Investments, tenth edition

7

  THE INVESTMENT DECISION 

can be viewed 

as a top-down process: (i)  Capital allocation

between the risky portfolio and risk-free 

assets, (ii)  asset allocation  in the risky portfo-

lio across broad asset classes (e.g., U.S. stocks, 

international stocks, and long-term bonds), 

and (iii)  security selection  of individual assets 

within each asset class. 

 Capital allocation, as we saw in Chapter 6, 

determines the investor’s exposure to risk. 

The optimal capital allocation is determined 

by risk aversion as well as expectations for the 

risk–return trade-off of the optimal risky port-

folio. In principle, asset allocation and security 

selection are technically identical; both aim 

at identifying that optimal risky portfolio, 

namely, the combination of risky assets that 

provides the best risk–return trade-off. In 

practice, however, asset allocation and secu-

rity selection are typically separated into two 

steps, in which the broad outlines of the port-

folio are established first (asset allocation), 

while details concerning specific securities 

are filled in later (security selection). After we 

show how the optimal risky portfolio may be 

constructed, we will consider the costs and 

benefits of pursuing this two-step approach. 

 We first motivate the discussion by illustrat-

ing the potential gains from simple diversifi-

cation into many assets. We then proceed to 

examine the process of  efficient  diversification 

from the ground up, starting with an invest-

ment menu of only two risky assets, then 

adding the risk-free asset, and finally, incor-

porating the entire universe of available risky 

securities. We learn how diversification can 

reduce risk without affecting expected returns. 

This accomplished, we re-examine the hierar-

chy of capital allocation, asset allocation, and 

security selection. Finally, we offer insight into 

the power of diversification by drawing an 

analogy between it and the workings of the 

insurance industry. 

 The portfolios we discuss in this and the fol-

lowing chapters are of a short-term-horizon—

even if the overall investment horizon is long, 

portfolio composition can be rebalanced or 

updated almost continuously. For these short 

horizons, the assumption of normality is suf-

ficiently accurate to describe holding-period 

returns, and we will be concerned only with 

portfolio means and variances. 

 In Appendix A, we demonstrate how con-

struction of the optimal risky portfolio can 

easily be accomplished with Excel. Appendix B 

provides a review of portfolio statistics with 

emphasis on the intuition behind covariance 

and correlation measures. Even if you have 

had a good quantitative methods course, it 

may well be worth skimming.