Remarks at XBRL International Conference
General Counsel, Investment Company Institute
December 4, 2007
Vancouver, British Columbia
Thank you for the introduction and good morning to you all. I’m truly pleased to be here and to be able to share with you my perspective on XBRL’s utility to the mutual fund industry.
This speech represents a number of firsts for me: it’s my first visit to Vancouver and it’s my first speech as General Counsel of the Investment Company Institute. It’s also the first time I’ve spoken to an audience made up almost entirely of techies and accountants. So, as an attorney, and a “twelve o’clock flasher” at that, I can’t help but feel as though I’ve shown up for a tennis match carrying a set of golf clubs.
But I’ll forge bravely ahead with my task, which is to explain how I think my industry—the mutual fund industry—will use XBRL. I expect we’ll use it quite differently from the way others use it, but I think we’re on the edge of something that could be quite exciting. But first, to provide context for my remarks, let me fill you in briefly on ICI and the industry it represents.
ICI is the association of American investment companies—mutual funds. We’re based in Washington, DC, and we represent an industry of over 15,000 funds, reaching almost 90 million shareholders, with U.S. $12.5 trillion invested. Like most trade organizations, ICI’s primary job is to represent our industry before the Congress and the federal government in support of public policies that serve the interests of our members and their shareholders.
One public policy change we’ve been advocating — since the era of Windows 3.1 — is better, clearer disclosure to fund shareholders of basic information about their investments.
Every American mutual fund has a prospectus, which contains a wealth of information about the fund, all of which is painstakingly required by rules and regulations. This information includes the fund’s investment strategies, objectives, and risks; what you’ll pay in fees if you invest; the fund’s performance over time, and that—surprise!—you may lose money if you buy shares.
Prospectuses also contain lots of other information that tends to be less important to most investors—often 50 pages’ worth or more! Frankly, they’re long, legalistic, and boring, and our research shows that hardly anybody reads them. They were described in one focus group as “oppressive,” “daunting,” and my personal favorite, “a lot of nuthin’.” I’m just curious, has anyone here read a mutual fund prospectus?
For years, we’ve encouraged the U.S. Securities and Exchange Commission, which regulates our industry, to provide important information to investors in a format that they may actually read.
And, days before our American Thanksgiving holiday, we got something to be thankful for. The SEC proposed a short-form, or “summary,” prospectus that would give investors the key information they need—in only 3–4 pages. All of the additional information contained in that 50-page behemoth would be made available on the Internet, or in paper upon request. Under this proposal, funds could provide investors with important information in a usable format, rather than sending a daunting book that usually went right to the circular file. And yet, thanks to the Internet, there would be no reduction in the amount of available information.
One of the SEC’s main objectives with the summary prospectus is to make it easier for shareholders to compare funds. They’ll find the same valuable information in the same place in every summary, so they’ll spend less time searching.
And here’s where XBRL comes in. The SEC, led by Chairman Chris Cox, doesn’t want to stop with a short-form prospectus, even though it will be tremendously helpful to investors. Well before the SEC formally proposed the summary prospectus concept, Chairman Cox enlisted ICI’s help in harnessing the power of XBRL to provide another way for investors to compare key information about mutual funds—without spreading prospectuses large or small—all over the kitchen table and marking them up with a hi-liter. In theory, with XBRL, an investor could retrieve information about performance, or fees, or objectives, for any number of funds, and view just that information.
While the Chairman’s vision had extraordinary implications, getting there was not going to be easy. Mutual fund investors aren’t all that interested in the fund’s financial statements. They want to know about how a fund invests its assets—its philosophy for creating wealth, so to speak, and this typically is disclosed in text and charts and graphs. As far as we know, other financial services organizations have used XBRL only for tagging financial statements—an entirely different animal.
To help transform a vision into a reality, we focused on the information in the first section of a fund’s full prospectus, known as the risk/return summary. This information will also comprise the key elements of the summary prospectus. But here was the challenge: This section of the prospectus, as I mentioned earlier, is largely displayed in text, charts, and graphs, as well as in numbers. Our task—the first of its kind—was to develop a taxonomy for information that is presented in these diverse forms. And how this information is presented is governed by a rigorous set of regulations—prescribing the order and presentation of the content, and sometimes even the words used.
To create the taxonomy, we convened a working group that drew upon the expertise of people involved in every stage of the production and use of fund prospectuses—fund sponsors and administrators, information aggregators, regulators, investor advocates, and technology providers—maybe even some of you were involved in this project. We couldn’t have done it without them and without PricewaterhouseCoopers, which provided valuable technical support.
It will come as no surprise when I say that our working group immediately confronted a number of daunting questions.
As I mentioned, a key component of mutual fund disclosure is the fund’s investment strategies. This information is typically provided in a paragraph or two of narrative text, and many funds use it to explain how they go about making investment decisions. How best to tag this information, which is intended to tell a story about the fund’s investment style and strategies? One solution, say at this end of the spectrum, was simply to put a single label on the narrative “strategy” section for each fund. That way, an investor who looks for the “investment strategies” tag could extract and compare strategies of various funds. But how does that approach help an investor narrow the field, so that she can look at just bond funds, or funds invested in large-cap stocks, or funds pursuing a growth strategy?
Over here at the opposite end of the spectrum, we could create a comprehensive list of investment strategies, and allow funds to select those that best describe their investment process. We started to go down that path—in fact, we reviewed about 100 prospectuses and made a list of all of the investment strategies they contained in them. In doing that, we noticed a few things. First, the list got VERY long. So long, that we doubted that it would be useful—it would just be too overwhelming.
More importantly, there were categories that had substantial overlap, but were not identical. Take, for example, “capital growth” and “long term capital growth.” Or a great example: “global securities” (which includes U.S. issuers) and “international securities” (which doesn’t). An investor looking for a certain type of fund might miss out, because the differences in categorization can be subtle.
