American Enterprise Institute

May 30, 2006

[Edited transcript from audio tapes]

11:45 a.m.	Registration
Noon	Lunch
12:30 p.m.	Introduction:	Peter J. Wallison, AEI
12:45	Keynote Address:	Christopher Cox, chairman, U.S. Securities and Exchange Commission
1:45		Panel I: XBRL, the New Computer Language: How It Creates Interactive Data
	Panelists:	Richard J. Daly, Automatic Data Processing, Inc.
		Mark Schnitzer, Morgan Stanley
		Louis M. Thompson, National Investor Relations Institute
		Mike Willis, PricewaterhouseCoopers
	Moderator:	James K. Glassman, AEI
3:30		Panel II: Enhanced Business Reporting: Why It Is Necessary and How It Works with XBRL
	Panelists:	Robert Eccles, Advisory Capital Partners
		John Philip, Infosys Technologies Limited
		Mike Willis, PricewaterhouseCoopers
	Moderator:	Peter J. Wallison, AEI
5:00	Adjournment

Proceedings:

Peter Wallison:  We’re very fortunate today to have an excellent keynote speaker to kick off this program on interactive data, and something called XBRL, which is the same thing.  I’ll do a little introduction and then Chris Cox will talk.  Then we’ll have a conference on XBRL and enhanced business reporting, which is an advance even beyond XBRL.  I hope you’ll all stay for that, which will be a fascinating program.

Every agency head, from the President on down, wants to leave a legacy of his time in office.  Few financial regulators have an opportunity to make the permanent changes that will amount to a legacy; changes that will last well beyond their time in office, and also improve the functioning of the economic sector under their jurisdiction.  In the case of the SEC, much of the Chairman’s tenure is often taken up with firefighting:  responding to Congressional demands, financial scandals, and problems in the securities market or the global financial economy.  There is little time left over in the usual case for initiatives that can actually make things better for investors, let alone public companies and other registrants.  In fact, it probably takes an act of will to break away from the day-to-day concerns of the Commission to think broadly about the nature of the agency’s role and how it can use its authority to improve the functioning of the securities market.

The current chairman, our keynote speaker today, appears to have done this.  He has made all of the familiar statements about protecting investors and enforcing securities laws against malefactors, but he has also focused much of his attention on a subject that goes beyond these customary goals.  He has recognized that investors have plenty of incentives to protect themselves, that software companies will produce the necessary applications if the demand is there, and that public companies would be delighted to reduce their reporting and filing costs if only they can be shown how to do it.

Our conference today focuses on one of the key initiatives of SEC Chairman Chris Cox.  It’s an idea he calls interactive data, but it also goes by the somewhat “techie” moniker, XBRL.  Interactive data is a good term for this important development because it implies communication between a database, such as the SEC’s files of company reports, and a software application that makes use of this data.  Another way of saying this is that it is interactive because it would make the SEC’s database of company reports machine readable. 

Today, if an investor wants to compare something as simple as the sales of all pharmaceutical companies, he or she must go to the SEC’s website, download their 10K reports, review the financial statements, and input the relevant numbers into an Excel spreadsheet.  Even this time-consuming effort is not likely to be sufficient, however, because some companies display their sales net of discounts, while others subtract the discounts at a later time. 

Thus, something as seemingly simple as sales numbers can require laborious searches through income statements and footnotes.  Interactive data, or XBRL, solves this problem by allowing a software application to extract precisely the comparable numbers from the SEC’s database of company reports.  In seconds.  And of course, sales numbers are only one of hundreds of important disclosures in a company’s financial statements and footnotes, all of which become accessible immediately through the use of XBRL. 

In the conference today, we will show you how this is done.  Obviously, this immediately empowers investors and analysts by making comparisons among companies easy to do.  As you will see, it also has the potentially vast ability to reduce costs that companies incur in preparing their financial statements and SEC reports, including the textual disclosure known as Management’s Discussion and Analysis, or MDA.  The same XBRL and interactive data capability will lead one day to better disclosure by companies, going beyond financial statements to indicators that show whether they are adding or destroying value through their operations. 

These important facts cannot be revealed by financial statements alone, but are vital if investors and analysts are to make the right choices in allocating capital.  In our second panel this afternoon, which will focus on what is called enhanced business reporting, we will discuss these developments and their relationship to XBRL. 

All this has been made topical by Chris Cox’s effort to bring this country’s system of disclosure to investors into the 21st Century, and none too soon one might add.  The development of an XBRL taxonomy or classification system has already been completed in the EU.  Financial statements prepared in international financial reporting standards, known as IFRS, will soon be searchable in a way that U.S. financial statements prepared in GAAP will not be.  Many commentators have already noted that a good deal of financial activity has migrated from the United States to financial markets in London, Paris, and Frankfurt.  In part, this is a result of our regulatory and litigation system, which has added costs and risks that are not common abroad.

To change this will take time, if it is ever accomplished at all, but the failure to adopt ideas like interactive data or XBRL will only place U.S. companies at a further disadvantage in competing for capital with foreign companies and the U.S. financial markets, at a greater disadvantage, in seeking the business of foreign issuers who can choose where they will list.  Under these circumstances, we should count ourselves lucky that the chairman of the SEC is focusing attention on what his predecessors almost wholly ignored.

But in a sense we shouldn’t be surprised.  Chris Cox’s career, from the time he received a joint Harvard Business School and Harvard Law School degree, through his work on class action, dividend, taxation, and budget process reform while he was a member of Congress, has reflected a much broader vision than his contemporaries.  When others have seen difficult or impossible hurdles, he has seen challenges and opportunities.  In numerous leadership roles in Congress, he has come to understand that little gets done unless someone commits himself to promoting a new idea or approach.  If he succeeds during his chairmanship in bringing this interactive data initiative to fruition, it will be a great achievement, no less important, I think, than the regularizing of financial disclosure through the development in the last century of generally accepted accounting principles.

Now, that’s a legacy.  We should all wish him well.  And please join me in welcoming SEC Chair Chris Cox.

[Applause]

Christopher Cox:  Thank you very much, Peter, and thanks to all of you at AEI for inviting me here today.  I am not at all surprised that Peter Wallison is in the Vanguard of the early adoption of an important technology initiative.  When I first met Peter 20 years ago, which in aggrandizing me in this introduction he failed to mention it was because he hired me into the White House when he was in the White House’s counsel with President Reagan, I noticed that Peter had on his desk in his West Wing office an IBM PC. 

Now, an IBM PC 20 years ago was not really high technology.  In fact, it’s quite a museum piece these days.  I asked him what he used it for and how he used it.  He said, “I was told that if I got this computer, it would simplify my life.”  And he said, “And it has.  Ever since I’ve gotten this computer, I’ve had no life.”  These days, computers are a little bit easier to maintain, and they can do wonderful things for us that certainly when man landed upon the moon, no one could imagine.  So it is very apropos that we are all getting together today to imagine how computers and innovation can help improve our capital markets. 

I hope that your minds are now all focused after a three-day weekend – you are sharper than ever after Memorial Day and your batteries are recharged.  You probably know that today, May 30, is actually Memorial Day, the anniversary of the first Memorial Day, so this is itself sort of a holiday for us, I suppose.  You may not have known that Memorial Day was founded not by the government, but by a private organization.  A retired U.S. Army general named John Logan, who was the head of a veterans’ organization called the Grand Army of the Republic, declared the first Memorial Day on May 30, 1868.  He said that Americans should honor the fallen by decorating their graves with flowers and right here at Arlington Memorial Cemetery over 5,000 people went and did just that on the very first Memorial Day.  While it immediately became an annual event, it was not signed into law until President Nixon did so in 1971.

So now back to AEI and where we are after entering into history.  You will be pleased to know that not only is the John Logan who made that first Memorial Day proclamation the same John Logan who was a Republican senator from Illinois and the Republican candidate for Vice President in 1884, but it is the same John Logan after whom the traffic circle just six blocks from here, Logan Circle, is named.

So today at AEI, we are talking about another national movement that was started in the private sector and that the government will eventually sign onto.  It’s the movement to the electronic exchange of financial information that will be standardized across many technologies, including personal financial software, corporate financial preparation software, and of course the Internet.  To make it sound high tech, which it in fact is, it goes by the name XBRL.  In plain English, as Peter noted, I call it interactive data. 

Either way, what we are talking about is very straightforward.  For users of financial information, we want better access, more accurate and reliable information, and significant new capabilities.  These are really quite simple concepts, but interactive data, which is truly revolutionizing and exciting, is often bogged down by its own terminology.  XBRL is not an abbreviation that rolls off the tip of the tongue.  Not only does calling something XBRL make people think you are talking about something on the Sci Fi channel, but then there’s taxonomies. 

I learned during my 17 years in Congress that anything that begins with “tax” causes native suspicion.  It seems to me we might be approaching this from the wrong end.  Instead of talking about all the gizmos that will make markets work better and give investors better tools than they have today, we ought to just skip to the end and start talking about the reasons that interactive data will make the lives of investors, companies, and even regulators better. 

