating services reaffirmed their confidence in the OCIP by giving the pension notes their highest ratings. In December 1994, a mere three months later, Orange County filed for bankruptcy.

It is indisputable that S&P and Moody's spectacularly misjudged the safety and stability of the OCIP. Like a bond mutual fund, OCIP was a pooled investment vehicle. The events in question occurred, at least in substantial part, after S&P had already introduced its bond fund risk ratings and purportedly had developed its "cutting edge" methodologies for assessing the safety and stability of pooled investment vehicles. The very fact that two of the most prominent rating services could make such egregious misjudgments in evaluating the interest rate risk of a pooled investment vehicle underscores the problems inherent in allowing their risk ratings to be disseminated in sales literature and possibly advertising to bond fund investors.

8. The Inherent Flaws in Fund Risk Ratings Cannot Be Readily Cured Through Additional Disclosure or Other Requirements

By their very nature, fund risk ratings are designed and marketed as an inviting "short-cut" for investors. This is why the ratings process, at its end point, is reduced to some shorthand designation—e.g., a scale consisting of alpha-numeric symbols or stock phrases. Because of the numerous troublesome issues that this process poses, we do not believe that any accompanying disclosure or other substantive requirements will suffice to protect investors. For example, any disclosure seeking to address these issues would have to inform investors that: (1) risk ratings are predictive, and strongly caution that future results may vary substantially from what is predicted; (2) risk ratings only measure certain types of risk, which may not be relevant in the case of any particular investor; (3) risk ratings are based on subjective analysis, and different rating services may reach different conclusions; (4) different rating services use different ratings methodologies and their ratings accordingly may not be comparable; (5) risk ratings are new, untested and different from credit ratings; (6) risk ratings may become outdated and, accordingly, investors should not rely on the ratings' future applicability to their fund; (7) rating services are paid to issue the funds' risk ratings; (8) rating services are subject to little or no regulatory oversight, particularly with respect to their risk ratings; and (9) the rating services disclaim any and all liability for the ratings.

The extent of the required disclosure in our judgment would be formidable—it would not be susceptible to a concise presentation appropriate for advertising or sales literature. More likely, it would take the form of lengthy "boilerplate"—the kind most likely to be ignored while investors focus solely on the shorthand of the assigned rating itself.

If, however, the NASDR concludes that it should permit the use of fund risk ratings in sales literature, we believe it is incumbent upon the NASDR to adopt rules assuring sufficient accompanying disclosures, along the lines set forth above. Moreover, the NASDR also should adopt substantive requirements governing the use of these ratings. This is, after all, required in the case of performance rankings, which, for the reasons noted above, raise far fewer investor protection concerns. In addition to extensive disclosure requirements, NASD Conduct Rule

IM-2210-3 strictly regulates the time periods for which performance rankings can be reported, prohibits rankings based on certain categories, and (of special note) prohibits the use of rankings that have been procured by an investment company. At the very least, it would be necessary for the NASDR to require fund risk ratings to be current as to some specified time period, to require that they be updated with some specified regularity, and to impose some standards on the entities issuing the ratings or, at the very least, independently review the methodology employed by the entities.⁵⁴

Most importantly, because of the highly detailed nature of these requirements, we strongly believe that they should be adopted by the NASDR only after specific proposals are issued to the NASD membership and the public for comment. Such careful and deliberate consideration is essential if the NASDR is going to rely on disclosure and substantive standards to address the numerous hazards inherent in fund risk ratings.

B. Other Means of Communicating Bond Fund Risks are Far More Appropriate Than Risk Ratings

S&P has charged that "[t]here appears to be great industry reluctance to disclose information about the risk characteristics of mutual funds."⁵⁵ That simply is not true. The Institute is committed to providing investors with information that they will be able to use appropriately and confidently in assessing a fund's risk characteristics.⁵⁶ As discussed below, Institute research demonstrates that investors do not favor and are not as confident of their ability to use risk information presented in the form of standardized measurements; instead, investors prefer simple narrative or graphic presentations of risk information.⁵⁷

The ICI Survey clearly indicates that mutual fund investors are concerned and seek to inform themselves about the risks of mutual fund investing. Ninety-seven percent of the recent fund investors who were participants in the survey agreed that mutual fund investing requires accepting some degree of risk.⁵⁸ Eighty-four percent indicated their willingness to take at least average risk to achieve average gain.⁵⁹ Risk information was reviewed by a much greater proportion of recent fund investors (seven in ten) than was information on any other aspect of their recent fund purchase except for performance. The survey also indicated that these investors obtained information on fund risks from numerous sources and in a variety of forms.

