Full service brokers are working with different revenue streams and ancillary businesses that greatly reduce their effectiveness as comparables. Waterhouse's relevance becomes suspect when you consider that they are no longer public. Their data is old - the beta of a discount brokerage in the telephone era is not especially useful to a decision about a discount brokerage entering the Internet era. The cost of capital for eTrade is 6.4%. As with Ameritrade they have no long-term debt and do not pay dividends. Ameritrade's cost of capital is going to be around 8%. Their borrowing cost is 8-9%, and we should take the lower figure to reflect that they have some equity, and that equity has thus far been free due to the lack of dividends. Joe Ricketts should view the cost of capital I have calculated as a rough guideline. The investment he is considering is unique in the history of Ameritrade. The data used in the calculation is based on data that does not necessarily reflect the company Ameritrade will become after...

That fact will make the historical data I have used to be less valuable. Moreover, the beta was an estimate based on another firm because of the short time in which Ameritrade has been public. Ameritrade's real beta will be different. This investment also changes the beta by making a material alteration in Ameritrade's business model. In essence, very little useful information has been provided. The undertaking is so different that it makes finding a useful comparable near impossible. The further we get from evaluating the investment on its merits, and start talking about companies with entirely different revenue streams and technologies, the less likely getting a good number from that data becomes. Unfortunately, this puts Joe in a situation where he isn't likely to get much more than a rough guideline regarding the risk of this investment.
Works Cited

Kester, W. Carl, Tofano, Peter, and Ruback, Richard S. (2005) Case Problems in Finance. 12th ed. McGraw-Hill, New York.