Watchmakers, after all, don’t sell their products by talking about tachometers and rotors.  They tell you that their watches keep perfect time.  You don’t have to know anything about movements to be able to tell time and to know that it’s always better if your watch doesn’t stop on your way to an important appointment.  With interactive data, the parts and the internal movements can likewise be daunting, but the result is to make it easier and better for the individual and for the markets as a whole.  Interactive data is a marriage made in heaven between investing and high technology.

Imagine if you will that you are reading the paper on a rainy Sunday late morning.  You’ve already read most of the sections.  Then you happen upon the classifieds, which you read just out of random interest.  You start reading the lonely heart section and you find these two ads sitting cheek by jowl.  The first one reads:  Young man eager to work for a secure future seeks honest mate to help him break out of traditionalist high-bound mindset and explore future possibilities. 

The other one says:  Young woman not afraid of the future, eager to explore new opportunities, frowns on secrecy and lies, thrives on open communication and honesty.  You think instantly that these two will notice each other’s messages and that they are sure to get together.  Imagine your disappointment when you notice that the same two ads run the next week and the week after and the week following.  You begin to despair that these two will never get together after all.

Sadly, this is how it appears sometimes for the traditional world of investment and the 21st Century world of high technology.  This may be a marriage made in heaven, but somehow they keep missing each other.

Technology has revolutionized almost every aspect of our lives:  the way we communicate with one another, the way we listen to music, and even the way that we play games.  Yet, with one of the most important things that we do in life, investing our savings for our future, what do we do?  We are still today relying on the open cry pit system of the 1700s and the printed page of Gutenberg’s 15th Century. 

This won’t do any longer.  Investing once was just for men – men wearing powdered wigs and knee britches no less.  It was once just for rich people.  No more.  Not in an America where nearly every working mother, every working family, and every retiree is directly or indirectly an investor.  Today over half of our citizens own stocks.  That’s a remarkable development.  Just 20 years ago when I first met Peter Wallison, only 19 percent of our citizens owned stocks.  What’s even more remarkable is that this trend is continuing and growing, as our citizens live and work longer than ever before.

When my father was born, the average life expectancy in the United States was 49 years.  At the end of the 20th Century, life expectancy was 77.  For children born since then, fully half will live to see their 90th birthday.  At least 10 percent of babies born today will live to be at least 100.  Demographers tell us that because they will be healthier for much longer, these of our fellow citizens will continue to work much longer than their parents did. 

This kind of revolutionary change in our society will demand corresponding changes in the way we save and invest.  Saving for one’s future will mean a much longer future than we’ve ever imagined before.  And with the movement away from defined benefit plans to 401(k)s and defined contributions, the investor is in charge.  So at the very same time that people are living and working longer, they are becoming active investors in droves.  But as with any participatory endeavor, the risk of injury increases. 

This obviously concerns us at the SEC, since our mission is investor protection.  And just in time, we have advances in technology that allow for this change in the game.  What we need is something that will give individuals faster access to better information that they can easily use and understand to make better investing decisions.  We need to make searches for information easier.  It should be easy to call up information about any company you choose.  You should be able easily to download it and use it in the personal software of your choice.  You should be able to easily analyze and compare the data with the same information from other companies.  We want to make the numbers derived from financial statements vastly more accurate, and we want to allow companies and investors to communicate with each other on a real-time basis:  24/7/365.

At a time when we have 24-hour news and 24-hour pizza delivery, why are we still living by the 10K and the 10Q?  If investors are going to be responsible for the growth of their investment, for picking which funds to put their nest egg in, they are going to need user-friendly responsive numbers that are easily accessible.

Interactive data does that by tagging numbers on financial statements and making them interactive.  So let’s say that you need to get ROI – return on investment – of all the companies in an industry.  Or you want to compare ROI for one company against others.  Interactive data will let you do that.  Let’s say you want to know about which company or which sector has the lowest beta.  Or you’ve heard somewhere somehow that the oil industry is the place to be for the next few years and you want to compare oil reserves among companies.  You want to make sure that when you call up reserves you get oil reserves and not accounting reserves for bad debts. 

The tags give each of the data points a unique ID.  That way, when an investor summons revenue or a prepaid expenses or goodwill, that is exactly what she gets.  Think of it, if you will, as every piece of data having its own GPS homing device. 

From the point-of-view of the Commission, our investor protection mission, there is a huge advantage if the investor is able to get this information before making her investment, and the numbers that she sees will be vastly more accurate than they have been up to this point.  People think that the numbers that analysts and investment professionals rely upon come straight from the SEC, that they are the same numbers that are filed with our agency.  Even most CEOs think this. 

But that’s wrong.  An entire cottage industry has developed to extract the information of the financial statements that companies file with us.  The error rate from this process of conversion is unacceptably high.  According to one study, for example, computers taking down information from our SEC website get the information right only 72 percent of the time.  People must then manually check every one of these hundreds of items from every company to get that 28 percent that was not captured.  This is, of course, time consuming and worse yet, because human fingers are involved, it is also error-prone.  After the human extraction has taken place, the error rates can range from 5 percent to 20 percent.

With all of the advancements that we have today, with all of the money that we spend on making sure that companies put out accurate numbers, why do we allow this low tech approach to continue?  EDGAR Online is a company based in Connecticut whose business it is to extract information from our filings and sell it to users downstream.  They have given us an idea, because they have aggressively adopted XBRL, of what the future could be like for all investors.  They have already tagged more than 10,000 companies’ statements, all the 10Ks and all the 10Qs going back several years.  EDGAR Online tells us that, with the aid of XBRL, it has attained a greater than 99 percent accuracy rate. 

Imagine retail investors having access to better numbers than professional analysts have ever dreamed og.  Imagine combining this accuracy with the power to mix and match and compare and analyze that data.  This can’t come a moment too soon because, according to press reports, research budgets have dropped precipitously since 2000 at all 7 of the largest Wall Street firms.  Some investment banks such as Wells Fargo have scrapped their research business altogether.  Research coverage has bounced back recently, but reports say that much of the new growth is on the buy side, in other words, research being sold exclusively to hedge funds and other money managers.  That kind of research isn’t available to individual retail investors.  Coverage of small-cap companies is still down from what it used to be. 

As Institutional Investor put it recently, “That’s not only a disadvantage to those retail investors who choose to pick their own stocks, it’s also hurting the ability of small companies, long a key driver of the U.S. economy, to raise capital, innovate, and grow.”  This is why the advent of interactive data is so important at this point.  It would help fill an informational gap that could be making markets less efficient.  In one case, it’s making the data more reliable; in the other case, it’s making analysis more widespread. 

Even beyond making the numbers more accurate and the research more readily available, interactive data will do one more thing.  It will help investors get real-time information that today is simply unavailable.  Think about it.  Isn’t it ironic that in this day and age we are still talking about 10Ks once a year and 10Qs four times a year?  Just as interactive data and the Internet can make the information within financial statements easier to extract and analyze, Internet technologies and web services such as RSS and Atom can get investors current financial information when they need it the most. 

Today, any investor who is looking for financial information first needs to find out when a company releases its information and then he needs to search for that information on the Internet, or wherever he can find it.  But there is a better way.  The same web services, like RSS, that today offer real-time automated delivery of news and sports, can enable the automated delivery of information to investors as it’s released based on their individual preferences.

As part of the XBRL voluntary program at the Securities and Exchange Commission, we are piloting the use of an RSS feed for all filings with XBRL documents.  That means that users will be notified of new filings as they occur, and they can access the data as it happens.  For professional analysts, the combination of interactive data and delivery mechanisms like RSS can fuel the real-time update of analysis models and risk screens based on user-defined criteria.

Of course, as you all know, this real-time communication is not limited to numbers.  XBRL can also tag narrative, so that means we can now get footnotes and management’s discussion and analysis and other text disclosures.  This combination of XBRL and RSS can create a constant stream of voluntary disclosure and, in so doing, vastly increase the quality of information that investors receive. 

As Peter and Robert E. Litan wrote in the book they wrote in 2000, The GAAP Gap, “The history of accounting has involved a continued tug-of-war between those who fear that too much disclosure will expose a company’s secrets to the world and those who believe that the more sunshine a company lets in, the better off investors and even the companies themselves will be in the long run.  The clear trend has been in the direction of more disclosure, both mandatory and voluntary.  We are simply advocating an acceleration of the trend.”  That’s the end of their quote.  This is perhaps why such blue chips as PepsiCo and General Electric this month asked to join the SEC’s XBRL test group. 

This conference is focused on the good many of the details of implementing what will be in effect a giant leap in investor empowerment, so it is worthwhile to step back a moment from the twigs and the branches so that we can see the forest.  Our nation has long been home to the richest, deepest, and most liquid capital markets in the history of the world. 

That market is a mighty engine that continues to power our economy at over 5 percent growth this year.  It has contributed to the creation of millions of new jobs and the national unemployment rate that is now down at 4.7 percent.  It is the source of a way forward for every young American looking forward to attending college.  It holds the retirement savings of our parents and indeed everyone in this room.  It is the well that our nation will tap to support the innovations in healthcare that will cure cancer and end diabetes.  It is worth doing everything that we can to keep that market working and to make it always work better.  In the end, we will all be the better for it. 

