Can online investors outperform the NASDAQ-100?

Willey, Thomas

ABSTRACT

The phenomenal growth of the Internet and online brokerage firms has lead to the development of the capability of the individual investor to buy and sell stocks online. The daily top ten purchasing and selling decisions for one of the largest online trading firms, Ameritrade, are used to create a portfolio for sixty trading days. The returns for the online investors fall short of those of a market index, even before adjusting for transaction costs and taxes. These results do not support the existence of a contradiction to market efficiency, or return anomaly, for this strategy of active investing.

INTRODUCTION

Online investing is one of the fastest growing areas of the personal finance landscape today. A recent survey from U.S. Bancorp Piper Jaffray, estimates that trades conducted online by individual investors grew from 17% of total trades in 1997 to 48% at the end of 1999, a compound growth rate of 68% (Carey, 2000). This extraordinary growth, combined with a similar acceleration in personal investing resources on the Internet, has allowed individuals to actively manage their own portfolios. Prior to the recent phenomena of Internet investing, the majority of financial researchers have examined the ability of printed periodicals and the broadcast media sources to generate returns that exceed the overall stock market returns. Barber and Loeffler (1993) examined the Wall Street Journal's regularly appearing dartboard stock picks versus analyst's recommendations and found positive abnormal returns and changes in trading volume around the publication date of the article. The ability of participants in Barron's Annual Roundtable to recommend stocks that would outperform the market was analyzed by Desai and Jain (1995). A positive abnormal return of approximately two percent (1.91%) was shown for the fourteen day window from the announcement date to the publication date, but this reverts to zero in post-publication holding periods of one to three years. Mathur and Waheed (1995) found a similar positive excess return (2.35%) when the "Inside Wall Street" column's stock recommendations from Business Week magazine were purchased by investors. Stock recommendations from the PBS weekly show, Wall $treet Week, were found to exceed the market return by 1.1% on the first trading day after the airing of the show (Griffin, Jones and Zmijewski, 1995). In each of these previously published studies, the effect on stock prices is positive around and on the event date and then the abnormal return disappears as the new information is incorporated into the new stock price.

Research on the impact of the Internet and the enormous volume of information, both good and bad, on stock prices and returns is at the beginning stages of financial research. McQueen and Thorley (1999) investigated the ability of the Motley Fool Website's Foolish Four Portfolio (www.fool.com) to outperform a market portfolio. This portfolio uses a variation of the highly publicized "Dogs of the Dow" high dividend yield investment theory to select stocks and uses an end-of-the-year portfolio rebalancing. This adjustment reflects changes in the relative ranking of the thirty stocks that make up the Dow Jones Industrial Average. The above authors found below-market returns for investors who followed the suggested strategy. Barber and Odean (1999) examined 1,607 investors who switched from trading over the phone to trading over the Internet from 1991 to 1996. Their results indicated a net decrease in returns of approximately five percent for the sample group. Before the switch, the phone investors outperformed the market by more than two percent annually, while after the move to online trading the same investor's performance lagged behind the market by more than three percent. An empirical investigation of The Motley Fool buy-sell recommendations is contained in a recent article by Hirschey, Richardson and Scholz (2000). Using an event study methodology, the three authors examined the effect of recommendations on the performance of five different portfolios recommended at the well-known website. Advice-to-buy resulted in average 1.62% increase in stock prices on the announcement date, while sell recommendations lead to a loss of 1.49% when posted to the website. Hirschey, Richardson and Scholz (2000) conclude, "... announcements are more newsworthy than second-hand buy-sell recommendations published in traditional print and electronic media." The impact of information available on Internet is quickly becoming a major force in the personal investment decisions made in financial markets of today and the focus of this research.

The study is organized as follows. The second section presents an overview of online investing and the Ameritrade Online Investor Index (AOII). The following part discusses the methodology used in creating the "AOII Net" portfolio. The portfolio results are presented in the fourth section. The study is summarized and the impact for investors is discussed in the final segment of the paper.

