Disclosure for Backtested Performance Information on the Simulated Strategies of Index Portfolios:

1. Index Funds Advisors, Inc. (IFA) was incorporated in March 1999 and placed its first independent client investments in early 2000. The performance information presented in the chart or table represents backtested performance based on combined simulated index data and live (or actual) mutual fund results from Jan 1, 1927 to period ending date shown using the strategy of buying, holding and annual rebalancing globally diversified portfolios of index funds. Backtested performance is hypothetical (it does not reflect trading in actual accounts) and is provided for informational purposes to indicate historical performance had the index portfolios been available over the relevant period. IFA did not offer the index portfolios until November 1999. Prior to 1999, IFA did not manage client assets. The IFA indexing investment strategy is based on the principles of the Modern Portfolio Theory and the Fama and French Three Factor Model for Equities and Two Factor Model for Fixed Income. Index portfolios are designed to provide substantial global diversification (approximately 15,000 companies in 35 countries) in order to reduce investment concentration and the resulting increased risk caused by the volatility of individual companies, indexes, or asset classes. Client portfolios are monitored and rebalanced, taking into consideration risk exposure consistency, transaction costs, and tax ramifications to maintain target asset allocations as shown in the twenty index portfolios.

2. A review of the IFA Index Data Sources and IFA Indexes Time Series Construction is an integral part of and should be read in conjunction with this explanation of backtested performance information. For detailed descriptions and definitions of the underlying criteria and data used to construct backtested performance, see these two links: Data Sources and IFA Indexes Time Series Construction. Simulated index data is based on the performance of indexes as described in the Data Sources page. The index mutual funds used in IFA’s twenty index portfolios are IFA's best estimate of a mutual fund that will come closest to the index data provided in the simulated indexes. Simulated index data is used for the period prior to the inception of the relevant live mutual fund data an equivalent mutual fund expense ratio is deducted from both live and simulated data. Live (or actual) mutual fund performance is used after the inception of each mutual fund.

The IFA Indexes Times Series Construction goes back to Jan. 1927 and consistently reflects a tilt towards small and value equities over time, with an increasing diversification to international markets and the real estate index as data became available. In Jan. 1927, there are 5 equities indexes and 2 bond indexes, in Feb. 1955 there are a total of 10 indexes, and there are 15 indexes in March 1998 to present. If the original 5 equity indexes from 1927 are held constant until July 2005, the annualized rate of return is 11.23%, after the deduction of a 0.9% IFA advisor fee and a standard deviation of 25.59%. The evolving IFA Indexes over the same period have a 12.02% annualized return after the same IFA fees and a 25.18% standard deviation. It is IFA's advice that the value of having a longer time series exceeds the concerns of index substitutions over the 1927 to present period. Due to the very high standard deviations of returns (25%) a 60 year or more sample size of data is recommended to reduce the standard error of the mean. In other words, in IFA's opinion, smaller sample sizes introduce larger errors than the errors introduced by stitching together indexes over time. This is the advice IFA provides to it's clients. Click HERE to see the analysis of the evolution of these portfolios.

Backtested performance is calculated by using a computer program and monthly returns data set that starts with the first day of the given time period and evaluates the returns of simulated indexes and index mutual funds, see Data Sources. In 1999, tax-managed funds became available for many different index funds. IFA uses tax-managed funds in taxable accounts. The tax-managed funds are consistent with the indexing strategy, however, they should not be expected to track the performance of corresponding taxable funds in the same or similar indexes. As such, the performance of portfolios using tax-managed funds will vary from portfolios that do not utilize these funds.

3. Backtested performance does not represent actual performance and should not be interpreted as an indication of such performance. Actual performance for client accounts may be materially lower than that of the index portfolios.Backtested performance results have certain inherent limitations. Such results do not represent the impact that material economic and market factors might have on an investment adviser's decision-making process if the adviser were actually managing client money. Backtested performance also differs from actual performance because it is achieved through the retroactive application of model portfolios (in this case, IFA’s twenty index portfolios) designed with the benefit of hindsight. As a result, the models theoretically may be changed from time to time to obtain more favorable performance results.