That’s what interactive data, XBRL if you will, and this conference are all about.  Thank you for your work, for your ideas, and your inspiration.  We at the SEC are grateful for your private sector leadership.

[Applause]

Peter Wallison:  Chris will take some questions, so have your questions ready.  I see a questioner in the back.  Would you identify yourself please?

Bert Ely:  Bert Ely.  Chris, thank you for being here today.  What are your thoughts as to when and under what circumstances the SEC would make XBRL mandatory for SEC filings and specifically annual reports:  10Ks, 10Qs, and the like?

Christopher Cox:  I’m hopeful that with the kind of private sector leadership that we’re seeing in this room and indeed that we are seeing around the world that it will be not only possible but easier for registrants with the SEC to file their information this way, using interactive data, so thinking of this as a regulatory mandate is not, I think, the right way to think about it.  The best way to think about this is a magnet.  We already have a voluntary program and people are volunteering to sign up for it.  I don’t think there’s any question that someday, and I hope that someday is very soon, this is going to become the norm.  Then, just as with Memorial Day, we will sign it into law.

Peter Wallison:  Jim, wait for the microphone.  Identify yourself.

Jim Smalhout:  I’m Jim Smalhout.  I’m a writer based here in the Washington area.  My question for you involves the possible implications for investors who are committed to index products and mutual funds.  I am wondering if you or some of your colleagues on the Commission have a view about what all of this means for them.

Christopher Cox:  Indeed we are beginning, or at least one of the beginnings that we are making, is with mutual funds and ETFs.  There is no question that with our focus on the retail investor at the SEC these days that’s where we should be, because mutual funds and ETFs are the investment of choice for the individual investor.  Furthermore, under the regime of the Investment Company Act, the kinds of disclosures that mutual funds make to the SEC and that are then distributed for the consumption of individual retail investors are more standardized and routinized and lend themselves to XBRL tagging.  The Investment Company Institute, again on a voluntary basis – this has not been result of a new SEC rule – has decided they want to work with us and complete the taxonomy necessary to do this so that all of the information that you get as an individual investor that they file with the SEC can come in this interactive form. 

If you own a mutual fund, imagine how this information might be used.  If all of the companies, ultimately, that are in the basket of the mutual fund, themselves report using interactive data, then it will be possible for computers operating in the background that the individual investor never sees, to sum up all this information, to use it in a way that yields some easily understandable numbers or benchmarks.  This kind of fund has this kind of risk or this kind of overall characteristic, based on very good information from each of the constituents of that fund.  Or this fund overlaps that other fund by 78.2 percent.  A lot of things that individual investors would find very useful can be made possible using interactive data.

Jeremy Grant:  Hello, yes, I’m Jeremy Grant from the Financial Times.  Two questions, if I may.  For somebody like me who is new to this pretty much, what got you onto this in the first place?  What was the red flag, if you like, that you saw that said, we’ve got to do this?  And secondly, I understand that the Europeans are reasonably far ahead on this.  Are you looking to some of their standards at all?  Are you involving them?  Are you consulting them as you look at this question of XBRL taxonomy?  Thanks.

Christopher Cox:  I’ll try to take both questions.  First, what got me interested in this is the fact that XBRL International and XBRL U.S. and the standard itself existed.  They were there.  They were ready to be seized.  It was simply a question of helping participate in the leadership of this effort, which the SEC is doing, to bring it about more quickly, to hasten it.  XML, of which XBRL is a close cousin, is ubiquitous in our economy.  If you are a young person, and there are several in this room, you probably already know about XML and RSS feeds.  If you are not, you might never need to learn it to get the benefit of it.  But there is no question that this is now ready.  This is something we can take advantage of.  And the mismatch between the way we do business now and the way we might do business in a matter of months is so great that we simply have got to take advantage of the opportunity.

Second, with respect to the rest of the world, this is really part of the attraction.  The rest of the world is involved in this.  This can be a global standard for the exchange of financial information.  I am not concerned that other countries might be ahead of us.  I’m thrilled that, as we move forward on interactive data, that we will be doing this in harmony with people all over the planet, so that the end result will be that our global capital markets, which are a fact of life, will function all the better for it. 

Just as one example, in Britain, as you know, Her Majesty’s revenue and customs is already set up to have all of the tax filings, on a mandatory basis, be made in XBRL format.  I am absolutely certain that the government agencies that are already doing this here in the United States or thinking about doing it, the Fed, for example, gets XBRL filings from banks and the IRS is beginning to look at this; we’ll be doing so in just a matter of time.

Lou Thompson:  Chairman Cox, I’m Lou Thompson with the National Investor Relations Institute.  If XBRL takes hold, do you think this might give the SEC the opportunity to, retaining the 10K, but move away from 10Qs and go toward real-time disclosure as outlined in section 409 of Sarbanes Oxford?

Christopher Cox:  Let me go you one better and tell you that we have, by last count, over 800 forms at the SEC.  It has been estimated that using interactive data, we could reduce that number to about 12, based on the unique data that are contained in each of these separate forms.  The idea would be, simply, you flip the model.  Instead of the form being the basis for the information, the information becomes the basis for the form, and the information can go wherever you need it to.  It can reorganize itself at your will.

Winter Casey:  Hi, my name is Winter Casey.  I’m from the national journal Tech Daily. When you first started talking about this, I believe it was not this spring but last spring, saying that this was a great idea, and then I remember in January you said we hope that companies will be adopting this within months.  As of yet, it seems like only I think the last release I received from the SEC was 20 companies.  Do you expect this figure to rapidly jump anytime soon and why?  Considering its been around for a couple of years now, it really doesn’t seem like companies have been adopting it all that quickly.

Christopher Cox:  It is indeed just a matter of months since I first started talking about this.  Remember that I was sworn in, not in the spring of last year, but in August, and I don’t want to cavil except that you’re talking about months and the difference between spring and August is four months.  We’re going ahead lickety-split here.  I am very pleased with the voluntary take up.  If you take a look at who are the participants in our voluntary XBRL filing program, it’s quite a cross-section of different kinds of companies – not exclusively technology companies by any means, although some of the top tech companies in the country are represented. 

United Technologies, which if memory services is the third largest industrial company in the United States, has told us that their costs of putting their financial statements in XBRL form was under $40,000.  It’s remarkable what we’re learning from this voluntary project.  And that, by the way, included start-up costs that they won’t have to repeat.

The utilization of this automated filing will make it cheaper and easier for registrants to supply the information so it isn’t all investor benefit, although that is what we’ve been talking about here today.  Investors surely will benefit, but there’s also a big registrant benefit, and I should add that in many ways companies are consumers of their own financial reporting.  Real-time management control information that is made possible by XBRL and interactive data is a great benefit to companies themselves.  For all of these reasons, I expect that we will in fact see a significant jump in the number of firms that are taking advantage of the opportunity.

Chris Whalen:  Chris Whalen from Institutional Risk Analytics.  Mr. Chairman, do you think the technical process of tagging and structuring corporate filings is going to influence the accounting debate over GAAP?  Both yourself and Chairman Baker and other spokesmen have talked about XBRL being used to essentially normalize the financial statements of different companies so that they are comparable.  You don’t have to normalize them by hand, which many of our colleagues on Wall Street do.  Do you think that this debate over the technology is going to also influence the debate over the GAAP accounting environment?

Christopher Cox:  It is a very important question because I think we need to all come to quick agreement that we are not trying to use XBRL as a driver of standards.  That is to say, we are not going to make new standards or more complicated standards or different standards as a result of the fact that we are reporting the information in this way, through interactive data.  I think that some people fear that because interactive data, XBRL if you will, can tag data all the way down to the checkbook level or the tribal voucher level, that somehow this is going to mean that we will have new government standards for all of these things. 

I want to make sure that is not what happens here, that rather XBRL and interactive data are the means by which we transmit this information but not an introduction of new complexity or new cost.

Peter Wallison:  We have time for one more question, in the back.

Lu Shing:  Lu Shing [phonetic] from China’s Beijing Magazine.  If I can use this opportunity to ask one question that Chinese companies are really concerned here, because in July, the Sarbanes-Oxley Act will be applied to China’s public companies listed in the United States.  What are your suggestions to them?  And do you think this new technology will help the Chinese companies listed in the United States to comply better with the tough accounting and disclosure legislation in this country?  Thanks.

Christopher Cox:  For accelerated filers – large, foreign, private issuers, it will of course be going forward with Sarbanes-Oxley’s implementation.  For others, you need to continue to await announcements on that subject.  With respect to the relationship between SOX compliance and interactive data, I think there is one.  I think that in the same way that interactive data will make financial reporting better, faster, easier, and less expensive for reporting companies and consumers of financial information alike, it will therefore improve the ability of companies to deal with their regulatory responsibility.  It will make that process less expensive, more accurate, cheaper, more routinized, and so on.

Peter Wallison:  Thank you very much.  All right. 

[Applause]

Christopher Cox:  Not that this is a revival meeting or we’re trying to get people before they leave, but we just signed up our 21st company, ADP, as of this morning.  So that’s good news.

