INDEX FACTS Assets Benchmarked (billions) Total assets........$2,000 Passive assets.......$450 Inst. assets........$1,500 ETFs.......................$27 Mutual Funds.........$320 ================= Number of Indexes....22 INVESTMENT FACTS Assets under Management (billions) Total Assets..........$131 US Mutual Funds...$20.6 ================= Number of Funds.....180 FUTURES & OPTIONS INFORMATION Futures traded at: CME, NYBOT, Eurex US, CFE Options traded at: AMEX, CBOE, ISE, PHLX, PCX Q: What are the major goals of your indexes? A: Russell's primary business is to research investment managers and fund managers to help our clients figure out which are suitable for their investment goals. We have a team of money manager analysts whose full-time job is to interview portfolio managers of separate accounts and funds. Just as Consumer Reports has their measurement tools to determine the quality of products, Russell has developed in the same way its own set of measurement tools on money managers. We built the broad-market Russell 3000 Index, which represents the largest 3,000 US companies by market value, as a measurement of the broad US equity market. We called all index fund managers in the United States and asked them how many stocks they could actually buy. The answer uniformly came back between 3,000 and 3,200. We kept two distinctive principles in mind when we started building the Russell 3000 Index. First, we not only help our clients with the selection of money managers, but we also work with them on asset allocation. Because our clients like to allocate assets selectively in US stocks or in non-US stocks, we constructed the Russell 3000 as an index that contains companies from only the United States. Second, we realized the importance of adjusting the ownership of the companies against how much of that was free floated. We applied this principle in cases of large privately held portions, which are not coming into the market any time soon, or for cross-ownership when a certain company holds a big chunk of another company. Q: What do you consider when building your indexes? A: The idea behind all of the enhancements to our products is to deliver easier and more appropriate tools for investors. We evaluate whether something is an improvement to our products by determining whether it makes them more representative of the universe of securities from which a group of managers might select. These groups encompass the mutual fund portfolio manager community as well as the institutional separate account community. We have licensed Barclays Global Investors to create iShares, their brandnamed ETFs, which are based on the Russell indexes. Presently, there are 12 iShares ETFs with $27 billion worth of assets invested in the fund family. In terms of ETFs, that shows that there are more dollars invested in ETFs for the Russell family than there are for the Dow or the NASDAQ family. In both 2003 and 2004 we had substantial asset growth in net cash flow into the Russell family of ETFs. Q: How do the Russell indexes differ from other benchmarks? A: If we compare the Russell 3000 Index against the S&P 1500 Index, we see that, obviously, the S&P 1500 has half as many stocks, so it is not as widely representative as the Russell 3000. In addition, S&P has quality screens on its indexes to make them in a way representative of credit worthy companies. At Russell, we represent the total market, whereas Standard & Poor's includes only what it views as high quality stocks. For example, Amazon.com is not in the S&P 1500, whereas it is in the Russell 3000 Index and the largecap Russell 1000 Index. While our goal is to represent the investable universe of securities, S&P's goal is somewhat different. The Wilshire 5000 Index is another appropriate example for highlighting certain differences between benchmarks. That index includes constituents that get priced on a given day, or in other words, it includes securities traded by appointment. In contrast, we want to know what constitutes the investable universe. The main difference is that the Wilshire 5000 is aimed at trying to cover everything, while the Russell 3000 focuses on everything you can buy. Furthermore, the Russell 3000 Index has been adjusted for free float since its inception in 1984, while the S&P and Wilshire are moving to free float over the course of the next 12 months. Their move in the same direction makes us believe they have realized that adopting this methodology is the right thing to do. Q:Will you describe the key differences between the Russell indexes? A: As some investment managers focus on large-cap stocks, others concentrate on investing in small-cap stocks. So, we divided the Russell 3000 into two pieces: the Russell 1000, which represents the largest 1,000 companies in the United States, and the small-cap Russell 2000 Index, which includes the stocks starting from 1,001 to 3,000. Again, we determined this distinction by talking to our money manager analyst team. We wanted to find out from them what constitutes the correct division between large-cap and small-cap. The feedback showed us that typical large-cap investment managers invest not only in the largest 500, but they really go down to the largest 1,000 stocks. Thus, our large-cap index is the Russell 1000 and the Russell 2000 turned out to be known across the industry as the small-cap benchmark. The next distinction is based on investment style. As our analysts identified in the mid-70s that money managers specialize in growth and value stocks, we decided to build style indexes that represent the universe of securities from which managers select. Since then, we have been measuring growth and value managers against growth and value indexes. We use those indexes to evaluate the money managers that go into the mutual funds we sell through RIAs. For example, if a money manager has a large-cap growth assignment in one of our mutual funds, he is being evaluated against the Russell 1000 Growth Index. Likewise, a smallcap value manager is being evaluated against the Russell 2000 Value Index. Q: Currently, how widely used are the Russell indexes? A: We do a survey on index funds once every six months when we approach all index