ONLINE INVESTING AND THE INTERNET

One of the most important developments in the field of finance, and frequently one of the most controversial, is the notion of the efficient markets hypothesis (EMH). In summary, the EMH describes how quickly new information about a company is incorporated into the current stock price and if investors can use information to outperform the market. Proponents of this theory believe the U.S. financial markets are very informationally efficient and therefore, recommend most investors employ a buy-and-hold strategy of a mutual fund that mirrors a major stock index. At the other end of the spectrum, many empirical studies have found pockets of inefficiency or average-return anomalies. Fama (1998) discusses the shortcomings in tests of asset pricing models, such as the Capital Asset Price Model and the Arbitrage Pricing Theory, which have uncovered patterns in historical returns of assets. In Fama's research, most of the long-term anomalies are explained away with changes in techniques, while many of these patterns have been readily accepted and adopted by the investing public. An example of one of these inefficiencies is the day-of the-week effect. Corrado and Jordan (2000) found the average daily return for the Standard & Poor's (S&P) 500 for Mondays for the period of July 1962 to December 1994 was -0.078%, compared to positive daily returns for each of the other days of the week. Similar results have been found in international markets and also in markets for other assets. No definitive explanation has been offered for these above results that were first documented in the early 1980s.

This example of a contradiction to market efficiency leads to the overall purpose of this study. Can a stock portfolio, consisting of the top net daily positions of a random sample of online investors, provide a daily return that will exceed the return generated by a market index for the sample period?

The Ameritrade Online Investor Index

One of the major online brokerage firms, Ameritrade, has began to publish the Ameritrade Online Investor Index (AOII), a daily measure of the amount of buyer participation based on a decisions made by the firm's online investors (Ameritrade Press Release, 12/1/1999). On every trading day, after the U.S. markets have closed, Ameritrade posts the Ameritrade Index page on the Internet. Appendix A presents the Ameritrade Index Webpage for the first day of the study. One of the goals of the index is to measure the individual investment decisions of online investment individuals. The AOII is presented as the percent of online traders that were buyers, and is found by dividing the number of buyers of equities by the sum of buyers and sellers of equities. For the initial day of the study, the AOII was reported as 37.68%, which indicated approximately 38% of all buyers and sellers would have been buying stocks and the remaining 62% would have been selling equities.

[FIGURE 1 OMITTED]

The study period for the AOII data begins on February 1, 2000 and ends on April 26, 2000, a total of sixty trading days. The maximum value for the AOII of 89.03% and therefore, the strongest bull sentiment for the study period was April 12. This strong buy signal occurred two days before the market index fell to its lowest point of the study on April 14. The strongest bearish value for the index of 14.18% was on February 23, which indicates that approximately 86% of online investors were selling securities on that day. This indication of increased online selling was found, as the trend in the market index was stable. Figure 1 contains a comparison of the AOII with the tracking stock for the NASDAQ-100 Index. This timeframe provides a challenging test for the ability of online investors to outperform a market index, since it contains a period of rising prices for roughly the first two-thirds and a gradual decline over the last one-third of the study. Based on a visual inspection of the two series, there seems to be a weak relationship between the two variables. The correlation coefficient between the two series is -0.2949 (p-value = 0.0198), which indicates a statistically significant negative relationship between the two variables. However, since the AOII is an indicator of online buying sentiment instead of a price index, a different type of analysis is required to measure the portfolio returns of the decisions of online investors.

METHODOLOGY

In addition to the AOII, the top ten equities that were bought and sold by Ameritrade's customers are available on the daily updates. The ratio of the summation of the top ten daily buys to the combined totals of the top ten daily buys and the top ten daily sells is a very close approximation of the reported AOII. From Appendix A, the sum of the online buyers for the first day of the study was 32.16%, while the online sellers equaled 52.93%, giving a total of the buys and sells of 85.09%. Dividing the buyers by the above total yields an estimate of the AOII (37.80%), which is a very close estimate of the reported index value of 37.68%. The top ten daily buys and sells will be the primary focus of this study and the respective positions in the stocks will be combined to create a portfolio consisting of the daily net holding online position investors take in an individual security. The absolute minimum number of stocks in the portfolio is ten, which would occur when the top ten daily buys and top ten daily sells are the same ten securities. Conversely, the maximum number of twenty stocks in portfolio would exist when none of the top ten daily buys appear in the list of the top ten daily sells. On an a priori basis, either one of these conditions would be a very rare occurrence. The daily net holdings are used to create the relative proportions invested in a given stock on a particular day. If the daily net holding is positive, a long position is created in the stock, while a short sell is indicated when this value is negative. The daily net holding of Security i on day t is defined as:

