Cfa Level Iii - Study Session 17

Front 1

Performance Attribution for Global
Total return for the International Portfolio
(SS17)

Back 1

Return in Local Currency
i. basic return in dollars
ii. basic return in euro
iii. basic return in yen

Return in Base Currency
i. basic return in dollars is also the return in base currency
ii. euro converted to dollars at start and finish, then basic return in dollars
iii. yen converted to dollars at start and finish, then basic return in dollars

Total return
i. Capital gain (portfolio local)
ii. Yield
iii. Currency component

Total return in base currency
= capital gain + yield + currency component

1. Most questions will have yield equal to zero by saying dividends aren't paid that period on indexes or stock
2. Capital gain (is in LOCAL currency)
3. Currency component does NOT always equal the change in the ER because there is a cross product from any capital gain/loss on the stock combined with the ER. Most of the time they are close though.

Front 2

Performance Attribution for Global
Total return for the International BENCHMARK
(as opposed to your portfolio)
(SS17)

Back 2

Total return
i. Capital gain (benchmark local)
ii. Currency component

Total return in base currency (benchmark)
= capital gain (benchmark local) + currency component

Benchmark weights
Rate of return in base benchmark dollars (assuming dollars are base currency)
Rate of return in converted benchmark euros to dollars
Rate of return in local benchmark euros
Rate of return in local benchmark yen

Front 3

Performance Attribution for Global
Total return for market and security selection

(SS17)

Back 3

Total return
i. Market return
ii. Security selection
iii. Yield component (most of the time zero)
iv. Currency component

Capital gain (local) = Market return + Security selection

1. The market return is the amount the manager would've made if they'd just invested their weights in a passive local market index
2. Security selection is the managers stock picking skill in beating return in that market against the passive local market index

2 Benchmark Return
-1 Market allocation contribution
1 Currency allocation contribution
0 Yield component
4 Security selection contribution
6 Portfolio return

Front 4

Currency Management
for Global Performance Evaluation

(SS17)

Back 4

Active currency management or currency hedging?
1. Passive currency management means manager is fully hedged against currency exposure NOT that s/he does nothing. Passive is defined relative to the benchmark being assigned.
2. Active means taking risk deliberately on currency.
3. Major issue in assessing currency mis/management is the benchmark and whether the benchmark is hedged or not.
4. Can also assess non-currency returns and use a 'currency overlay manager' to manage the currency.
5. Derivates: used to manage currency exposure like forwards/futures

Front 5

Multiperiod Attribution Analysis
(where you cross multiply..)

(SS17)

Back 5

Two periods
i. Portfolio returns 25% y1 and 20% y2
ii. Market returns 20% y1 and 10% y2
Therefore the contribution of security selection (ATTRIBUTE RETURN) was 5% in y1 and 10% in y2.
NOT to sum returns across periods
NOT to just chainlink attributes
OK to chainlink benchmark then
OK to chainlink pf
OK to take the difference of the two

1. Past attribute (y1 10%) compounds with benchmark in y2 if there is no active decisions in y2
2. If there IS active decisions in y2, the contribution of choice is earned on pf value at end of y1.

Front 6

Performance Appraisal for Global
in terms of Risk

(SS17)

Back 6

Two types of risk measures
1. Total or Absolute Risk (no benchmark) - all about standard deviation
2. Relative Risk (tracking error = returns in excess of benchmark)
er = rP - rB
Tracking Risk = ACTIVE RISK
σ = √ [ 1/(t-1) sum(er - average er)²]
3. Short term good returns may be due to chance, use long evaluation horizon. Performance may be coming from taking extra risk, rather than any particular skill.

Front 7

Risk-adjusted Performance
International Portfolio

(SS17)

Back 7

1. Sharpe Ratio
i. known as reward to variability
ii. reward is return over risk free
iii. variability is total risk of the pf
iv. useable only for global portfolio
2. Information Ratio - relative to benchmark. Is the excess return (alpha) large relative to the tracking error.
excess return (alpha)
IR = --------------------
tracking error
Managers here deliberately take active bets and knowingly takes on tracking error.
IR = 0.5 is good
IR = 1 is exceptional

Front 8

Risk Budgeting

(SS17)

Back 8

Depending on which managers have which skills leads you to how much risk to allocate to them.
i. Some are allowed large tracking error (because they are seen as able to add value)
ii. Some have small allowed tracking error
1. Absolute risk allocation comes down from the asset allocation approach
2. Relative risk allocation comes from active management away from benchmarks.