The picture that emerges from the Institute's research is that investors currently are availing themselves of a wide variety of tools that exist in order to assess the risk of mutual fund investing. At the same time, however, significant efforts have been made to identify even more effective means for informing investors on this critically important subject. The purpose of the ICI survey was to have fund investors themselves assess the effectiveness of competing risk disclosure methods—including "plain English" narrative descriptions of risk, a graphic depiction of the variability over ten years of a fund's total return, and three different standardized, quantitative risk measurements (standard deviation, beta and duration). Investors were asked to assess whether the competing disclosure approaches simplified risk analysis, appeared dependable across all market conditions, and seemed suited for investors like themselves. Finally, investors were asked how confident they would be in using the various methods to assess the risks of a fund and which of the methods they preferred.

Based on these criteria, investors expressed significantly greater confidence in and a stronger preference for narrative and graphic presentation of risk information as compared to any of the standardized measurements. Significantly, even those investors who had previous experience in using such measurements preferred narrative and graphic presentations of risk information. Those investors who had no prior familiarity with standard deviation, beta or duration, however, seemed to have unjustified confidence in their ability to use these measurements effectively after only a brief initial exposure—a fact that confirms our belief that investors will unduly rely on risk ratings without fully understanding their limitations.

Based on the Institute's research into the needs and preferences of investors themselves, we believe that there are a number of alternative methods of risk disclosure that would be more appropriate than the inclusion of fund risk ratings in supplemental sales literature. These measures were recommended to the SEC in response to its solicitation for comment on ways to improve mutual fund risk disclosure,⁶⁰ and the Institute continues to urge that the SEC take early action on them. In its letter, the Institute made the following recommendations:

First, narrative disclosure should be improved by requiring that prospectuses focus on the overall risks of a fund portfolio rather than on the individual securities held by the fund. The Institute survey confirms that narrative disclosure significantly assists investor understanding of mutual fund risk, and that further refinement of narrative disclosure requirements would benefit many investors.⁶¹ In recognition of the importance of enhanced narrative disclosures, the SEC recently issued its "Plain English Disclosure" release.⁶² That release stresses the importance of presenting risk information in a clear and explanative format.

Second, fund prospectuses should be required to include a ten year annual total return bar graph, together with a presentation of the fund's average annual returns over one, five and ten year periods. The Institute survey indicates that such graphic presentation of the variability of a fund's returns would significantly assist investors in their consideration of mutual fund risk, and would provide information that investors could readily understand and properly employ both to compare the risks of several funds and to evaluate the risks of a single fund.⁶³

Third, all funds that hold themselves out as having a stated maturity policy and hence some specific interest rate sensitivity (such as in their name or marketing materials) should be required to have a commensurate portfolio duration policy, at least under normal market conditions.

We also believe that a "profile prospectus" of the type the SEC is presently considering would be a valuable tool for investors in making investment decisions and comparing the risks of one fund with another.⁶⁴ A profile prospectus can provide the investor both with a narrative description of the principal risks facing the fund as well as a graphic demonstration of fund volatility, thereby making it clear to the investor that the fund can lose money as well as make it.

V. Conclusion

As the foregoing discussion demonstrates, bond fund risk ratings are inherently flawed in several significant respects—they are predictive; they purport to reduce risk to a single measure that focuses on short-term volatility; they are subjectively determined; they may quickly become outdated or stale; and they are subject to strikingly little regulatory oversight. All of this is compounded by the fact that there is no reason to doubt—and every reason to believe—that they will be relied upon to an inappropriate degree by investors, especially the less sophisticated ones to whom bond fund supplemental sales literature is directed. The NASDR would be unwise—in our judgment, even irresponsible—to overturn years of precedent to permit the use of these ratings in sales literature.

ENDNOTES