Daily Net [Holding.sub.i, t] = % Buyers of [Security.sub.i, t] - % Sellers of [Security.sub.i, t] (1)

Next, the daily returns for each of the stocks included in the portfolio were found by:

Daily [Return.sub.i, t] - ([Price.sub.i, t] - [Price.sub.i, t-1])/[Price.sub.i, t-1] (2)

The daily net holding is multiplied by the daily return to find the daily gain (loss) resulting from taking a position in each security. The gains and losses are summed up across all of the stocks to create the "AOII Net" portfolio return.

"AOII Net" Portfolio [Return.sub.t] = (Daily Net [Holding.sub.1, t] X Daily [Return.sub.1, t]) + (Daily Net [Holding.sub.2, t] X Daily [Return.sub.2, t]) + ... + (Daily Net [Holding.sub.i, t] X Daily [Return.sub.i, t]) (3)

The daily portfolio return will be compared to the daily return of the NASDAQ-100 Index. The return on this market portfolio was proxied by the tradable index mutual fund known as the Cube (Ticker symbol: QQQ). This security tracks the performance of the technology-heavy NASDAQ-100 Index (Lucchetti and Brown, 2000). This index was chosen as a proxy for the market portfolio, since most of the year-to-date online buys and sells by Ameritrade online investors have been concentrated in the high tech sector.

The return on the market index is defined as:

Market Index [Return.sub.t] = ([QQQ.sub.t] - [QQQ.sub.t-1])/[QQQ.sub.t-1] (4)

The ability of this active investing style of online investors to outperform the passive approach of buying the index fund will be measured by the calculation of the "AOII Net" advantage. This return differential is calculated as:

"AOII Net" [Advantage.sub.t] = "AOII Net" Portfolio Daily [Return.sub.t] - Market Index [Return.sub.t] (5)

If this advantage is positive, investors following this style of buying individual securities would have outperformed the market index. Otherwise, an investor would have benefited by investing in the index fund. The composition of the stocks and the corresponding returns of the online investor's portfolio and the "AOII Net" advantage will be readjusted on a daily basis until the end of the study (April 28, 2000). The sample size is 60 daily observations.

The "AOII Net" Portfolio

Table 1 presents the raw frequencies of the top ten daily buys and sells for the sixty trading days. An overall total of 126 individual stocks were bought and/or sold by online investors for the study period. However, the majority of online investors buys and sells were concentrated in the same eight stocks. The "most popular" stock, as well as the "least popular" stock was JDS Uniphase Corporation. This security was purchased fifty-seven of the sixty days (95% of the trading days) and sold on fifty days (83.33%). In addition, Cisco Systems, bought on 75% and sold on 77% of the study dates, and Qualcomm, purchased 63% and sold 75% of the study period, occupied the second and third positions on the daily buy and sell rankings. Online investor's buys and sells were heavily concentrated in the stocks of the high technology sector over the period of the study.

The calculation of the daily net holding for a given stock uses the daily buys for a stock minus the daily sells. Forty-one stocks were found only on the daily buy list (32.54% of the 126 total stocks), while 18.25% of stocks were found only on the daily sell list (23 stocks). The remaining 62 stocks (49.21%) were classified as net positions, defined as a minimum of one day with a mention on the daily buy, as well as the daily sell list.