fund providers. Data from our research as of June 30 shows that there were $450 billion, including $27 billion in ETFs, worth of assets in Russell index funds across the industry. In a broader context, the institutional assets, the passive assets and the mutual fund assets benchmarked against the Russell family of indexes now exceed $2 trillion in total. In the ETF world, the Russell 2000 Index ETF is the largest, while in the passive world the most commonly used index for index funds is the Russell 3000, which appeals to investors because of its broad market aspects. The second most commonly used benchmark in this category is the Russell 1000 Index. As for derivatives, there are now futures traded against the Russell 2000 and the Russell 1000. There are also options traded on the ETFs on our indexes as well as options traded on the indexes themselves. Q: Could you provide the sources of information that you use to estimate the popularity of your indexes? A: In general, we look at two sources of information. The first of them is Nelson’s Marketplace Database, which is a yellow pages service for institutional products offered to large institutions such as pension funds. Most of the products in that database are listed along with their benchmarks. There are currently more assets in more products benchmarked to the Russell family of indexes than any other family of indexes. Another source of data is Morningstar’s DataLab, which has a list of benchmarks for most mutual funds. According to their most recent data, there are 722 funds benchmarked to the Russell family of indexes. Needless to say, S&P is still the big player there, but we are starting to make inroads into the retail space, both in the ETF and mutual fund arena. Q: How has the significance of the Russell indexes changed historically? A: Around 20 years ago, the way a stock began developing an institutional following and institutional holdings was with the help of a sell-side analyst who would start covering it. In recent years, however, we have noticed a change, defined by four distinctive trends. First, we have seen a reduction in the number of sell-side analysts. Second, the remaining sell-side analysts have reduced the number of stocks that they cover. Third, portfolio managers have become more benchmark-sensitive. Finally, we have seen the popularity of Russell indexes increase as a whole. As a result, we have reached the conclusion that the way a stock now starts developing institutional ownership is as it becomes a member of the Russell 3000. As some of the new small stocks, ranging from 200 to 300, join our index each year, many of them show on the radars of various portfolio managers because they are now part of their benchmarks. At this moment, portfolio managers start looking at the stocks and sometimes they buy them. Q: Will you discuss the adjustment methodology and process? How often do you adjust your indexes? A: Some index providers adjust their indexes by committee. Russell's methodology is rules-based and objective. Our rules are available on www.russell.com. Our methodology is strictly based on three rules that determine the ways in which the Russell 3000 undergoes changes. First, we have a process called index reconstitution, which takes place once a year in June. The purpose of this process is to re-determine what the 3,000 largest companies in the United States are in order to reflect the changes in the marketplace. This rule ensures that our indexes are representative of the market. By implementing our second index change rule, we make sure that we delete companies that have disappeared due to different corporate actions over the course of the year, but we do not replace these companies in the indexes until the following June. Finally, the third rule comes into effect when, once a quarter, we review the IPOs that have come out in the previous quarter and determine if any of them are big enough to be in the Russell 3000. If these IPO companies meet our criteria, we will add them to the indexes at the end of the quarter with an announcement on the additions a month in advance. As we remain dedicated to the representativeness of our indexes, we have considered changing the frequency of the adjustment process. However, our research with our client advisory board, which consists of people from the index fund community, the pension plan sponsor community, the brokerage community, the active money manager community and the consulting community, indicated that its members were pleased with the way we conduct the reconstitution process. Q: What measures have you put in place to overcome the problems associated with index adjustment? A: First of all, we believe that transparency and objectivity are crucial principles for our business. We follow these principles by ensuring that we publish our rules well in advance and provide information related to the probable consequences from the adjustments within our indexes. Each year in May and June we receive calls from companies that try to convince us not to delete them from our indexes. By staying true to our discipline, however, we eventually remove between 5% and 10% of these companies from our indexes. We are currently working on an innovative tool called provisional indexes, which are going to calculate the rate of return on a portfolio as if it were reconstituted in the process of making the reconstitution announcement. Thus, an investor who wants to reconstitute on June 14 will be able to mix the provisional index with the actual index to come up with the correct benchmark. Additionally, we are going to publish the legacy index that will calculate the return of the old index for a month after the reconstitution process. With this tool an investor will be able to reconstitute after the fact. The message for our clients is: Choose the day on which you want to do the reconstitution. Some people will choose a day other than the official reconstitution day for various reasons. Some people will switch the reconstitution day to the last Friday of June, while others will stick to June 30 because it is a suitable day for accounting purposes. As a result, the reconstitution effect will be spread out over time.