Ameritrade – Cost of Capital

In mid-1997, Joe Ricketts, Chairman and CEO of Ameritrade Holding Corporation, wanted to improve his company’s competitive position in deep-discount brokerage1 by taking advantage of emerging economies of scale. The success of the strategy required Ameritrade grow its customer base. The growth would require substantial investments in technology to improve service and capacity, and in advertising, to increase customer awareness. The strategy would require large expenditures relative to Ameritrade’s existing capital. In order to evaluate whether the strategy would generate sufficient future cash flows to merit the investment, Ricketts needed an estimate of the project’s risk.

DUKE UNIVERSITY Fuqua School of Business FINANCE 251F/351 Individual Assignment #1 Cost of Capital at Ameritrade Prof. Simon Gervais Spring 2010 – Term 1 In this case, you have to use data from comparables to estimate the cost of capital at Ameritrade. The process involves a few stages that this handout will guide you through. First, we need to determine which set of firms to use as comparable firms. You should try two different sets. The first set will include three discount brokerage firms: Charles Schwab Corp, Quick & Reilly Group, and Waterhouse Investor Services.1 The second set will include six investment services firms: A G Edwards, Bear Sterns, Merrill Lynch, Morgan Stanley Dean Witter, Paine Webber, and Raymond James Financial. Stock

It is based on the key assumption of rational, mean-variance optimising investors with the identical use and access to information, who can borrow or lend at a common risk-free rate as well as invest in public traded assets in a single period without taxes and transaction costs. Hence this, each investor held the Sharpe-ratio maximising market portfolio and optimises its utility by leverage or de-leverage it, based on their rate of risk aversion. Holding a well-diversified portfolio, the company specific risk – defined as the beta in the CAPM – is the only factor affecting the expected returns. Subsequently, the Security Market Line (SML) can be obtained as the expected return-beta relationship.

Ameritrade went public only from March 1997, and this project investment decision is being presumed around August 1997. In the initial phase of a publicly-traded company, the stock return can be extremely volatile and difficult to estimate. This also presents a challenge that we cannot accurately calculate an equity beta (βE or βL) due to lack of sufficient historical data for variances and the expected return based on statistical tools. In such a scenario, we seek to use the data from similar companies within the same industry. It is prudent to use the data of companies that have similar capital structure, size, investments and business operations.

Subject Abstract 1. Summary 2. Total Return Model 3. After-Tax Discounting 4. Derivation of Risk-Adjusted Discount Rate and Liability Beta Figure l : Baseline Risk / Return Line vs Leverage 5. Liability Beta Figure 2: Equity vs Liability Beta Figure 3: Equity Beta vs Risk-Adjusted Discount Rate (After-Tax) 6. Underwriting Profit Margin Figure 4: Underwriting Profit Margin vs Loss Payout Figure 5: Underwriting Profit Margin vs Investment Yield Figure 6: Underwriting Profit Margin vs Market Risk Premium Figure

Now, using the Value Weighted NYSE index as the proxy for the market return, compute equity betas for each company. Use the past 5 years of data to compute the beta. After computing equity beta, using the Debt-value ratio in Exhibit 4, unlever the equity beta, (i.e., compute the asset beta from equity beta).

If Wild Widgets, Inc., (WWI) were an all-equity firm, it would have a beta of 0.9. WWI has a target debt-to-equity ratio of 0.50. The expected return on the market portfolio is 16%, and Treasury bills currently yield 8% per annum. WWI one-year, $1,000 par value bonds carry a 7% annual coupon and are currently selling for $972.73. The yield on WWI’s longer term debt is equal to the yield on its one-year bonds. The corporate tax rate is 34%.

In mid-1997, Joe Ricketts, Chairman and CEO of Ameritrade Holding Corporation, wanted to improve his company’s competitive position in deep-discount brokerage1 by taking advantage of emerging economies of scale. The success of the strategy required Ameritrade grow its customer base. The growth would require substantial investments in technology to improve service and capacity, and in advertising, to increase customer awareness. The strategy would require large expenditures relative to Ameritrade’s existing capital. In order to evaluate whether the strategy would generate sufficient future cash flows to merit the investment, Ricketts needed an

F&F (1992) has investigated the association between CS stock returns and the five common suggested factors (the beta, the size of the firm, the BTMR, the EPR, and the DER) on the U.S. Stock Market. According to F&F, the impacts of the DER and the EPR can be absorbed by the BTMR and the size of firm factors. In preparation for examining the strength of the size of the firm and the BTMR impacts, they separated the examination period into three sub-periods. The authors discovered that the market beta was affirmative only during a single period. However, it could not be statistically substantial. The impact of the firm's size was insignificant, from 1977 to 1990, but it was substantially associated to stock returns.