Table 2 contains the top ten daily maximum and minimum net buys and sells. Cisco Systems had the largest positive net position of 37.45%, when 38.61% of Ameritrade's online investors were buying the stock and only 1.16% were selling the security on April 12, 2000. The largest single negative net position of -32.58% was on February 23, when 33.02% of online traders were selling Qualcomm versus 0.44% for the buyers of the security. The online investors were "approximately" right in these two specific instances. The daily return for the long position in Cisco was -5.96%, while the short position in Qualcomm had a stock price change of -4.89%. In comparison with the raw frequencies of daily buys and sells concentrated in eight companies, the largest net purchase and net selling decisions by online investors were focused on the four stocks of Cisco Systems, Qualcomm, America Online and MCI Worldcom.

PORTFOLIO RESULTS

The results for the first day of the study, February 1, 2000, are presented in Table 3. There were five net positions from a maximum of +8.53% to a minimum of -18.75%, five buys ranging from +1.56% to +0.37% and a range of -0.67% to -7.74% for the five sells in the fifteen stock portfolio. Since the top ten daily buys and sells are not posted until after the market's close on a given day, the closing price for the stock for the first day was used as the purchase price. This action assumes the opening price on the following trading day is approximately equal to the closing price on the previous day. The closing price on the second day was used as the selling price. The return was defined as the percentage change in the price of the stock for the one day holding period, using the strategy of buying (selling) at the opening price and selling (buying) at the closing price. The daily net holdings for the portfolio for the second day of the study were recreated after the information was available on the Internet. The daily net holdings for the first day ranged from a maximum of +8.53% in America Online to a short position of -18.75% in Qualcomm. The total of the daily net holdings is -20.77% (Column 3), which corresponds to the difference of the top ten buys of 32.16% minus the top ten sells of 52.93% from Appendix A. This results indicate a overall short position in the portfolio, where nine of the fifteen stocks (60%) were sold short and the remaining six stocks (40%) were held in a long position. For the first day's returns, America Online provided the largest daily return of +10.18%, versus a -9.20% change in the price for Plug Power. The largest positive component of the portfolio's first day return was 0.8684% for America Online (a daily net holding of +8.53% times the daily return of 10.18%). Qualcomm generated a weighted average loss of -0.8268%, calculated from a daily net holding of -18.75% multiplied times a return of +4.41%. The daily net holdings, or weights, were multiplied by the daily returns and summed up to find the "AOII Net" return of +0.2929%. The first day's return for the buy-and-hold strategy of purchasing shares of the tracking security for the NASDAQ-100 Stock Index was +0.8483%. The "AOII Net" advantage, defined as the "AOII Net" return minus the index return, was a loss of -0.5556%. For the first day of the sample, the return on the online investor's net portfolio was lower than the return on the index portfolio.

The trading results for the sixty days of the study are presented in Figure 2. As one might expect in an informationally efficient market, there are periods when the "AOII Net" portfolio advantage is positive, and other periods when the advantage is negative. For the overall period, the advantage was positive for thirty days and negative for thirty days. However, there are significant positive and negative "streaks" in the portfolio. The best results were found near the end of the study period (April 7 to April 13), when five days of positive returns and an average daily return of +3.734% were generated by the online investors. However, these positive results were almost exactly offset by a four-day run of negative results. An average daily loss of -3.395% was generated by the portfolio for the period of March 20 (Monday) to March 23 (Thursday).

[FIGURE 2 OMITTED]

Table 4 contains the sixty daily results and the return-risk characteristics for the market index, the "AOII Net" portfolio return and the "AOII Net" Advantage before transactions costs. Before trading costs, the average geometric return for online investors was -0.2053%, which is -0.2988% lower than the geometric return, due to compounding, on the market index of -0.0154%. This loss would be further accelerated when trading costs are considered. The primary benefit of the strategy exists in a slight reduction of risk, as shown by a decrease in the standard deviation of returns of 0.31%. The variability of the "AOII Net" Advantage was 3.36% versus the standard deviation of the market index returns of 3.67%. For the study period, there is no indication that online investors can outperform, in fact, they would have gained by buying-and-holding shares in the market index. There is no evidence of the emergence of contradiction to market efficiency, or the existence of an average return anomaly, based on the investment decisions of online traders.