[Applause]

Peter Wallison:  Come to XBRL, come to XBRL.  Okay, we’re going to right into our panel, so can I ask the first panel to come up to the podium.

Sit down and I will introduce you, and then we’ll take it from there.

This panel, which will address XBRL itself rather than enhanced business reporting, which is the subject of the next panel, is going to be moderated by Jim Glassman, my colleague here at the American Enterprise Institute. 

Jim specializes in issues involving economics and financial markets.  He is also the host and the co-founder of techcentralstation.com, a website that concentrates on matters of technology and public policy.  In September 2004, Jim launched a new organization, Investors Action, for which he serves as chairman.  Investors Action aims to educate America’s 90 million investors and represent their interests in the public policy arena.  He also writes a weekly op-ed column on economic and political topics for Scripps Howard, and his most recent book, The Secret Code of the Superior Investor, was named one of the top ten investing books of 2002.  There’s a lot more on Jim in the folder that you’ve gotten, as well as the other people who will be speaking today. 

Jim will then introduce the panel and then moderate it.  Thank you, Jim.

James Glassman:  Thank you, Peter.  I’m just going to sit here and introduce people.  So welcome to the post-perennial part of the conference today.  We are incredibly pleased and honored to have a panel like this today.  Peter Wallison put this conference together and it is very exciting to me to be a participant in it.  I’m not going to make any introductory comments.  We’re going to go straight to the panel.  Let me just introduce them briefly in the order in which they are going to speak.

Mike Willis is a partner with PricewaterhouseCoopers.  He served as a founding chairman of XBRL International, which today is composed of 450 leading software accounting and finance companies from 26 countries around the world.

Richard Daly is group president and executive committee member of Automatic Data Processing.  ADP is the largest provider of shareholder communications services. 

Then, after these two presentations, we will have comments by Mark Schnitzer, who is an executive director at Morgan Stanley, where he is responsible for managing data strategies for the ModelWare business.  He was also the founder and president of FreeEDGAR, which he sold to EDGAR Online in 1998. And he knows this business very well.

Finally, Lou Thompson, whom you just heard from in a question.  He is the president and chief executive officer of the National Investor Relations Institute and is an expert on matters of disclosure, governance, and corporate strategic management topics.  A while back, he was assistant White House press secretary to Gerald Ford.

We’re just going to jump right into it with Mike Willis from XBRL International.

Mike Willis:  Thank you, Jim.  I appreciate the opportunity to be here.  Peter, thank you very much for hosting this event.  What I’m going to cover is really some XBRL topics:  what is it; why is it relevant?  And then I’m going to show you how XBRL can be used to enhance delivery, to lower cost of analysis, and to enhance the compliance processes for registrants.  Those will be my primary talking points and focus for our conversation this afternoon.

So why is XBRL relevant?  Well, it’s relevant because the current paper-based model puts us in a manual supply chain.  Once we report in paper, even electronic paper, these electronic documents, they function in a supply chain that forces everybody downstream into a bunch of manual processes.  I’ll walk through that supply chain in a few minutes, but the basic idea here is, if these other supply chains can be automated and apply automation, why can’t the supply chain that’s serving the capital markets be more automated, and why can’t we apply technology and the Internet to that idea?  That’s what XBRL is all about.  It’s about leveraging the Internet to make the information supply chain feeding the capital markets more efficient.

So what’s XBRL?  You’ve got this in your slide deck.  It’s basically an international information format for business information.  Chairman Cox mentioned that it is a dialect of XML, and XML is an Internet standard, and that makes XBRL an Internet standard.  XBRL is in fact an XML dialect.  So when someone says XBRL--that is the same as XML.  There is no difference.  It’s the difference between the kind of English that your physician uses and the kind of English that your attorney uses.  They both speak the same language; they may use the same terms, but they mean different things.  So XBRL is XML for business reporting.  It functions very much like the barcode that is on this bottle of water. 

If we think about the barcode for a second, the barcode wasn’t around 25 or 30 years ago.  It was an idea.  And what was the ramifications of the barcode?  Well, about 30 years ago, the average grocery store had on average 6,000 units on the shelves.  So did the barcode standardize those units or did it standardize the way a unit can be described?  Clearly, it standardized the way a product could be described.  It did not force people into a standard type of description.  It allowed you to standardize the way you described a product.  The outcome today is there is roughly 36,000 units on the average grocery store shelves. 

So an increase of 6 in terms of the number of units that are the shelf, while decreasing the cost of managing inventory in the retail grocery sector by $17 billion per year.  The point I want to make to you here is that standardization, like the barcode, dropped the cost dramatically while increasing the volume of inventory that could be managed:  lower cost, more efficiency. 

So XBRL is that same idea around business information.  It’s a way to describe information but very simply it’s an international information format.  It is international because there’s consortium, members, and jurisdictions in over 26 countries around the world.  The members--the individual companies--actually join at the country level; then they are represented at the international steering committee.  So like other Internet consortiums, this is driven by market participants, and in this case market participants literally from all over the world.  Just a few weeks ago, the United Arab Emirates joined and the stock exchange there began a program very similar to what the SEC has announced here in the U.S.  So international in nature, represented by members in over 26 countries.

Mike, what does XBRL look like?  Well, I’m going to take you to the Reuters website and we’ll look at the results from Reuters and it looks like this.  This is what XBRL looks like.  It’s not for a human being to read.  It’s actually for a machine to read.  Now let me actually just go to the Reuters website and show you what this looks like.  Here is the PDF document.  That Adobe PDF document has actually been rendered from an XBRL document, so to a human being this looks no different than the Adobe file that you had in the prior quarter. 

So Mike, what’s the difference?  Well, underneath the covers is an XBRL document that your machine can reuse.  Also, if you mouse over a particular number or an item, it will actually tell you exactly what it is, in this case revenues in accordance with the IFRS vocabulary, in pounds and give you the exact dollar amount.

There’s a lot more intelligence here than your normal paper document or your regular PDF document, my point being that underneath the covers is something that looks like this and you can see the revenue number I’ve highlighted here - $2.4 billion – but if I actually go over to the PDF document, you can see the same $2.4 billion there again.  It’s just a different format for the same exact information.  Now, you could consume it as a PDF document as a human being.  It looks no different.  It looks just like a piece of paper.  But your machine can also consume that same information, and I’ll show you what that means in just a second.  You can also scroll down to the bottom of the Adobe document and click on one of these paperclips.  It will actually kind of release the XBRL document that is actually embedded in the Adobe file. 

I just want to give you a flavor of what this looks like.  To a human, it could look just like a piece of paper, but to a machine, it actually looks like something that’s very structured and therefore very reusable. 

Let me back up for a second and actually talk about business reporting processes, because I talked about the supply chain a minute ago, but I want to focus in on the idea of supply chain processes.  The FFIEC and the banking regulators here in the U.S. have to collect information from all the banks, they have to validate that information, they have to then analyze it, and then they have to prepare a report that might go to the Federal Reserve.  That sort of process is something that occurs at the regulatory level.  That same set of processes also occurs at the banking level.  The controller for a bank that is reporting to the regulator goes through the same set of processes, of collecting information, validating it, analyzing it, and reporting it. 

These processes actually occur at every step of the supply chain, so rather than thinking about a process in terms of what application is it in, I want you to think about the process in terms of these basic functions, which recur at every step of the supply chain.  And because the supply chain is currently manual and paper-based, these processes are manually executed.  The idea of collecting it is manual; I have to cut and paste.  For validation, I have to check to make sure the numbers that were entered were entered correctly. 

Analysis is done manually.  Think about the way people do analysis in Excel today.  They build models that are very complex, and those models have a lot of information that is embedded in the Excel worksheet – all done manually.  When you begin to look at this, the supply chain includes operational activities – at the very lowest end transaction information, with policy making on the far right. 

What’s XBRL do for us?  It does a couple of things, but I want to start on the far left with the operational activities, because XBRL is not alone on the supply chain.  There is a lot of information that is actually at the operational level.  I’m going to use ACORD.  Many of you probably have not heard of ACORD, but ACORD is the XML language for the insurance sector.  It’s an international consortium of insurance companies, and they have gotten together and described transactional information for insurance products:  property and casualty; life; reinsurance; etc.  As a result, they can exchange information much more efficiently at the transaction level in the insurance sector.  At the operational level, we might be using something like ACORD. 

At the ledger level, though, you would be using something like the XBRL ledger.  That’s a different idea than the XBRL you would be using for the reporting with the FDIC or the SEC.  The XBRL ledger is a way to describe information within an enterprise.  It’s a way to describe a ledger.  It’s not a general ledger.  It’s really any ledger.  So think about this as a universal way to describe information in an enterprise, much like the barcode.  It’s not the 35 million terms that were going to be in a ledger.  It’s merely a way to describe a ledger:  account number, description, business unit, debit and credit – those kinds of ideas.