Since Asset Evaluation discount rate is determined by return on capital employed (or ROI) assessment project, Therefore, the capital asset pricing model has a wide range of application in assessment. The central role of CAPM method is to analyze the portfolio and securities value, then find the cheap securities. It provides a standard for evaluate the value of securities. Expected rate of return of each security shall be equal to the risk-free rate plus a risk premium measured by the coefficient β: Ri= Rf+ βi( Rm- Rf). When the expected of return of the market portfolio is estimated and the β of the securities is estimated, then the expected of return under the market equilibrium can be calculated. In addition, there is an expected value in market arising from the future income( dividends and terminal value). Ri= ( dividends + terminal value)/ initial value - 1. In an equilibrium state, these two expected rate of return should be the equal, and the initial value should be set at (dividends + terminal value ) /(Ri + 1)(Da & Jagannathan, 2012). Compare the current actual market price with the equilibrium of the initial price. If they are not equal,it is indicated that the market price is set by mistake. The mistake price should have the return requirement, it can obtain excess returns by using this(Da & Jagannathan, 2012). Specifically, when the actual price is lower than the equilibrium price, indicating that the stock is cheap

The advantage of CAPM is that it provide a clear and intuitively explicit forecast in regard to how to measure risk and the connection between risk and expected return(Fama & French , 2004).Accordance to the provisions of CAPM, Beta coefficient is used to measure an asset systemic risk, it is used to measure the

The evidence gained from examination done by Nimal and Fernando (2013) concerning Tokyo Stock Exchange (TSE) and the Colombo Stock Exchange (CSE) confirmed not only that beta is a useful tool in expanding deviations in market premium, but also that there is a relation between return and beta. However, the previous research done in the Japanese market by Yonezawa and Hin, (1992) did not confirm the validity of the CAPM. In their research, they checked monthly returns from January 1952 to December 1986 and concluded that limited diversification was the main cause of CAPM failure.

This study explores the (troubling) empirical evidence bearing on capital asset pricing theories. General formulas for the coefficient on beta and it standard error are derived, which show that the outcomes of cross-sectional tests have no causal relation to the pricing models. If a test refutes a model, this could be because the model is misspecified or because poor proxies for true expected returns and betas are used. Simulation and calibration results suggest that realized returns are a much poorer proxy than estimated betas are. The noise in realized returns typically inflates the estimated standard error, with drastic effects on the statistical power. Inferences based on ex ante returns are more powerful but suffer from a serious size problem. JEL: G12, C31, C52.

Capital Asset Pricing Model is the foundation stone of modern finance theory. It reveals the basic operation rules of the capital market and it is important in market practice and theory research(). By use of this model, the relation between risk and expected return is accurately predicted, and provides a method to estimate the yield of potential investment projects and help us to predict the expected rate of return of market in future. Although the Capital Asset Pricing Model is not entirely consistent with the empirical test, however, because of the simplicity of the model, it is still widely used. A model may have highly realistic assumptions, but if it has no predictive power, it is largely worthless. Most researchers have attempted to test the CAPM to see if it works, looking at the relationship between observed beta values and average returns(). One phenomenon is that the actual slope of the Security Market Line is slightly less than the predicted slope of the CAPM. This essay is going to discuss potential explanations of this phenomenon.

In their robustness checks they firstly adjust the MPEG model for predictable forecast error and also use model generated earnings forecasts in place of analyst forecasts. Each paper has advantages and disadvantages in their estimation of cost of equity capital, Francis et al. (2004) uses two radically different models (one based on analysts’ forecasts, and one on PE ratios) which could lead to more robust results; but suffer from unenviable biases in their set. Conversely, Artiach and Clarkson (2014) have the luxury of choosing from an excellently ranked menu of cost of equity models thanks to Botosan (2011); however are limited by data in not being able to use the two most sophisticated, although do still have a more updated approach than Francis et al. (2004). The fundamental difference here is that Francis’ et al. (2004) primary results are driven by an analyst forecast model, whereas Artiach and Clarkson’s (2014) are driven by PE ratio model; as a result the results will be fundamentally different.