Chat Shipping and Handling A

1
Chat Shipping and Handling (A)

• International Finance
• Dick Sweeney

2
Take Aways

• Decision-making with portfolio analysis The
  mean-variance, risk-return approach
• Its mechanics its usefulness
• Comparison with Asset Pricing Model approaches
• Which approach for a particular decision?
• Applying scenario analysis
• Sample versus population moments

3
I. Analyzing Chats Problem

• How does including European operation look?
  Compared to US operation by itself
• Base case is US operation by itself
• Use risk-return tradeoff approach from portfolio
  theory
• Answer looks good. Combination has higher
  expected return, lower variance of return than
  U.S. alone

4
Risk-Return Trade-Off
EF
ER
Eur
With two assets choice set has only positions
on curve. With many assets, have positions on
and inside curve. With just two assets, still get
curve ? do not have to have large number of
assets to get boundary curve.
_
Com
US
Sigma
0
5
I. Analyzing (cont.)

• Combination "better" than US alone
• Why? Covariance is not negative
• negative covariance not necessary for result
• small enough covariance does the trick
• In this case, even large covariance is OK
• If combination not necessarily better, might use
  Sharpe measure (need risk-free rate)see graph.

6
II. Mechanics General approach(general,
computer program approach)

• Find means of both projects
• Find variances of both projects
• Find covariance of two projects
• Then
• Find mean of combination (need weights)linear in
  weights
• Find variance of combination (need covariance,
  weights)quadratic in weights

7
More

• Scenario returns () prob.
• US Europe
• 1. Boom times 23 28 .60
• 2. Hard times - 22 - 17 .30
• 3. Euro weakness 23 4 .10
• Note You can use scenario analysis in other
  problem, e.g., to get expected cash flows for NPV
  calculations
• Note Could use market in place of US for CAPM

8
More (cont.)

• Means (population! not sample!)
• ?Eur (28 x .60) (-17 x .30) (4 x .10)
  12.1
• ?US (23 x .60) (-22 x .30)(23 x .10)
  9.5
• In percentages (could use decimal equivalents)

9
More (cont.)

• Variances built on squared deviations
• ?2Eur (28 - 12.1)2 x .60 (-17 - 12.1)2 x .30
  (4 - 12.1)2 x .10 412.29 2
• ?2US (23 - 9.5)2 x .60 (-22 - 9.5)2 x .30
  (23 - 9.5)2 x .10 425.25 2
• ?Eur 20.305, ?US 20.62
• But diversification helps

10
More (cont.)

• Covariance built on products of deviations
• ?Eur,US (28 - 12.1) x (23 - 9.5) x .60
  (-17 - 12.1) x (-22 - 9.5)
  x .30 (4 - 12.1) x
  (23 - 9.5) x .10 392.85 2
• Variances, covariances are in percentage
  squared or decimal equivalent
• 10 / 100 0.10
• 102 / (100)2 102 / 10,000 0.001
• 392.85 / 10,000 0.039285, not 3.93

11
More (cont.)

• Combination wUS .50, wEur .50
• Can vary weights to trace out EF (see handouts)
• ?com (ERUS x wUS) (EREur x wEur)
  ( 9.5 x .50) (12.1 x .50)
  10.8
• ?2com ?2Eur (wEur)2 ?2US (wUS)2
  2 ?Eur,US (wEur) (wUS)
  425.25 (.50)2 412.29
  (.50)2 2 (392.85)
  (.50) (.50) 405.81 2
• ?com 20.145 lt ?Eur 20.305 lt ?US 20.62

12
Note on Covariances

• Covariance can also be found from
• Standard deviations, ?Eur and ?US, and
• correlation coefficient for RUS and Reur?US,Eur
  
• ?Eur,US ?US,Eur (?Eur x ?US)
• or ?US,Eur ?Eur,US / (?Eur x ?US)
• Market beta ?Z ?Z,Mkt / ?2Mkt
• ?Z,Mkt (?Z x ?Mkt) / ?2Mkt
• ?Z,Mkt (?Z / ?Mkt)

13
III. Deeper issues

• Project looks good in terms of covariance with US
  operation
• Perhaps should look good in terms of covariance
  with diversified portfolio
• Optimally diversified portfolios ? Asset Pricing
  Models
• Simplest APM ? CAPM
• Could use more general APMs ? more risk factors
  (Sweeney and Warga in packet)

14
III. Deeper Issues (cont.)

• Good Europe project might fail SML test
• In CAPM, need beta, risk-free rate, premium on
  market ? SML
• How does project look in SML analysis?
• Have to add some more information
• rf 6, ERM - rf 5, ?Mkt 15,
  ?Eur,Mkt ??
• Decision sensitive to data
• See Chat (B) on this

15
IV. Which analysis to use?

• Key is whether Chat decision makers are well
  diversified
• If not, likely want answers above
• If decision makers are well diversified, they
  should consider a ? analysis
• Even if the stockholders are well diversified,
  un-diversified managers might go with the
  analysis in II. above