Additionally, you have XBRL for external reporting.  This is what people like the SEC and the FDIC and the IRS will be interested in.  These are separate topics.  They are related, but they are separate.  There’s the idea that someone might be able to drill all the way down from a 10Q back into a ledger.  That’s not really technically feasible, although you could do it by exposing your ledger, but if you give someone your 10K mapped in the XBRL external reporting language, it’s not going to give them de facto access to underlying ledgers.  It’s two separate ideas.  It’s really two separate concepts.  However, they can be related for efficiency purposes in your company.

My point here in this supply chain is that you have this supply chain that is currently being conducted on a manual basis.  What XBRL allows you to do is to begin automating these processes of collection, validation, analysis, and reporting at literally every step of the supply chain, so that I can begin to automate these steps that are currently done manually.

What are the outcomes of a manual supply chain?  Well, we already know that access is manual – cutting and pasting.  Manual validation.  And validation is a user problem.  Validation is all on the user’s shoulders.  How do you validate something?  You have to go and look it up again.  That is not a good idea.  XBRL allows you to make validation a preparer problem.  The FDIC implemented that concept and, as a result, they were able to increase their data quality in a very substantial way – I think it’s over 60 percent or so increase in data quality because validation became the bank’s responsibility, not the regulator’s responsibility, much as it should be.

Coverage is dwindling.  I think that the manual access to data is actually one of the contributing factors in terms of how much information analysts are accessing to actually perform an analysis.  For instance, things like the discount rate of a defined benefit plan is a piece of publicly available information, but because it is buried down in a note, it may or may not be timely included in an evaluation analysis.  I think Mark Schnitzer will speak to that later. 

Reuse is impacted by the timeliness of a manual supply chain.  How quickly can an analyst actually include the information reported in their evaluation model?  Well, if it’s manual, the answer is not very quickly.  If I ask you to go get the information for a company for the last three or four years and put it into a model, how long would that take you?  Are you thinking hours, days, or minutes?  Or are you thinking seconds?  That’s the type of change that XBRL will bring – not hours, but literally seconds to obtain information for decisions.

The zero context, this is a point that I actually want to show you, but it’s much like the Reuters example I just showed you a picture of.  The ability for information to actually relate to other concepts, whether it be the GAAP standards or the actual business rules that might be appropriate for specific pieces of information. 

So if I just step back for one more second, I want to emphasize the impact of standardization.  These ideas are applicable to XBRL, but they are also applicable to the universal product code.  They are applicable to things like the containerization in the shipping industry, railroad ties, the railroad gages were standardized – all these concepts we think of today as almost infrastructure that were at one point proprietary technologies, much as we think of information today.  XBRL is about taking the concept of information and making it just like a piece of infrastructure.  I can plug my application into the Internet and I can get any information I want.  I want to show you that in just a second.

Let’s talk about the second area here, around enhancing analysis.  Well, what are the current problems with analysis?  Proprietary formats adversely impact the availability of information and require manual transfers of information.  As a result, we end up with untimely analysis, we end up with limited scope of information included in analysis, and unforeseen transformations where third parties normalize data, and that is a problem for analysts who are trying to quickly include it in our models. 

So let me show you what XBRL actually means to an analyst.  Chairman Cox talked about EDGAR Online, so I want to show you an actual model.  Here I have ADP’s results in a template.  What I’m going to do is, I’m going to go get Microsoft’s information in this template.  Now what is occurring here is that I am in Excel, I’m going to query a web service – which Chairman Cox talked about.  This is a broadcasting idea.  A website is a publishing idea in which human beings go and look at information.  A web service is a broadcast of the data, so I can sit in my application, never go anywhere, and the data will come to me. 

So I am going to simply put the Microsoft ticker in here.  There is the Microsoft data.  I’m going to put the IBM ticker in here; there’s the IBM data.  I’m going to put the Dell information in here; there’s the Dell data.  I’m going to put the Intel ticker, and there’s the Intel data.  I just blew through five or six companies in a matter of seconds.  That’s what this means to an analyst.  To a retail investor, to a buy-side or sell-side, their ability to access information goes right through the roof using the Internet.

There are some other interesting ideas here.  If I actually mouse over this idea of total assets, in an Excel worksheet, this would actually show a number up here.  But in this model, it doesn’t.  It actually refers to the taxonomy, or the term.  That’s interesting, Mike; what’s your point?  My point is this:  if I come over to current ratio, I can now show you a term that is actually a macro. 

So think about how your macros work today.  They are physical battleship games.  C2 divided by D2.  Not transferable, not shareable, and certainly not manageable other than for someone drilling down into your Excel worksheet.  In the XBRL world, macros become shareable intellectual property.  They can be shared with someone else using Excel.  They can be shared with someone else using another application.  Because we are using standardized terms – descriptions – I can build business rules or formulas based on those standards.  If I’m an analyst, XBRL becomes a way for me to manage my intellectual property associated with my business rules or modeling concepts.  You can see that right here in this tool.  This is a good example of what it means to analysts, but this is within the EDGAR Online IMetrix product. 

I’m going to close that one and I’ll show you a different idea.  Here is just a simple idea.  I want to go get net income for Dell for the last five years.  I’m going to build it myself.  I’ve just copied this set of cells over here.  I’ll put in the same companies again:  IBM...this is how quickly you can actually access information.  And it’s not physically contained – I just put the wrong ticker in – because whatever you put in there will be rendered immediately.  The speed of access here is a completely different idea using Internet standards, web service’s delivery, and a structure like XBRL to actually identify what the information is, just like the barcode.  Let me close out of this one.  This is all using the IMetrix web service. 

Let me show you a different tool.  This is a tool called Zimba.  This is made by Hitachi and it is also an Excel plug-in.  What I’m going to do is I’m going to demonstrate what Chairman Cox talked about with the SEC RSS feed.  I’m going to request that the SEC EDGARs will now come to me.  In Excel, here are the companies that have actually filed in XBRL within EDGAR.  Notice at no time did my fingers leave my hand.  I never went to the Internet.  The information came to me.  That’s the big transformation here.  One of our panel participants this afternoon is John Philip from Infosys.  He’s here visiting with us, so I am going to request his information.  Once I’ve determined what information I want, I simply hit the Open button and the 10Q is delivered to me inside of Excel.  Now I have in just a matter of seconds the 10Q for Infosys in my Excel document.  There it is.

The idea of going to EDGAR and digging something out, that idea evaporates in the XBRL model.  You’re using Internet protocol as a web service to deliver the information at a broadcasting manner so that an analyst can sit at their model and consume this information on the fly.

Those are two examples of consumption tools from an analytical perspective.  Since I showed you those live, we don’t need to go through the deck. 

How does XBRL do that?  Here I’m going to pull open the covers a little bit and show you what XBRL actually is.  Most people think XBRL is merely a vocabulary of terms or a dictionary idea.  It is very much that with this concept of a label.  In the middle you have an item that could be any concept reported by a company that is part of their vocabulary.  Additionally, though, XBRL has something called a presentation idea.  What this means is that the computer might see the label, but the human being would see the presentation representation. 

Why is that important?  In Korea, the stock exchange there would like to access the international capital markets.  Since not many of you probably read Korean, then you have a difficult time reading their top 20 companies’ reports.  What they did is they got the companies to produce their reports in the XBRL language.  Then they used the presentation in XBRL to allow the investment community to see those labels in other languages:  French, Dutch, German, Chinese, Japanese, Romanian, Russian, whatever.  In one document, the labels can be rendered in multiple languages.  The companies here can actually reach more investors as a result of this concept. 

Mike, that’s very interesting technically, but what was the economic outcome of the Korean stock exchange program?  Well, in the three weeks following their projects, those top 20 companies saw the degree of international investment double.  It went from the low teens to the high twenties in a three-week period, just due to the additional transparency and reusability of the information that was reported by the companies.  Why is the Korean stock exchange doing that?  Because the Shanghai stock exchange in China has already mandated XBRL as a filing format.  If you go to the Shanghai exchange, every public company on that exchange, some 843, have already reported in XBRL for several quarters, so it is a competitive deal there in Asia to access the international capital markets.  That’s the purpose of the presentation idea.

There is also an idea called the reference, which allows you to take any item and relate it to something else.  In today’s Internet model, this would look like HTML, but HTML is a physical idea and this is one that is logical.  As I showed you what the Reuters 10Q, you can actually point at an element and be referenced to what it is, in that case IFRS.  If you go and start looking at company reports, the XBRL taxonomies allow you to navigate from a reported concept to the related FASB standard or to the related SX rule.  Within the taxonomies, there are actually relationships between an item, an object of reporting, and the related rules.  These rules can also be company policies, so it does not have to be just compliance ideas that are external to the company; they can also be company compliance activities, like an accounting manual or an accounting policy guide. 

There are some other concepts at the bottom here that have to do with validation.  Those are the concepts that relate to business rules – that’s the formula concept – and calculation, which is one of the ideas that helped the FDIC dramatically improve the quality of information they were receiving.  The point of this slide is that XBRL is a lot more than definitions.  It includes a very dimensional orientation around information and a platform that you can use in a lot of different ways.  It’s not just about exchanging information; it’s about the better use of information.

