Reading 3.4

Question 1

Strategic asset allocation is

Answer 1

long-term target asset allocation based on investor objectives and long-term risk-return expectations

Question 2

Tactical asset allocation is

Answer 2

making short-term portfolio decisions to change a portfolio’s systematic risks to generate alpha. Most investment managers also make tactical decisions

Question 3

Modern portfolio theory (MPT), which was pioneered by the economist Harry Markowitz (Markowitz 1952), states that

Based on the principle of _____, MPT suggests that allocation choices in perfect markets are mean-variance efficient portfolios which means?

Answer 3

assets that are less than perfectly correlated can be combined to
maximize return for a given level of risk.

diversification

optimal portfolios that maximize expected return given a risk aversion level

Question 4

To create a mean-variance efficient portfolio (under the MPT theory), investors do the following:

Answer 4

combine the risk-free asset and the market portfolio

Question 5

Why is applying MPT to alternative investments is challenging?

What is the solution to dealing with the issue?

Answer 5

due to the lack of quality data

A “satisficing” approach is used (coined by the economist Herbert Simon) = searching through available alternatives until an acceptable (good enought) solution is identified

Question 6

probability-weighted expected return & probability-weighted standard deviation of returns may be expressed as

Answer 6

Question 7

Asset owners’ risk-return preferences may be stated in terms of their utility, which is measure of

Answer 7

satisfaction gained from investment wealth or return.

Question 8

Utility function U(*) describes what?

An investor’s utility may be expressed as a function of wealth W as

Answer 8

the relationship that converts an investment’s wealth or return into the investor’s level of utility

U(W)

Question 9

Expected utility is the ___

Answer 9

probability-weighted average utility over all possible outcomes.

Question 10

An investor’s expected utility with two potential outcomes to wealth, W1 and W2, may be expressed as

Answer 10

Question 11

Risk-averse investors (i.e., those who demand higher expected return to bear risk) have concave utility functions. What does it mean?

Risk-averse investors do not typically take risks on investments with ___

Answer 11

increasing at a decreasing rate

no expected payoffs

Question 12

investment’s expected utility may be expressed in terms of its expected return μ and its variance of returns o2 (st dev) as:

Answer 12

Question 13

assumed investor preferences are that

Answer 13

most investors dislike variance and kurtosis and like positive skewness

Question 14

Expected UTILITY FUNCTIONS WITH HIGHER MOMENTS (formula)

Answer 14

Question 15

Expected utility may be expressed using value at risk (VaR) as the risk measure instead of variance

Answer 15

Question 16

The degree of risk aversion, may be expressed as the ratio of the optimal portfolio’s expected excess return to its variance, formula:

Answer 16

Question 17

MEAN-VARIANCE OPTIMIZATION (MVO); The return Rp on a portfolio of N risky assets and one riskless asset may be expressed as:

Answer 17

Question 18

Efficient Frontier is a graph of

Portfolios on the efficient frontier represent portfolios with the ____ for a given ____

Answer 18

expected returns and standard deviations of risk-averse investors’ optimal portfolios

highest expected return

standard deviation (SD)

Question 19

In general, MVO does not generate ____

Answer 19

realistic weights

Question 20

hurdle rate is the ____

The hurdle rate for a new asset may be expressed as:

Answer 20

minimum expected return that an asset must earn to add value to an optimal portfolio (to be included in the portfolio)

Question 21

if an analyst overstates an asset’s mean return, then an optimization model’s
suggested weight for the asset is likely to be ___ than would be considered reasonable.

Alternatively, if the asset’s mean return is understated, other assets would be _____.

Answer 21

larger

largely omitted from the portfolio

Question 22

The mean-variance optimization (MVO) process may be described as follows:

Answer 22

1. The portfolio manager estimates the mean return and the variance and covariance of returns for all assets.
2. The optimizer produces an extreme portfolio allocation (which is unrealistic) = Extremely large allocations are made to seemingly attractive assets with high mean returns and low return volatilities, and little or no allocations are made to seemingly less desirable assets with low mean returns.
3. To generate more reasonable portfolio allocations, the portfolio manager subjectively adjusts the model by adding constraints or changing the inputs (e.g., mean return).

Question 23

Accuracy of estimated variance and covariance measures can be improved using _____ data. However, this data is not available for most ____.