In addition to poor performance relative to the market performance, an important additional cost to implementing this active investing style is the trading costs of rebalancing the portfolio on a daily basis. The amount invested in an existing individual stock in the portfolio, as well as additions and deletions to the composition of portfolio changed daily and would contribute to high brokerage fees. Figure 3 contains the number of individual stocks that make up the "AOII Net" portfolio. The absolute minimum number of stocks is ten, while the maximum is twenty. These two extremes are approached on the low side with eleven stocks in the portfolio on two days and on the high side, with three days with eighteen stocks in the mix. The daily turnover is quite high, and although while the costs of online trading is relatively low, the explicit cost of implementing this would be relatively expensive.

There were four consecutive days when a constant number of stocks (fifteen) were held in the portfolio and seven other instances when a constant number of securities were held for two successive days. However, even in these eight instances, trading costs would be incurred since the daily net holding position in a security was never constant. The benefits and costs of an investment strategy using a high turnover of stocks were measured in a recent article by Odean and Barber (2000). Using brokerage account data from more than 60,000 households for the period of 1991 to 1996, the two researchers found the most active traders, defined as individuals with a 258% turnover rates, exhibited the lowest annual returns. The differences in annual returns were quite significant, the least active traders generated a 7.1% advantage over the most active group, while the average household enjoyed a 5% advantage over the very active traders. The results from the present study support these findings, specifically, a lower return on the actively managed "AOII Net" portfolio than that available by passively investing in a market index.

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

The compound returns of a hypothetical $100 investment in each of the two strategies over the study period are shown in Figure 4. For the buy-and-hold, or passive, strategy of investing in the market index, an investor would have suffered a loss of $0.92, a compound growth rate of -0.02%. For the active approach of the online investor, the ending amount of the "AOII Net" portfolio would have been $88.40. This ending amount represents a compound growth rate of -0.21% for the study period. In other words, online investors would have faced a net loss of $10.68 ($88.40-$99.08) on their initial investment of $100. This negative return would be even more when the substantial brokerage fees to implement this very active portfolio management technique are considered.

SUMMARY

The explosion of financial information on the Internet, combined with the acceptance of many investors to conduct online transactions, has played a major role in transforming the world of personal investing today. Online brokerage houses use extensive television and Internet advertising to increase the volume of trading at their Websites. Many of these ads encourage investors with the abundance of information available, the speed of execution of the trade and the low transaction cost of buying and/or selling. These external factors, combined with an individual's human nature, have created many cognitive biases. Researchers in human behavior have documented three possible biases that may affect participants in online trading: self-attribution overconfidence, the illusion of knowledge and the illusion of control (Barber and Odean, 1999).

Given some degree of success in any task, it is only human nature to believe most, if not all, of the credit is due to the individual's ability instead of chance or luck. If a mistake occurs in trading, it most likely is someone else's fault or just bad luck. For the online investors in this study, this bias may have existed through the first twenty days of the sample. During this early part of the study period, the portfolio of online investor's net positions grew by approximately 5%, only to give up the gain and decrease to the ending position of losing about 12% from the initial investment. However, the market index increased by more than 25% for the first two-thirds of the study period, only to end with a loss of 1% from the initial stake. Online investors would have increased their net return by 11% by buying and holding the market index, as compared to actively managing their portfolio. The active style of portfolio management has other costs, such as brokerage fees, the bid-ask spread and the opportunity costs of the time committed to analyzing, evaluating and monitoring the portfolio. Assuming an average trade size of $5,000, an $8 per trade brokerage fee and a bid-ask spread of 1%, the cost of selling one stock and buying another to adjust the portfolio is $66 ($8 + $8 + $50). This cost would be multiplied by fifteen stocks, the average number of stocks in the online investor's portfolio, to calculate an expected daily transaction fee of $990. This fee would have to be offset by an increase in the replacement stock's price to justify the daily cost of adjusting the portfolio.