Mike, you showed us how it can enhance analysis.  Let’s talk about how it can lower the cost of information delivery.  What are the current problems?  Again, proprietary formats combine content and information.  When I want to show you something, I have to show it to you in the presentation representation that you want to see it.  That means that if I am going to prepare a graph, I have to prepare the graph in this way, in that way, so that I end up with a bunch of presentation activity.  There is no reuse of those concepts.  In the XBRL world, I can give one XBRL document that, not only can be presented in different languages, it can also be presented in different formats and representations. 

I’ll show you that quickly.  Here’s United Technologies Corporation’s XBRL document.  I am going to highlight the income tax note and a disclosure of numeric context in the note.  Both of these concepts are defined.  Then I’m going to render that same document in Excel, just like I did with the Infosys report earlier.  Here is the narrative content from United Technologies Corporation rendered in Excel in just a matter of seconds.  Additionally, I can render that in a browser just like a piece of paper and I can read it just over my Internet browser.

So XBRL allows you to look at this information in multiple formats, in multiple presentation concepts, literally with a click of the mouse, again lowering the cost of providing information to the marketplace.

Let’s go inside the company and talk about how XBRL can enhance compliance activities.  This is the real economic benefit to registrants.  What are the current problems?  Well, the first one is disparate proprietary data stores across the company, and as a result, a lot of manual processes and manual controls.  Here’s a current process.  On the left I have business units, in this case I have 3, and there could be as many as 100.  In many large multi-nationals they may have 100 separate ERP systems at a business-unit level, not simply 3.  Those are typically consolidated with some type of aggregation tool – Oracle Financials, Hyperion, Cartesis, SAP, something that would aggregate those disparate data stores.  The first thing we do is we actually produce a report.  This is the draft 10Q. 

Here is the fallacy with this idea.  This is pretty pervasive in most companies.  As soon as you produce this document, you have now relegated yourself to a manual review process, because you have produced a document that is basically only human-readable.  If you have a set of compliance requirements, the way you figure out whether or not this document includes all the required compliance activities is to have someone go through and manually check them off.  Of course, then the auditors get to do the same thing, and then every time you version this document, you go through the same set of manual control processes.

In addition to that, there may be a lot of other information that is not included in the databases.  You are actually using Excel to aggregate a bunch of supplemental information.  I am going to refer to this affectionately as “spreadsheet hell,” because this is the area where lots of manual processes, transformation errors, simple mathematical problems occur:  either macros blow up, people change the format of the Excel template, some kind of manual problem occurs, and it generates all kinds of problems downstream because, again, we are relying upon a manual compliance activity to catch that.

Here is the review cycle.  These are technical terms.  The first one is a serial process I’ll refer to as the “pinwheel process.”  What you review is serial in nature; it’s passed along among the executives that need to review the document.  The alternative is the “ping-pong method,” in which you send it out to each person individually and then the combination of the two is, of course, the “Ferris wheel,” which gets you back to the same area.  Either way, it’s a major problem because these are all done manually.

Here’s the issue:  once we get to this state and we’ve actually filed a report, what do we do next?  Then we decided we’re going to apply the XBRL tags.

[Laughter]

Again, we go through the same set of review processes over here and it’s a problem.  Then we get the opportunity to do this for each of our stakeholder groups, be it regulators, investors, etc.

Let me go back to the barcode.  When barcodes were first introduced, they were actually applied by inventory clerks in the grocery stores.  How long do you think it took the grocery store owners to figure out that was a dumb idea?  It took about 3 months and they began asking the manufacturers to apply the tags back further in the supply chain.  Let’s try that idea here.  This is the XBRL model.  Rather than applying the tags at the end of the supply chain, let’s apply them earlier back.  Let’s apply them back against the ERP systems and against the actual spreadsheets. 

Now, Mike, this sounds very high tech – has anybody done this?  The answer is yes.  In fact, there is a company in Japan called Wacoal that did this 2 years ago.  There’s a case study that walks through the benefits:  lower costs, greater flexibility, better compliance processes, and most importantly better access to information by management.  Their option was to put in a big ERP system everywhere.  It cost them 3 years at $150 million.  The path they chose was this path, and this path only cost them 6 months and about $50 million.  That was on a full-blown ERP implementation idea. 

The point is that the use of XBRL inside the enterprise is a cost-effective alternative to other solutions.  It has been done.  In fact, Wacoal has been followed by Kirin Brewing in Japan again, and there are some companies in the U.S. that are beginning to test this at various priority processes in their organizations.

Once you have actually gotten the XBRL applied at the lower level, then you can review the content in its source – not in a document.  I’ll call this a collaborative review process.  Additionally, you could apply automated tools to assess if your report has everything it should have in it.  I can automate a great deal of that because now I have the structure and I can apply tools to it.  I no longer am relying on manual human beings to go through manual control processes.  Once I have done my review, I can simply file export an XBRL document to any of the constituents that I want to use.

The ability is here is actually to reduce the cost of reporting in a very significant manner while also enhancing the control environment and the ability for management to actually manage what is going on.  The point here is that XBRL allows you to enhance your controls at a lower cost. 

Mike, who can allow us to do this?  Here is one example.  What I am going to show you is a consolidation reporting application right here – Hyperion, Oracle, SAP, Cartesis, FRX – they all sort of sit in this area.  What I’m going to show you next is Cartesis in some spring shots.  What you can do is Cartesis has a publishing idea around XBRL.  You can pull a taxonomy in.  You can drag and drop and map your reporting elements against these taxonomies.  You can then review the report before it goes out.  Then, as an added benefit, much like I showed you earlier with the analytical concepts, in my consolidation application, I can actually pull in market data and begin to compare it.  Here’s the big benefit to management.  A lot of business intelligence is about rear-view window information.  It’s information in the company. 

What I am showing you now is within my consolidation application, I’m actually looking at the side-view mirror.  I’m looking at how I am doing in the marketplace.  That’s a lot better information for management when they need it, rather than just relying on historical period-over-period growth rates.  I may be growing at 8 percent, but if the market is growing at 3 or 4 times that, it would probably be good for management to know that, and not just at an enterprise level, on a business unit level as well.

Here are some examples of the information that is being delivered to them and actually populated through their Cartesis application – just an example, but again it is being rendered in graphs and tables; however you want to see it, it can be rendered very quickly from this format. 

Here are the key messages.  XBRL enables transformation of pervasively manual processes.  It allows you to automate access, validation, and analysis in an exponential way.  Automated processes mean less manual processes.  They mean less manual controls, less processing time.  All that adds up to lower cost.  The point that I would like to emphasize is XBRL is all about an economic incentive.  If you look at the Internet today, every public company in the U.S. posts their reports on their website.  There is no SEC rule that requires that.  It’s done because companies find it economically beneficial.  XBRL is simply the next step in Internet technologies to allow companies to achieve lower costs and greater communication capabilities.  The repair and process improvements are in your deck, and that wraps up for me on the XBRL overview.  Thank you.

[Applause]

James Glassman: Thank you, Mike.  Thanks for that very clear explanation of a very complicated subject.  Let me just ask you, you make a powerful case for XBRL.  What is the public policy role here?  What is the role of the SEC?  Is there one?  Is it important?  Or can we just go ahead with this in the private sector?

Mike Willis:  It’s a supply chain and the SEC is a participant in the supply chain, so they need to participate like the rest of us.  Their public policy role would be to facilitate collaboration, discussion, and the infrastructure components that may be needed, most specifically, I would say the taxonomy development arena and how the market participants can actually come together to work on taxonomies that can be used in the public domain.  That is probably the biggest public good of relevance to the capital markets is the taxonomies.

James Glassman:  Is it that the United States is behind foreign countries in this adoption?

Mike Willis:  I don’t know that I would phrase it as “behind.”  I would phrase it that other countries have a more aggressive political style.  I think you can see the countries who have taken a mandatory approach have a government that is maybe more enforcement-oriented or more restrictive than we find here in the U.S.  I think that is probably the primary indicator of those countries being able to basically mandate that.  I think that the economic incentives, though, that are available to us will result in a much deeper implementation in the U.S. than we may see by other countries’ mandatory implementations.

James Glassman:  Lou wants to make a comment and then we will go to Richard Daly next.

Louis Thompson:  I would say that the focus in the U.S. over the past several years has been on implementation of the 404 internal controls process of Sarbanes-Oxley.  It is a multi-million dollar process.  Companies are still reeling from the costs of that and really questioning whether the benefits are incommensurate with the costs.  If you look at XBRL, I think as companies realize the economies of this, they are going to have a far different appreciation of the benefits of XBRL compared with the costs.

James Glassman:  Thank you.  Richard Daly, group president of ADP.

Richard Daly:  Well, Jim, thank you.  I want to start by thanking Peter Wallison and Jim Glassman for the opportunity to be here today.  I thought I would start by trying to personalize a little bit what all this data really means. 

About 6 years ago, before my daughter turned 17, I had actually made a deal with her when she entered high school that if she kept her grades up and avoided teenage difficulties, when she was 17 we would talk about getting a car.  She kept her end of the deal and now here we are and it became time that I had to keep my end of the deal, so we reached an agreement that I would match certain money that she had saved and indeed found a car.  When we got the car, I gave her the dad speech about how significant this is, it’s very valuable, it’s very important, you really have to be careful, heaven forbid something happened to this, and certainly I would never want her to get hurt.  All of that is fine. 

