BKM Chapter 7 Flashcards
Reasons investors distinguish between asset allocation & security selection (3)
(BKM - 7)
- demand for investment mgmt has increased over time
- financial markets have become too sophisticated for amateur investors
- economies of scale in investment analysis
Main sources of uncertainty (2) - description & examples
BKM - 7
- systematic - risks common to all stocks, including: inflation, interest, and foreign exchange rates (aka non-diversifiable)
- non-systematic - risks from firm-specific influences such as firm sucess in R&D and personnel changes
Insurance principle
BKM - 7
risk reduction from spreading exposures across many independent risk sources (eliminates firm-specific risk)
Efficient portfolios
BKM - 7
risky portfolios with the maximum expected return for a given level of risk
Expected return for the complete portfolio
BKM - 7
weighted sum of expected returns
Variance of the risky portfolio
BKM - 7
= double sum of w(i) * w(j) * Cov(r(i), r(j))
Diversification benefit
BKM - 7
as long as correlation (rho) <> 0, then the standard deviation of the total portfolio is < the weighted average standard deviations of the individual assets
there is no diversification benefit if rho = 1
Cov(r(i), r(j))
BKM - 7
Cov(r(i), r(j)) = rho(i,j) * sigma(i) * sigma(j)
Hedge asset
BKM - 7
asset that has a negative correlation with other portfolio assets
Diversification benefit with perfect correlation (rho = 1)
BKM - 7
no diversification benefit, all risk is systematic risk
Perfect hedge
BKM - 7
perfect negative correlation (rho = -1)
Optimal weights for 2 risky assets in the optimal risky portfolio
(BKM - 7)
w(D) = {E[R(D)] * sigma^2(E) - E[R(E)] * Cov (R(D), R(E))} / {[E[R(D)] * sigma^2(E) + E[R(E)] * sigma^2(D) - (E[R(D)] + E[R(E))] * Cov(R(D), R(E))}
w(E) = 1 - w(D)
D = debt (bonds) E = equity (stocks) R = excess return
apply w(D) and w(E) to y* (% in risky portfolio) to get weights for individual assets relative to the total portfolio
Separation property (2)
BKM - 7
portfolio choice comes down to:
- determining the optimal risky portfolio (objective - same portfolio for every investor)
- determining the proper capital allocation (amount allocated to risk-free vs. risky portfolio) based on investor’s risk aversion (subjective - different portfolio for every investor)
Risk pooling (aka insurance principle)
BKM - 7
adding uncorrelated risky projects to a portfolio (spreading exposures across uncorrelated projects)
Risk-sharing
BKM - 7
allowing other investors to share the risk for a portfolio
Risk-pooling, risk-sharing, and risk reduction
BKM - 7
risk-pooling by itself does not reduce risk, but does improve the Sharpe ratio
combining risk-pooling and risk-sharing reduces risk (b/c the total size of the portfolio is unchanged vs. adding risks and increasing the investment)
Markowitz portfolio optimization model (3 + 2 sub-steps)
BKM - 7
- determine available risk-return opportunities
a. create input list of expected returns, variances, and covariance estimates
b. optimization model finds the efficient frontier - find the CAL that maximizes the Sharpe ratio
- identify the optimal mix b/w the optimal risky portfolio & risk-free asset
Optimization considerations (3)
BKM - 7
- whether shorting is permitted
- minimum dividend yield
- exclude ethically or politically controversial investments
Minimum variance risky portfolio weights
BKM - 7
w(D) = [sigma^2(E) - Cov(r(D), r(E))] / [sigma^2(E) + sigma^2(D) - 2 * Cov(r(D), r(E))]
w(E) = 1 - w(D)
Goal of the optimal risky portfolio vs. minimum variance portfolio
(BKM - 7)
optimal risky portfolio: maximize the Sharpe ratio
minimum variance risky portfolio: minimize overall variance (may produce a lower, less attractive Sharpe ratio)
Time diversification myth
BKM - 7
risk is not reduced by spreading investments across time
instead, it is preferable to invest in an efficient portfolio over many periods, reducing the amount in the risky asset in each period
Total variance for an equally weighted portfolio where all securities share the same sigma & correlation
(BKM - 7)
sigma^2(p) = (1 / n) * sigma^2 + ((n - 1) / n) * rho * sigma^2
n = # of securities
Systematic risk for an equally weighted portfolio of n securities
(BKM - 7)
systematic risk = rho * sigma^2
*does not depend on n
Firm-specific risk for an equally weighted portfolio of n securities
(BKM - 7)
firm-specific risk = total risk - systematic risk
“risk” = variance
Correlation and risk as the number of securities increases
BKM - 7
when rho > 0, as n increases, portfolio risk approaches systematic risk (firm-specific risk is diversified away)
when rho = 0, as n increases, portfolio risk approaches 0 (there is no systematic risk & all firm-specific risk is diversified away)