The illusion of knowledge is directly related to the amount of financial and stock information available to online investors. Online investors have access to Websites devoted to fundamental and technical analysis, historic and real-time quotes and stock recommendation discussion rooms, to name just a few of the many resources on the Internet. The implication is clear, the more data an investor has, the more capable she or he is to make an informed investment decision and the probability will also increase that a successful trade will be made. However, the assumption that what happened in the past will be repeated in the future may not hold. For example, in a coin flipping experiment, the probability is evenly distributed between the two outcomes. An investor could research where the coin was manufactured, the exact composition of metals, the manufacturing date, the weight of the coin and computer simulations of coin flipping exercises, to aid in the prediction of the next flip of the coin. However, even with all of this information readily available, the chance of predicting the next coin flip has not increased. Some predictions will be right, while others will be wrong, regardless of the amount of information accessed.

The final bias that may be evident in the actions of online traders is the illusion of control. Lopes (1987) states individuals have a desire to master their surroundings and find it discomforting to feel a complete lack of control. Perhaps, many online investors feel they have a great amount of control over their financial futures and are more likely to be frequent traders. An example of the breadth of online trading for the study period exists in the 126 different securities that were either bought and/or sold during the sixty trading days examined.

The results of this study support two well-documented facts in the field of finance. First, it is very difficult to outperform the portfolio returns of a market index. The portfolio of the net positions of Ameritrade's online investors generated a loss of -0.21%, compared to the return on the market index of -0.02%, yielding a net loss before transactions costs of -0.30%. For the study period, an investor would have benefited by purchasing and holding the market portfolio. Secondly, the daily trading fees for implementing this active style of investing would lower the above loss. Although the cost of executing an online trade has decreased in many cases to less than $10, active investors would still face the more substantial additional cost of the bid-ask spread. Suggestions for future research in this area of online investing are to create a "buyers" portfolio, made up of the top ten buys of online investors, as well as the complimentary "sellers" ten stock portfolio. The returns of either of these additional portfolios could be evaluated against a market index to continue the search for an average return anomaly. Online investing has generated an increase in trading activities for many individual investors, but for the buyers and sellers at the Ameritrade website the costs of actively managing a portfolio exceed the benefits of this strategy. In other words, in agreement with the consensus of previous literature, no exception to financial market efficiency was found and the ability of this group of investors to outperform a market index was found was not supported.

REFERENCES

Ameritrade Press Release 1999. Ameritrade launches online investor index: First daily measurement of behavior of online investors. (December 1).

Barber, B.M. & Loeffler, D. 1993. The "dartboard" column: Second-hand information and price pressure. Journal of Financial and Quantitative Analysis (28), 273-284.

Barber, B.M. & Odean, T. 1999. Online investors: Do the slow die first? University of California-Davis Working Paper, (September).

Carey, T. W. 2000. Better, not just bigger. Barron's (March 13).

Corrado, C. J. & Jordan, B. D. 2000. Fundamentals of Investing: Valuation and Management. The McGraw-Hill Companies, Inc., New York.

Desai, H. & Hain, P. C. 1995. An analysis of the recommendations of the "superstar" money managers at Barron's annual roundtable. Journal of Finance (50), 1257-1273.

Fama, E.F. 1998. Market efficiency, long-term returns and behavioral finance. Journal of Finance (49), 283-306.

Griffin, P.A., J.J. Jones & M. Zimijewski 1995. How useful are wall Street week recommendations? Journal of Financial Statement Analysis (1), 33-52.

Hirschey, M., V. J. Richardson & S. Scholz 2000. Stock-price effects of internet buy-sell recommendations: The motley fool case. The Financial Review (35), 147-174.

Lopes, L. 1987 Between hope and fear: The psychology of risk. Advances in Experimental Social Psychology (20), 255-295.

Lucchetti, A. & Brown, K. 2000. Amex is back, thanks to a variety of tradable index mutual funds. The Wall Street Journal, (February 22).

Mathur, I. & Waheed, A. 1995. Stock price reactions to securities recommended in Business Week's inside wall street. Financial Review (30), 583-604.

McQueen, G. & Thorley, S. 1999. Mining fool's gold. Financial Analysts Journal (55), 61-72.

Odean, T. & Barber, B.M. 2000. Trading is hazardous to your wealth. Journal of Finance (55), 773-806.