4. Index portfolios 10, 30, 50, 70 and 90 were originally suggested by Dimensional Fund Advisors (DFA), merely as a example of globally diversified investments using their many custom index mutual funds, back in 1991 with moderate modifications in 1996 to reflect the availability of index funds that tracked the emerging markets asset class. Portfolios between each of the above listed portfolios were created by IFA in 1999 by interpolating between the above portfolios. Portfolios 5, 95 and 100 were created by Index Funds Advisors in 1999, as a lower and higher extension of the DFA 1991 risk and return line. Due to the high similarity of the 1999 versions of index portfolios 95 and 100 to index portfolio 90, the 95 and 100 portfolios were moderately modified in November 2002 to have higher exposure to small and value equities through out the world. According to the extensive research of Eugene Fama, Kenneth French and Jim Davis, utilizing data from the Center for Research of Security Prices (CRSP) over a 68 year period from July 1929 to June 1997, this change has higher risk and return expectations than the previous versions of 95 and 100. (see Characteristics, Covariances, and Average Returns: 1929-1997) In January 2004, IFA changed the computer program setting to calculate annual rebalancing on the various indexes in the index portfolios. Previous to that they were rebalanced monthly. Annual rebalancing is closer to the actual rebalancing of client accounts, therefore it was adopted as the new method in January 2005.

5. Backtested performance results assume the reinvestment of dividends, ordinary and capital gains and annual rebalancing. The performance of the strategy reflects and is net of the effect of IFA’s annual investment management fee of 0.9%, billed monthly. Monthly fee deduction is a requirement of our software used for backtesting. Actual IFA advisory fees are deducted quarterly, in advance. This fee is the highest fee IFA has ever charged. Depending on the size of your assets under management, your investment management fee may be less. Backtested risk and return data is a combination of live (or actual) mutual fund results and simulated index data, and mutual fund fees and expenses have been deducted from both the live (or actual) results and the simulated index data.

Although index mutual funds minimize tax liabilities from short and long term capital gains, any resulting tax liability is not deducted from performance results. Performance results also do not reflect transaction fees (as seen here) and other expenses charged by broker-dealers, which reduce returns. IFA is not paid any brokerage commissions, sales loads, 12b1 fees, or any form of compensation from any mutual fund company or broker dealer. The only source of compensation from client investments is obtained from asset based advisory fees paid by the client.

More information about advisory fees, expenses, no-load mutual fund fees, prospectuses for no-load index mutual funds, brokerage and custodian fees can be found on the HERE and on the Fee link in the gold navigation bar below and on every page of this internet site.

6. For all data periods, annualized standard deviation is presented as an approximation by multiplying the monthly standard deviation number by the square root of twelve. Please note that the number computed from annual data may differ materially from this estimate. We have chosen this methodology because Morningstar uses the same method. (see IFA Indexes Time Series Construction)

7. Not all of IFA clients follow our recommendations and depending on unique and changing client and market situations we may customize the construction and implementation of the index portfolios for particular clients, including the use of tax-managed mutual funds, tax-harvesting techniques and rebalancing frequency and precision. In taxable accounts, IFA uses tax-managed index funds to manage client assets. However, the tax-managed index funds are not used in calculating the backtested performance of the index portfolios, unless specified in the table or chart. Some clients substitute the mutual funds recommended by IFA with investment options available through their 401k or other accounts, thereby creating a custom asset allocation. The performance of custom asset allocations may differ materially from (and may be lower than) that of the index portfolios. See composite returns for All Accounts and for just Tax Deferred Accounts. Note that the last few pages provide descriptions of calculations used in various columns of the composite returns reports.