About 3 days after the car was in the driveway, I got home from work and as I’m coming up the stairs she asked me to come into her room and she was on the Internet.  To my surprise, she had ADP’s financials from EDGAR Online and she was looking at the executive comp section.  She pointed out that, as pleased as she was to have the car and as significant as it was, somehow if something happened to the car, she thought we would still get by.

I can only imagine a world of XBRL how much easier it would have been for her to reach that conclusion and to research data.  I just hope someday my younger daughter doesn’t talk to me about my growth rate versus the other business units within ADP.

As a panelist today, both Peter and Jim asked me to really cover two hats.  First, as the most recent XBRL filer, what was it like for ADP?  And second, as the leading distributor of proxy information and financial information in the U.S., what do we really see that the benefits of XBRL could be; how do we see implementation going?  I really will spend most of my time talking about our experiences in that capacity.

Speaking first as an XBRL filer, it is relatively uneventful is what I could say.  We actually filed our 10Q today and somebody said, why don’t you try to compare it to SOX?  It’s almost beyond an apples and oranges comparison.  There really is no comparison.  SOX involved every staff person within ADP.  It was over a period of 18 months at least.  It was a regular topic at every executive committee meeting and every other activity.  In terms of XBRL, I asked to be on the emails that transpired.  We used outside consultants at a cost of less than $10,000.  It was a handful of people. 

I’ll call it a relatively-normal level of emails going back and forth about who was doing what, etc.  I would say with a limited amount of resources and a reasonable amount of focus on a few key financial people in our organization, it was what I would call an uneventful activity, and one in which even in my own internal activities – when I think about the amount of reports I get as an officer of ADP – I am actually looking forward to being able to analyze our own data, which at times seems to be almost a full-time job just to analyze the data of our own company.

I’m alive again?  All right.

Let me talk a little now about putting the other hat on.  As the leader/proprietor of shareholder communications and proxy voting services, we see tremendous potential for XBRL as an element of Chairman Cox’s vision for qualitatively improving the information experience, the issuers, and investors. 

Let’s talk a little bit about what the process is today.  There are 8,000 public companies, 87 million investors, of which these 87 million investors get to those companies through 800 banks and broker/dealers.  Over 13,000 shareholder meetings have to be managed annually.  One-third of those meetings occur in about a six-week period, which most people refer to as proxy season.  Investors need to be notified in a timely fashion.  Annual financial statements and proxy statements setting forth matters to be voted on need to be reviewed.  Votes associated with 300 billion shares need to be cast, tabulated, and accurately reported so that management, directors, investors in public companies can hold meetings and operate with effective corporate government.

ADP provides much of that plumbing and the technology behind the scenes to ensure that these processes work effectively and accurately.  Let me try to size that plumbing a little bit for you.  Right now we have 8 million lines of code, which have 300,000 function points.  A function point, to put into layman’s terms, means that something requires a decision.  So there’s 300,000 different activities that require a decision in this process.  We have 150 full-time systems programming professionals to support, maintain, and upgrade this, and that’s behind a dual operating system in dual locations with over 12,000 nips of computer support and hundreds of systems professionals supporting those activities.  So this is not a simplistic process; this is a rather complex process, bringing all these pieces together.

Thanks to a piece we call Proxy Edge – which is really close to being interactive data for institutions – XBRL will certainly take it to the next level.  Twenty-five hundred institutions today vote billions of shares automatically online; they analyze it; they vote their shares; and they do this in a very automated process.  If they were to try to do this manually, as they tried to do back in the late 1980s, early 1990s, the reality is, many of those votes would either never get voted or ultimately never make it to the corporations that need those votes. 

Technology is indeed something that we at ADP believe is our job.  We need to be aggressive appliers of real technology.  We are convinced that XBRL is real technology.  We are committed to aggressively apply that technology to take the process indeed to the next level.

Going forward, Chris Cox envisions a world of interactive data for retail investors, and as I’ve said, we share that vision:  successful shareholder communications and proxy voting with interactive data.  XBRL requires 3 major elements for issuers and investors.  Behind that, there are thousands of details in between that need not to get lost as we apply this new technology.  First, if you want to investors to have information, it is important to push it in front of them.  Studies clearly show that investors do not participate when they have to search for information and take steps to get it.  Today 11 million investors receive information electronically through investordelivery.com.  That’s ADP’s generic website out there.  They vote in high percentages.

We successfully work with the SEC staff to implement e-delivery, just as we did working with the SEC staff as successfully in the household.  This was all done on an opt-in basis so that no investor would be left behind.  To get to these 11 million investors, we sent out 3 billion notices.  Every time you receive a piece of paper through your broker or bank, there is a notice there that says go to www.investordelivery.com; save a tree.  We’ve tried all sorts of different techniques.  But the reality is, I think XBRL could be the magic bullet, because investors go to websites for their convenience, not for our convenience.  Getting better data is why investors go to the Internet right now.  XBRL could be that better data.

The second thing is, if you want investors to be informed, it is important to closely link proxy information right to the proxy ballot.  You want them to have it at the same time.  This information experience must be made easy and convenient.  People do not use the Internet again for our convenience.  They use it for their convenience because it offers more value to them.  Today 85 percent of investors look at the material that we sent to them right now in paper form.  If it is better information, we can get that 11 million closer to the 87 million investors and we can have them looking at it in more detail on a timelier basis with more benefit to them.

Third and finally, if you want efficient shareholder participation, it is important to closely link vote execution to information delivery.  Today 14 million investors representing 85 percent of Wall Street votes vote electronically through Proxy Edgy or through proxyvote.com.  They do this because we have made it easy through technology.  They didn’t do it because it saves companies money; they didn’t do it because it was more efficient for ADP to tabulate votes.  They did it because we made the technology easy for them and those who wanted to participate found this easier than participating with paper ballots. 

So let’s not lose sight of the end here.  The end is the investor.  They are the customer.  We have got to give them what they want.  XBRL, we believe at ADP, could again be that magic ingredient to have more of those 87 million investors join the 11 million investors right now who receive material electronically to go out there and get their information electronically.  We believe the key to successful implementation of interactive data is for investors to be aware and willing to serve U.S. markets or if it must meet the needs and preferences of individuals in terms of delivery, analysis, and execution, let’s not lose customer preference in this process.

Thank you.  You’ve covered it so much, Mike, I’m going to skip some of the data as it relates to XBRL.  I’m going to revise my data on explanation of XBRL based on what I heard today.

[Laughter]

Today, tremendous strives have been made in technologies for delivering shareholder information and for voting shares.  These strives are demonstrated by increasing levels of investor participation.  E-delivery and e-voting are fast and efficient.  E-delivery gets the information to investors.  It comes in an email message that they recognize and trust.  It comes from their brokerage firm tagged as their brokerage firm, a source that they are confident in.  E-voting is bundled with information delivery, so that voting is informed.  But today, the technology of information presentation and analysis is admittedly not at the same level of sophistication that we bought e-delivery of that information and e-voting. 

Electronic information is largely delivered in PDF and HTML files today.  We all know what it’s like wading through a PDF document page-by-page, trying to find information about a company in which you invest.  It can be slow and it can be frustrating.  A PDF document is not easily searchable.  It’s searchable, but I wouldn’t use the term “easily,” although as I pointed out, my daughter managed to find where she needed to get to for my information. 

Comparisons between peer companies are manual, or at best, you have to leave the current session and go to another website to find or calculate comparative information.  This causes an interrupt or, as retailers on the web are well-known, an abandoned shopping cart.  We don’t want to abandon investors in here.  We don’t want lower shareholder participation.  This is why we believe that today over 2 million investors that originally agreed to receive their information from us electronically, subsequently came back and asked us to get back to paper.  We believe XBRL will answer their needs and not give them this frustrating experience.

In an interactive data world, with XBRL, the information experience for investors is qualitatively improved.  For example, say as an investor you are looking at an annual report and you want to review the company’s information on capital spending.  For issuers, that file, financial reports in XBRL.  For issuers in XBRL, you can easily enter a search for that information and much other information. Thanks to XBRL taxonomy, you can examine key ratios, and these are automatically calculated for you.  An example, up here I’m showing Cap X as a percentage of revenue.  I’m also showing Cap X as a percentage of free cash flow.  This is one of thousands and thousands and thousands of examples of how people will be able to research information and what’s important to them.

Once there is a history of such filings and greater participation by corporate issuers in filing XBRL, even more possibilities open up for qualitatively improving the investors’ information experience.  For an example, an investor could more easily examine trends in capital spending over time, as the examples up here show.

Let’s talk about informed voting by shareholders.  Today XBRL taxonomies have been developed for most elements of financial statement reporting.  The need is for more companies to file in XBRL.  Based on ADP’s first time, though, we don’t think it’s going to be a big deal, and we suspect that benefits and turns are more efficient, finance and accounting processes made are going to be well worth the effort.  Again, when I talk about my own experiences inside of ADP, on a monthly basis I get no less than 3 inches of paper, of all of ADP’s internal reporting documents.  It is virtually impossible for me to get through those.  As I think about filing those documents on an XBRL basis, the information that I would like, comparing my own data to previous activities, will benefit me and will also benefit investors. 