An appealing solution would seem to exist somewhere in the middle: what if we created a limited list of broad strategies, maybe 10 or 12, such as “growth,” “income,” and “fixed income”? That had a lot of appeal, and it’s similar to the process used by some well-known fund information service providers, such as Morningstar.
But while these “style boxes” are popular with investors, many fund companies find them narrow and arbitrary. For one thing, one person’s growth fund might be another person’s value fund—there are no fixed definitions for these classifications. Our working group had lengthy debates about where to draw the lines and come up with a common set of definitions — and there wasn’t much agreement.
More importantly, while independent information service providers can impose any definitions they like, our taxonomy was intended to be used for regulatory filings, which are subject to the federal securities laws. And as a lawyer, I can assure you that there is a big legal difference between a third party assigning a label to a fund and the fund itself reporting to a regulator under that label.
We’re still wrestling with this issue, but rather than let the perfect get in the way of getting it done, we’ve chosen not to require any strategy tags—even though we recognize that our decision may make the tagging less precise, and therefore less useful for investors and other users.
Here’s another big-picture issue: is it necessarily beneficial to let investors extract a single paragraph or section of narrative, without the section before or after it? As you all know, one benefit of XBRL is that it allows users to isolate individual pieces of data. Great for financials and similar information, but some funds use their prospectus to tell a narrative story. They explain their objectives, risks, and strategies in a way that strings together the benefits and risks of the fund. Taking one paragraph without the others, a user might lose important context. And from an investor protection standpoint, always on the forefront of our minds, how can we be sure that investors will see certain cautionary information if they’re going to extract only isolated parts of the document?
That’s just a sample of the questions that came up as the working group developed the draft taxonomy, and honestly, we’re not sure that we have always arrived at the right answers.
Fortunately, though, we have a chance to find out. In August, the SEC adopted a rule allowing funds to test the taxonomy voluntarily. Nine fund complexes so far have submitted their risk/return summaries in XBRL format. Not a huge sample, I’ll grant you, but it’s a start, and we hope that, in time, it will help us answer some of our questions and refine our solutions.
We spoke to some of these firms about their experiences so far—and let me remind you that “so far” is all of three and a half months. But even though we’ve barely dipped our toe in the water, we’ve heard some consistent remarks. One common theme was that the process of tagging individual risk/return summaries wasn’t difficult, but it was tedious and time-consuming. There simply aren’t good tools available yet to make the process easier. Now, this may be a chicken and egg problem—good tools for tagging data will emerge when more funds are demanding better tools. We hope that some of you might help the industry out with that.
More broadly, we heard concerns that were reminiscent of the issues I described earlier. For example, there is still no consensus on whether we should try to describe funds using a common vocabulary, which would help investors compare similar funds, or whether it is more important to allow funds the flexibility to describe themselves in their own words.
Nor is there consensus on the benefits of allowing investors to pull certain data points in isolation. Funds with very simple strategies, such as those that track a common index, may not lose much in translation—there’s less of a story to tell. But most funds are actively managed, and some of those funds have expressed valid concerns about investors losing context if they can isolate individual data points. One member described this to me as being akin to listing all of the ingredients for a recipe, but without providing instructions on how to put them together.
Despite these concerns, we are very aware of XBRL’s potential. And, as I said, we’ve only been at this a very short while. If we can come to some agreement about how to deal with the thorny issues I’ve described, we—the Institute and the fund complexes we represent—recognize that XBRL could be very powerful.
For one thing, it is an incredible data management tool. With XBRL, fund complexes, some of which have 100 or more funds (I think Fidelity is well over the 300 mark), could maintain a centralized data bank that they could use for multiple purposes, including generating internal reports, creating marketing materials, or quickly assembling disclosure documents and other regulatory filings, a process that now requires a great deal of manual labor and, therefore cost. It also could eliminate the need to create and maintain duplicative data in more than one place within a single organization, and mitigate problems arising after merger transactions, when the legacy systems of the combined organization can’t talk to each other.
XBRL could also help ensure that fund investors get accurate information from third parties, such as the on-line services that offer fund ratings or other comparisons. Did you know that today, those services manually copy data from prospectuses and other regulatory filings—and their error rate is not trivial. If funds were tagging their data in XBRL, and the rating services were retrieving it in that format, fund companies and investors alike could have greater confidence in the accuracy of the available information. And after all, that’s a lot of what it’s all about.
Looking ahead, perhaps a long way ahead, maybe when my 12-year old is all grown up, XBRL could move us to a world where funds could just file data, rather than the multitude of different, but overlapping, regulatory filings prescribed by current rules. As an added advantage, investors and other users could retrieve fund data in a manner completely personalized to their own needs. We live in a world of personalized services—personal shoppers, customized websites, even different ring-tones for friends and family—why not investor-created prospectuses?
Finally, mutual funds themselves are huge investors, and can benefit as end users of XBRL from the ready availability of accurate data about potential investments. Good data is a key to investing success, and speed in retrieval often means getting the critical trade in before the other guy. We find that good research about micro and small cap companies, in particular, is growing more limited, and XBRL just may allow a diligent portfolio manager to discover the next Microsoft. And for that, my friends, we don’t have to wait for my kid to grow up.
We still have a lot of thinking to do about how XBRL fits into the mutual fund world. We’ve barely scratched the surface, both in terms of how to make it work, and of what the potential could be.
Sometime around 1980, an executive of a technology corporation was asked if his company planned to get into personal computers. He said: “Well, we thought about it, but we couldn’t figure out why anybody would want to have a computer in their home.”
We don’t intend to make that mistake with XBRL.
See you in the future, folks.