Furthermore, assets that lack ____tend to be illiquid, with reported prices/returns typically based on ____

Answer 23

higher frequency

alternative assets

high-frequency prices

appraisals

Question 24

Anson (2016) empirically analyzes the effects of unsmoothing private equity and venture capital returns on asset allocation. He finds that unsmoothing results in:

Answer 24

1) considerably higher volatility estimates (twice those of smoothed returns)
2) considerably higher estimated return correlations of private equity and venture capital with equities, credit, and government bonds.
3) better estimates of higher moments (e.g., skew and excess kurtosis)

Question 25

Another issue with estimating covariance of returns occurs when optimization involves portfolios with large numbers of assets, optimization requires estimates of the covariance between ____, and this can be very large for large portfolios

Answer 25

each pair of assets in the portfolio

Question 25

Regarding asset allocation, Anson finds that optimal allocations based on unsmoothed returns were ___ lower for private equity and ___ lower for venture capital

Answer 25

20%

30%

Question 25

An issue with the MVO is that it considers only the first two moments of a return distribution (i.e., mean and variance) and ignores the higher-order moments (e.g., skewness and kurtosis).

How do you deal with it?

Answer 25

1. The optimization model is expanded to include skewness and perhaps kurtosis (issue with this approach is that skewness and kurtosis are difficult to predict accurately, and they are disproportionately influenced by a few extreme observations)
2. The MVO model includes skewness and kurtosis as constraints.
   * For instance, a portfolio’s skewness may be constrained to be positive and its excess kurtosis constrained to be less than three.
   * An issue with this approach is that it may not be possible to construct a portfolio with the desired skewness or kurtosis values.
3. Weights of investments with undesirable skewness or kurtosis values can be constrained. For instance, the maximum allocation to relative value hedge funds may be set at 10%.

Question 26

There are two types of illiquidity risk

Answer 26

1. Market liquidity risk - risk of being forced to sell a fairly illiquid asset in a market with few active market participants. In this and in particularly highly stressed environments, the asset will likely need to be radically reduced in price in order to be sold.
2. Funding liquidity risk - risk of borrowers or investors being unable to immediately pay what they owe. This may result in forced liquidation of assets; for instance, if investors in the futures market cannot meet margin calls and clearing houses close out the affected positions.

Question 27

What is a liquidity penalty function?

What does its incorporation allow for?

Answer 27

A mathematical adjustment made to the Mean-Variance Optimization (MVO) model to account for illiquidity in asset markets. It reflects investors’ preference for liquidity and the cost associated with holding illiquid assets.

By incorporating this function into the MVO model, investors can better assess the trade-offs between expected returns and liquidity constraints when allocating assets, especially when dealing with illiquid assets.

Question 28

How to reduce error estimation risk for small and large portfolios (in terms of quanity of investments)?

Answer 28

For small portfolios, subjective adjustments and consideration of assets with short return histories or low trading frequencies help mitigate estimation errors.

In large portfolios, resampling and shrinkage techniques control for estimation error risk and stabilize allocation decisions.

Question 29

two key approaches for resampling returns for portfolio optimization:

Answer 29

1. Hypothetical Returns Simulation: Estimates statistics from historical data, then simulates new return scenarios based on these estimates.
2. Hypothetical Samples of Actual Returns: Draws random samples with replacement from the historical data, creating new combinations of past performance.

Question 30

Explain the resampling returns technique (how it works and what is the goal)

Answer 30

Resampling returns is a technique that shuffles historical data to create multiple possible futures, helping build more robust investment portfolios.

The ultimate goal is to improve portfolio optimization

Question 31

What is the Shrinkage technique in reducing estimation errors

Answer 31

It is accomplished by reducing CONFIDENCE INTERVALS of statistical parameters, with the goal of improving forecasts of the statistical parameters (i.e., achieving better predictive capabilities) and thus mitigating extreme portfolio allocations and improving risk-adjusted performance.

Question 32

How does the The Black-Litterman (BL) approach tackle MVO’s risk of extreme allocations?

Answer 32

by letting managers inject their investment views about asset classes, leading to a final portfolio that blends historical data with expert judgment.

Question 33

Another way to circumvent the issues and sensitivities of MVO (including estimation errors) is to add additional constraints to the MVO model.

4 Common constraints used include the following:

Answer 33

1) Non-negative Weights (No Short Sales) - most common constraint, preventing negative portfolio weights (borrowing and selling assets). It helps avoid overly speculative strategies based on extreme asset views and potentially controls estimation errors in historical data.

2) Limits on Individual Asset Allocation: This sets a maximum investment amount for any single asset, ensuring diversification and reducing portfolio risk overdependence on any one investment.

3) Correlation Limits: Constraints can be placed on the estimated correlations between assets. This helps manage the model’s reliance on potentially inaccurate correlation estimates.

4) Tracking Error Limits: This limits the portfolio’s deviation from a benchmark or market index. This ensures the portfolio stays aligned with a target investment style or overall market performance.

Question 34

Overusing constraints for thhe MVO model can result in?

Answer 34

Restricting the optimization process and lead to allocations heavily influenced by the constraints themselves, not necessarily the best return-risk balance.