Front 9

Biases in Return and Risk for global performance evaluation

(SS17)

Back 9

Switching managers is costly.
Main problem is differentiating between luck and skill that will lead to superior performance in the future.
i. Infrequently traded assets (smoothing and downward bias on risk)
ii. Options (understate true losses)
iii. Survivorship bias (for returns, +ve return managers don't get fired)
iv. Survivorship bias (for risk, large -ve return hedge funds disappear)

Front 10

Implementation of Performance Evaluation
(Global Benchmarks)

(SS17)

Back 10

1. Manager Universes (peer group)
NOT comparing apples with apples. Managers have different styles and strategies. If they manager is told that they can pursue whichever strategy then this is an appropriate benchmark.
If they are told to be mid-cap (i.e. a specific benchmark) then it is an unfair benchmark
2. Customized Benchmarks
OK because it compares like with like, not against all managers, just against managers that are out there with the same style/strategy as the manager being assessed.
3. Customised benchmarks consider
i. Individual country weights (weight on their GDP or index)
ii. Countries, industries and styles
iii. Currrency hedging

Front 11

Global Benchmarks

(SS17)

Back 11

1. Individual Country/ Market Weights
NOT good to use a GDP adjusted weight because you are constantly rebalancing.
2. Countries, Industries and Style
Is industry allocation now more important than country allocation?
A value manager assigned an international value benchmark is good allowing him/her to invest internationally in value stocks.
3. Currency Hedging
Is the manager taking currency risk
Do they have a currency overlay manager
If there is hedged risk, then a hedged benchmark should be used.
If there is active risk, then it shouldn't be benchmarked against a hedged benchmark.
Arbitrary hedge ratio of 50%

Front 12

Evaluating Portfolio Performance

(SS17)

Back 12

Importance of assessing performance because
FUND SPONSORS PERSPECTIVE
1. It is quality control for the fund sponsor. Control mechanism.
2. It is the feedback step and brings everything full circle back to the investment policy
3. Gives evidence to the trustees that the money is being invested in an appropriate manner.

NOT ok to invest in wacky strange things that have not been authorised.

INVESTMENT MANAGERS PERS.
1. Shows them where they are being effective/ineffective

Front 13

Three components of Performance Evaluation

(SS17)

Back 13

Performance Evaluation is based on
1. Performance Measurement
What was the account's performance?
2. Performance Attribution
Why did it produce that performance? (relative to benchmark)
3. Performance Appraisal
Is it luck or skill (what is the quality)

Front 14

Performance Measurement without intraperiod external cash flows

Back 14

Just adding or subtracting assets into the account shouldn't affect return.
1. MV₁ - MV₀
rf = ------------------------
MV₀

2. Cash flow at the beginning
MV₁ - ( MV₀ + CF )
rf = ---------------------------------
MV₀ + CF

3. Cash flow at the end
( MV₁ - CF) - MV₀
rf = ------------------------------
MV₀

If you look at performance on a daily basis then this cash flow issue doesn't really matter.

Front 15

Time-Weighted Rate of Return

(SS17)

Back 15

Time weighting revalues the account ever time an external cash flow occurs. It's WEALTH RELATIVE i.e. compounded for $1 invested (money-weighted is IRR or compound rate of ALL funds invested)
1. Shows compound growth
Subperiod 1 = Days 1-5 $30,000
End Day 5 = $1,045,000
Subperiod 2 = Days 6-16 $20,000
End Day 16 $1,060,000
Subperiod 3 = Days 17-30
End $1,080,000

Return for sub1
= [($1,045,000 - $30,000) - $1,000,000]/ $1,000,000
= 0.0150 SO 1.5%

Subperiod returns can be chainlinked