Thomas Willey, Grand Valley State University

Table 1: Raw Frequencies of Top Ten Daily Buys and Sells
(n = 60 trading days)

Company Days Bought Percent

JDS Uniphase Corp. 57 95.00%
Cisco Systems Inc. 45 75.00%
Qualcomm Inc. 38 63.33%
Microsoft Corp. 35 58.33%
America Online Inc. 33 55.00%
Oracle Corp. 29 48.33%
Ameritrade Holding Corp. 27 45.00%
CM GI Inc. 18 30.00%
Broadvision Inc. 17 28.33%
MCI Worldcom Inc. 16 26.67%

Company Days Sold Percent

JDS Uniphase Corp. 50 83.33%
Cisco Systems Inc. 46 76.67%
Qualcomm Inc. 45 75.00%
America Online Inc. 45 75.00%
Oracle Corp. 37 61.67%
Microsoft Corp. 25 41.67%
Dell Computer Corp. 23 38.33%
Yahoo! Inc. 21 35.00%
Ameritrade Holding Corp. 21 35.00%
CMGI Inc. 21 35.00%

Table 2: Top Ten Daily Maximum and Minimum Net Buys and Sells
(n = 60 trading days)

Company Percent

Cisco Systems Inc. 37.45%
Palm Inc. 35.51%
Microsoft Corp. 28.21%
Yahoo! Inc. 20.18%
America Online Inc. 16.36%
Proctor and Gamble Co. 14.61%
Microstrategy Inc. 12.68%
MCI Worldcom Inc. 11.91%
Qualcomm Inc. 10.38%
Exodus Communications 9.62%
Qualcomm Inc. -32.58%
America Online Inc. -29.01%
JDS Uniphase Corp. -18.71%
Lucent Tech Inc. -18.19%
Oracle Corp. -17.79%
Amazon.com Inc. -15.79%
Cisco Systems Inc. -14.81%
Ameritrade Holding Corp. -11.42%
Dell Computer Corp. -11.07%
MCI Worldcom Inc. -10.68%

Table 3: Ameritrade Online Investors Index (AOII) "Net" Portfolio
for February 1, 2000

In this example, the net portfolio would be created on February 1,
2000 and rebalanced on the following trading day. The daily return
is defined as the daily percentage change in the price of the security
or the index. The AOII net advantage for the first day of the study
is -0.5556 %. For the first day of the sample, the online investor's
net portfolio did not outperform the market index.

Company Minus Daily Net Holding %
 Sells %

Qualcomm Inc. 15.46-34.21 -18.75
America Online Inc. 9.45-0.92 8.53
JDS Uniphase Corp. 2.24-3.51 -1.27
E*Trade Group Inc. 1.56-0.00 1.56
Nokia Corp. 1.25-0.00 1.25
Tyco International Ltd. 0.52-0.00 0.52
Oracle Corp. 0.51-1.83 -1.32
Plug Power Inc. 0.42-0.82 -0.40
Yahoo! Inc. 0.38-0.00 0.38
Sykes Enterprises Inc. 0.37-0.00 0.37
Cisco Systems Inc. 0.00-7.74 -7.74
Amazon.com Inc. 0.00-1.61 -1.61
CMGI Inc. 0.00-0.92 -0.92
Exodus Communications 0.00-0.70 -0.70
Sun Microsystems Inc. 0.00-0.67 -0.67
AOII Net Return
Less: Index Return
Equals: "AOII Net" Advantage

Company Daily Return % Holding *
 Daily Return %

Qualcomm Inc. 4.41 -0.8268
America Online Inc. 10.18 0.8684
JDS Uniphase Corp. 2.35 -0.0298
E*Trade Group Inc. 1.57 0.0245
Nokia Corp. 0.81 0.0102
Tyco International Ltd. -2.85 -0.0148
Oracle Corp. 0.58 -0.0076
Plug Power Inc. -9.20 0.0368
Yahoo! Inc. 3.35 0.0127
Sykes Enterprises Inc. 2.36 0.0087
Cisco Systems Inc. -3.34 0.2587
Amazon.com Inc. 2.97 -0.0477
CMGI Inc. 3.52 -0.0324
Exodus Communications -5.15 0.0361
Sun Microsystems Inc. 0.62 -0.0041
AOII Net Return 0.2929
Less: Index Return -0.8483
Equals: "AOII Net" Advantage -0.5556

