Performance evlaution and portfolio analytics

Question 1

Describe the sharpe ratio and advantages and disadvantages

Answer 1

measures return compensation per unit of total risk, with risk, measured by std. dev. -> slope of CAL

Advantages
 Simple to use
 Allows comparison across asset classes
 Intuitive conceptual interpretation (reward per unit of total risk)
 Not affected by leverage

 Disadvantages
 Assumes that returns follow normal distribution
 Does not distinguish between good and bad volatility
 Does not distinguish between systematic and idiosyncratic risk
 Comparison needed

Question 2

Describe the sortino ratio and lists advantages and disadvantages

Answer 2

Measures return compensation per unit of bad riks, which is measured by downside deviation (stdev of all returns that are lower than RF)

Advantages
 Distinguishes between good and bad volatility
 Does not assume that returns follow normal distribution
 Allows comparison across asset classes
 Intuitive conceptual interpretation

 Disadvantages
 Noisier than Sharpe ratio when return distribution is symmetric
 Does not distinguish between systematic and idiosyncratic risk
 Comparison needed

Question 3

Describe the traynor ratio, and what are advantages and disadvantages

Answer 3

Return compensation per unit of systematic risk, measured by market beta

Advantages
 Distinguishes between systematic and idiosyncratic risk
 Intuitive conceptual interpretation

Disadvantages
 Assumes that beta is positive and constant
 Does not distinguish between good and bad beta
 Requires beta estimation -> sensitive to chosen benchmark
 Comparison needed

Question 4

Describe jensen’s alpha, and list advantages and disadvantages

Answer 4

Excess return of the portfolio above CAPM

 Alpha is the difference between actual return and predicted return
 Appropriate when portfolio is part of broader, fully diversified portfolio

 Advantages
 Simply to compute
 Statistical significance easy to test
 Level has intuitive interpretation
 Direct measure of return added by manager

 Disadvantages
 Noise in beta estimation leads to noise in alpha estimation
 Beta may vary over time due to timing or style switching
 Does not distinguish between god and bad beta
 Alpha depends on benchmark (factors) in models
 Alpha can be simply boosted by using

Question 5

What are problems for performance measurement?

Answer 5

-Survivorship bias
-Luck
-‘‘Beta in disguise’’
-Lack of data/data manipulation
-> smoothing of returns, non-linear strategies, deviating from proclaimed investment style (style drift)

Question 6

Why do HF managers increase risk?

Answer 6

Flow-performance relationship: outflows from bad performance are smaller than inflows from good performance -> may give poorly performing managers incentive to gamble

Question 7

Why do we want to measure performance?

Answer 7

-identifies talent
-Criteria for compensation (only for abnormal returns that cannot be achieved through passive investing)
-Incentivizes HF managers to decide well

Question 8

What is the information ratio, and its advantages and disadvtages?

Answer 8

Alpha divided by idiosyncratic risk

Pros:
accounts for idiosyncratic risk (deviation from benchmark)
Intuitive conceptual interpretation (active return/active risk)

Cons:
Same as for any measure that uses alpha as input, the same drawbacks as jensen’s alpha
It needs to be compared, it doesn’t say much by itself. IN practice IR of 0.5 is good, IR of >1 is excetional

Question 8

Describe style analysis, and the steps

Answer 8

We choose style analysis to evaluate funds comapred to passive similar styles. So to evaluate fund strategies, in some cases to replicate them, to build liquid alternatives and to manage risk exposure.

The process
1. regress fund returns on passive style portfolios/benchmarks
2.Regression coefficients constrainted to be nonnegative (MFs cannot short sell) and sum to 1 -> solver required
a. benchmark coefficients reveal implicit weight of each asset class
b. intercept measures average return due to security selection and timing
c r2 measures % of returrn variation due to style choice rather than selection -> style choice matters more than security selection specifics

Question 9

Describe style analysis issues

Answer 9

-Ambiguity on definitions: what is a value style?
-Style drift
-Depends on appropriate selection of benchmarks
-Not immune to manipulation (return smoothing)
-Statistical testing is difficult, bnechmarks are strongly correlateid
-Yields average style but does not capture rapid style change

Question 10

Style analysis pros

Answer 10

-Can handle any strategy for which passive indices exist
-Uses only return information (no holding needed)
-Powerful analysis which detects style drift
-Gives insights in how to replicate the funds returns

Question 11

If alpha is 2%, what is the expected return of a market neutral position?

Answer 11

Alpha + RF