The taxonomy for the proxy statement, however, have not yet been defined.  One of the things we are looking to do at ADP, given our role in the proxy process, is help lead that with folks like Niri, with folks like the Society of Corporate Secretaries, the SIA, and others, and lead to collaborate with the proxy process participants in developing and quickly advancing taxonomies for the proxy statement or, as Mike pointed out, it’s just about the XBRL being applied to a different set of definitions, a different set of standards.

We could envision how search and comparison capabilities for proxy statements would be useful for issuers and investors in the voting process.  It would be easier to find and compare important information, that be it on stock performance, directors, board or committee reports, executive comp, or other proxy information.  With such information in hand, investors could also quickly and efficiently cast their votes.  By pushing interactive data, investors who recognize the benefits would participate at greater levels and continue to participate and increase in participation.

Let me see if I can just wrap up here.  Rolling out a new technology is both a science and an art.  Mike really covered the science.  It’s real work and it has to work right.  The art is, investors need to be part of it.  Every study shows that if you push data in front of investors, they are more likely to look at it.  The better the data, the better the participation.  We are committed to work with technology leaders and the SEC just as we have for the last two decades to enable investors to be better informed and to continue to participate at even greater levels. 

We believe that XBRL will be more effective than the evolution we helped lead, whether it be with electronic voting or with house holding, because in this case, it is in the investor who is the customer here, who is going to be far better informed, and it is going to be their benefit to participate as well.  We are willing to put our money where our mouth is.  The easy part was the filing of ADP and XBRL.  We think the more important part is making every one of those 87 million investors aware that XBRL is available, available to them, and can indeed empower them as investors, as Chairman Cox said.  We are committed to do that – not over the next decade, but over the next months and years.  Thank you.

[Applause]

James Glassman:  Thank you, and thanks for telling us about your own experience with ADP and XBRL and your daughter.  I just want to highlight one thing.  Let me ask you a direct question.  There is a lot of talk about the SEC moving toward mandated e-delivery or e-voting.  If that happened tomorrow, before XBRL was being extensively used, what do you think the response of shareholders would be?

Richard Daly:  The studies right now, particularly the Forrester studies, show that investors today, 92 percent of them, prefer to receive their financial information in paper.  There is a variety of reasons for that.  There are security reasons; it’s the convenience; it’s the portability.  But I really believe is that the real reason is that it’s no better than the data that they would be forced to go and get online by them searching for it, versus the convenience of the data being pushed to them today.  So, I think if it was forced today, it would be premature, and investor participation would go down, which I know Chairman Cox and the SEC overall is very concerned about happening.

The idea of moving to an XBRL, though, where better data is out there, combined with pushing it out there, we think is the one-two punch that would work best for the roll-out and for investors.  Now what our plan is, in a world of XBRL, through our broker clients, we have millions of emails, which right now people have opted not to use to receive their information. 

We think in a world of XBRL, and even prior to a world of XBRL we can convert to a dual distribution model, which is if you haven’t opted to receive electronic delivery, we will still push the paper in front of you, but simultaneously we will push you an email with the same data.  That will get people used to receiving the data electronically.  As we convert to XBRL, as they open those emails, they will say, gee, this is actually better information than I am getting in the paper.  And by making it very easy with a mouse click or two, to turn off the paper and become a bona fide opted-in electronic receiver, we think we will again get that 11 million number much closer to the 87 million over a much shorter period of time.

James Glassman:  Thank you very much.  Mark Schnitzer of Morgan Stanley.

Mark Schnitzer:  Thank you very much for the invitation to speak here today on behalf of Morgan Stanley.  I would like to speak with you all today about how interactive data is going to help analysts and investors to make better-informed investment decisions.

At Morgan Stanley, I manage a global team that’s responsible for building more robust valuation models.  Our research analysts use a system we created called ModelWare to categorize the data in their valuation models on a consistent basis and validate the data using a set of rules.  This whole system is based on XBRL, and our analysts don’t even know that they are using XML or XBRL.  Our research, management, accounting valuation experts, and firm management have been committed to making XBRL and interactive data a reality since joining XBRL in 1999, and we are making this investment in XBRL to enable our clients to make better-informed investment decisions.  So let’s take a look over the next few minutes at how interactive data will help investors.

As we all know, accounting disclosures vary by jurisdiction, by the industry a company happens to be in, by individual companies inside of an industry, and even forgiving company disclosures change from one quarter to the next, from one year to the next.  The task for taking all of this data as an analyst or as an investor and tying it together to compare data across companies or even across one company over time is an onerous task, very difficult to do even using the technologies we have today. 

Interactive data is going to enable us to compare data across all these different dimensions.  It’s not going to eliminate all the work.  It will not make financial analysis be a one-click process; however, it will alleviate a lot of the tedium of comparing data and enable analysts and investors to spend more of their time on making informed investment decisions based on solid analysis, not on who can build the fastest and most interesting spreadsheets.

Business reporting as we know it today has made only modest advances since the current regulatory regime was put in place in the period after the Great Depression.  Even though earnings releases, 10Ks, and 10Qs are now available online rather than just on paper, they are not actually anymore understandable or not significantly more understandable today than they were back in the 1930.  With the EDGAR Online, with the access to the EDGAR system, with the SEC system, investors still have the same limited access to data that they have always had.  As a result of [indiscernible], Chairman Cox made the point before, a lot of the data that the investors work with that’s re-keyed from 10Ks, 10Qs, and other documents is error-prone.

So let’s look at sell-side research today.  When a company issues an earnings release, it is basically a race amongst the sell-side analysts to put out a quick note for the buy-side, basically re-telling the story of what the company just reported in their quarterly earnings report.  As a result, if you had a second or third analyst that’s putting out your comment on that quarterly earnings, the buy-side doesn’t really pay any attention to it, because they have already seen the reports.  Even if the third went out and you had interesting points that you picked up from the earnings release, the chances of the buy-side actually picking up on that and pulling information from that are actually surprisingly low because the email got deleted from the inbox before it actually got read.  That’s the reality of how the quarterly reports get disseminated in the marketplace today.

As you can see in the lower report over there, we also provide more thematic pieces on a company over a longer period.  The analysts frequently supplement the data from the regulatory filings with data that they pick up either through surveys or doing additionally work, really looking at company-specific data from other sources.  This whole world is, of course, going to change with interactive data.  I can slide back a couple of slides.  So the current workflow that we discussed is that data gets reported in HTML or PDF.  The analyst or vendor then takes the data from that HTML or PDF document and re-keys that data or copies and pastes that data into a spreadsheet model.  Then analysts race to get the information out. 

With XBRL the process changes significantly.  Basically, when the information is available in XBRL as Mike showed nicely before, there are already software tools available that will automatically load that data on a tagged basis directly into your Excel spreadsheet, that you can then spend your time actually coming up with investment insights and getting your investment note out rather than simply re-keying the data.

I’d like to talk a little bit about some of the work that we can do with interactive data to actually incorporate interactive data directly into financial analysis that goes on.  These next couple of slides that I’m going to run through show you a little bit of how we actually internally use the tagging of financial data to actually enable us to look at the data on a consistent basis for companies that we cover around the globe.  In this particular instance, I’m showing you a chart that shows various components of obligations of a company and how it would be categorized across time. 

Here on this slide, you can see if you just look at what is the reported debt for these various companies that are shown below, it shows that some companies have more net debt and other companies have less net debt, but that picture is actually misleading.  When you add in the operating leases and capitalize the operating leases, which is a component of how we do our uniform calculations for all companies, you see a very different story than when you were just looking at the net debt as reported in the face financial statements.

Here we are looking at pension plans, pension obligations.  It says basically the same story that we were just looking at a moment ago with the leases, where, when you really drill into the data as it is reported and take that a step forward and look at that data with the adjustments that you would want to make, you see a very different picture than you see just on the surface with the basic financials. 

So what we are going to see with XBRL and interactive data is the ability to provide not just a spreadsheet analysis that we discussed earlier but the ability to apply model constructs on a very consistent basis across a disaggregated information that is reported through the interactive data for companies regardless of the accounting jurisdiction that they happen to be in.

Interactive data is going to have a significant impact on the various analyses that we’ve put together.  At Morgan Stanley, we’ve put together not only the company reports that I was just showing you a few minutes ago, we also put together thematic pieces.  These thematic pieces have taken literally months to put together as we collect data from companies across all industries or look at data across hundreds or in some cases several thousand companies.  The analyses that we are doing are very limited because we constantly have to make a trade off between the data collection time period and our ability to get a report out into the marketplace. 

We are working on a thematic piece, whether it’s discussing pensions, Job Creation Act, or whatever the report is, if we take too long to collect the data and somebody else gets out with the report before we do, we have just diminished the market value of our report significantly.  Getting that report out in a timely fashion is critical.

Interactive data enables to pull in