Table 4--Returns for the NASDAQ-100 Index and "AOII Net" Portfolios,
02/01/2000 to 04/26/2000

Date QQQ Return "AOII Net" "AOII Net"
 Portfolio Advantage
 Return Before
 Transaction
 Costs

01-Feb-00 0.85% 0.29% -0.56%
02-Feb-00 3.90% 0.36% -3.55%
03-Feb-00 0.68% 1.66% 0.98%
04-Feb-00 1.38% -2.13% -3.51%
07-Feb-00 4.19% 0.64% -3.55%
08-Feb-00 -3.53% 0.85% 4.38%
09-Feb-00 3.28% 0.79% -2.50%
10-Feb-00 -2.32% 0.65% 2.98%
11-Feb-00 -0.13% 0.36% 0.49%
14-Feb-00 0.72% -1.15% -1.87%
15-Feb-00 -0.47% -0.22% 0.25%
16-Feb-00 2.13% -0.14% -2.26%
17-Feb-00 -3.61% -0.17% 3.44%
18-Feb-00 0.32% -0.02% -0.34%
22-Feb-00 5.57% 2.68% -2.89%
23-Feb-00 2.16% 1.03% -1.13%
24-Feb-00 -2.29% -0.47% 1.82%
25-Feb-00 -0.06% 0.83% 0.89%
28-Feb-00 2.64% 0.16% -2.48%
29-Feb-00 0.82% -1.24% -2.06%
01-Mar-00 -2.32% -1.29% 1.02%
02-Mar-00 5.64% -5.27% -10.90%
03-Mar-00 0.79% -2.40% -3.18%
06-Mar-00 -1.51% 0.67% 2.18%
07-Mar-00 0.91% -0.59% -1.50%
08-Mar-00 3.37% 0.43% -2.94%
09-Mar-00 -0.43% 1.43% 1.87%
10-Mar-00 -2.62% 0.68% 3.30%
13-Mar-00 -3.70% -0.33% 3.37%
14-Mar-00 -4.19% -0.35% 3.84%
15-Mar-00 5.71% 1.76% -3.95%
16-Mar-00 1.90% -0.21% -2.11%
17-Mar-00 -2.82% 0.16% 2.98%
20-Mar-00 3.77% -1.53% -5.30%
21-Mar-00 2.01% -0.47% -2.48%
22-Mar-00 1.75% -0.75% -2.51%
23-Mar-00 1.35% -1.95% -3.29%
24-Mar-00 0.16% 0.24% 0.08%
27-Mar-00 -2.55% -0.34% 2.20%
28-Mar-00 -4.14% -2.15% 1.99%
29-Mar-00 -2.50% 0.02% 2.52%
30-Mar-00 2.10% 1.50% -0.60%
31-Mar-00 -5.99% -0.67% 5.32%
03-Apr-00 -0.97% -1.02% -0.05%
04-Apr-00 -1.78% -0.43% 1.35%
05-Apr-00 2.37% -0.38% -2.75%
06-Apr-00 4.02% -1.03% -5.05%
07-Apr-00 -6.10% 1.27% 7.37%
10-Apr-00 -2.37% -0.82% 1.55%
11-Apr-00 -5.88% -3.12% 2.76%
12-Apr-00 -4.35% -2.67% 1.68%
13-Apr-00 -8.66% -3.35% 5.31%
14-Apr-00 11.51% 5.61% -5.90%
17-Apr-00 2.65% 0.05% -2.60%
18-Apr-00 -3.06% 0.15% 3.21%
19-Apr-00 -2.17% 0.05% 2.22%
20-Apr-00 -2.87% -2.11% 0.75%
24-Apr-00 6.75% 3.37% -3.38%
25-Apr-00 -3.42% 0.36% 3.78%
26-Apr-00 4.43% -0.82% -5.25%
Arithmetic Average 0.05% -0.19% -0.24%
Geometric Average -0.02% -0.21% -0.30%
Standard Deviation 3.67% 1.63% 3.36%