8. Performance results for clients that invested in accordance with the Index Portfolios will vary from the backtested performance provided on the site due to market conditions and other factors, including investments cash flows, mutual fund allocations, frequency and precision of rebalancing, tax-management strategies, cash balances, lower than 0.9% advisory fees, varying custodian fees, and/or the timing of fee deductions. As the result of these and potentially other variances, our clients have not and are not expected to have achieved the exact results shown since November 1999, when we placed our first investment. Actual performance for client accounts may differ materially from (and may be lower than) that of the index portfolios. Clients should consult their account statements for information about how their actual performance compares to that of the index portfolios.

9. As with any investment strategy, there is potential for profit as well as the possibility of loss. IFA does not guarantee any minimum level of investment performance or the success of any index portfolio or investment strategy. All investments involve risk (the amount of which may vary significantly) and investment recommendations will not always be profitable.

10. Past performance does not guarantee future results.

11. WHY GO TO ALL THIS TROUBLE?

This type of analysis is important because a shorter time period introduces a large statistical sampling error for both risk and average returns. Past performance does not predict future performance, however, analyzing 30 years or more of simulated risk and return data is a more reliable source of information concerning the cost of capital for firms and their shareholders and the resulting expected returns for investors who trade their cash for shares and bonds of those firms. That is the essence of capitalism.

The result of this data is a probability distribution with an average return and a standard deviation around the average, which best characterizes future random events that are totally unpredictable like the roll of the dice or flip of a coin, yet these random events over long time horizons, like 30 years or more, accumulate to new distributions. These distributions are, to varying degrees, similar to a large sample of previous distributions, such as 30 years. Shorter time horizons demand lower risk investments, while longer time horizons allow for regression to the mean. The "mean" refers to the average expected outcome of returns, which is also the most probable outcome. The distribution of historical market data is a leptokurtic distribution, meaning it is not conclusive in any way as to the limits of losses or gains (see Leptokurtic distributions). The dice and coin flip does have limits, but the market does not. There is an unlimited risk on stock market investments that can not be clear in even very long term historical data. For example, in the stock market crash of 1929, the market declined 89% and many investors had leveraged their capital and lost all of their investment. The stock market is a risky investment and investors can lose all or nearly all of their money because of the risk of firms going out of business, general macroeconomic and political risk, and challenges to the ideas of capitalism in general.

However, this analysis is far more useful than the traditional 1, 3, or 5 year returns and risk data used by the great majority of individual and professional investors. Without such longer term analysis, investors would be merely speculating on the risks and expected returns of their investments with a statistically unacceptable sample, like a gambler in a casino hopelessly trying to beat the casino statistician, who may be referred to as the dice, card, and roulette wheel actuary. This is in fact what investors do and several of studies have confirmed it is the source of their near zero average returns over the last 17 years, after inflation and taxes. As Louis Bachelier stated in the first published paper on the random character of stock market data, The Theory of Speculation (1900), "the expected return of speculation is zero." Statistically speaking, investors have a relatively high standard error of the mean (average return) with data of less than 30 years.

Because Index Funds Advisors is recommending mutual funds that correlate to the investment criteria of the simulated index data, there is a greater chance that the data is useful to index funds advisors than it is to actively managed mutual fund advisors that do not replicate the index and therefore engage in style drift. Past performance for active managers is an especially poor indicator of future results, due to the relatively small number of years of performance data available for each active manager and the fact that even during that period they are style drifting.

This analysis and investment strategy is consistent with the Modern Portfolio Theory, which is the term used to summarize the combined research of Harry Markowitz, William Sharpe and Merton Miller. They were awarded the Nobel Prize for Economics in 1990 for their efforts to describe how financial markets work and how to build efficient portfolios.

12. Your use of this site is acknowledgement that you have read and understood the full disclaimer. Index portfolios times series standard deviations and returns source: DFA FA Returnw. © Copyright 1999-2005.