Looking for the optimal value tilt
Northfield Annual Research Conference,
The Greenbrier
October 25, 2006
Edouard Senechal, CFA, FRM
Senior Risk Manager
Not intended for public distribution. For important additional information, please see the Additional Disclosures at the end of the presentation.
Table of contents
Section 1
Simple hedge of a value exposure
Section 2
Assessing the value premium
Section 3
Measuring the impact of the value tilt on the information ratio
Section 4
Finding the optimal exposure
1
SECTION 1
Simple hedge of a value exposure
SB – I assume we’re talking
about GSP – penciled that in
Asset allocation process
Portfolio is broadly diversified through top-down and bottom-up input from
Investment Management & Research
Asset allocation
Risk management
Currency
Global Investment Solutions
Global Securities
Portfolio
Security selection
US equities
Global(ex-US)
equities
Emerging
markets equity
US
bonds
Global(ex-US)
bonds
High yield
bonds
Emerging
markets debt
Performance of a portfolio with value tilt
Long-term performance does not help in the short term!
Cummulative Returns since May 1991
8
Portfolio
Benchmark
7
Active
6
5
4
3
2
1
0
-1
Jan90
Jul92
Jan95
Jul97
Jan00
Portfolio: U.S. Large Cap Equity Model since inception Benchmark: Russell 1000.
Jul02
Jan05
Jul07
Source : UBS Global Asset Management
♦ How can we hedge this exposure?
♦ In order to hedge our value exposure we need to measure it.
4
Measuring the value tilt: The academic perspective
♦ Fama and French (1992): Book value to Market Value (B/P) is a better measure of
value/growth than earnings to price (E/P). B/P explains a greater proportion of the
cross-section of asset returns. The role of E/P in explaining the cross-sectional
variations of asset returns can be captured by a combination of asset size and B/P.
♦ Lakonishok et. al (1994): find a larger spread between value and growth portfolio
performance when defining value/growth with cash flow to price (CF/P) or with a
combination of CF/P and past growth in sales.
♦ Lie and Lie (2002): Test the accuracy of the different valuation multiples used to
assess corporate value. B/P provides the most accurate estimate of corporate
value when using industry average multiples.
♦ Chan and Lakonishok (2004) use a cross-sectional regression to estimate how to
combine B/P, CF/P, E/P and S/P (sales to prices) to get to an overall value
indicator.
Sources:
Fama, E., French, K.,“The Cross-Section of Expected Stock Returns” - The Journal of Finance, 1992
Lakonishok, J., Shleifer, A., Vishny, R.,. “Contrarian investment, extrapolation, and risk.” Journal of Finance 49, 1541–1578. 1994
Chan,L., Lakonishok, J. “Value and Growth Investing: Review and Update” Financial Analysts Journal January 2004, Vol. 60, No. 1: 71-86
5
Lie, E., Lie, H., “Multiples Used to Estimate Corporate Value” Financial Analysts Journal, Mar 2002, Vol. 58, No. 2: 44-54.
Measuring the value tilt: Industry standards
♦ Model providers and index providers each have a different definition for value.
♦ However, B/P is the most commonly used measure for value.
Northfield
(US model)
MSCI
S&P
Citigroup
Russell
Barra
Style
Research
B/P
X
X
X
X
X
X
Growth*
X
X
X
X
X
X
D/P
X
X
X
X
X
E/P
X
X
X
X
C/P
X
X
S/P
X
X
EBITDA/P
* Growth is defined differently across vendors
X
Source : UBS Global Asset Management
B/P= Book to Price, D/P= Dividend to Price, E/P= Earnings to Price, C/P= Cash-Flow to Price, S/P= Sales
to Price and EBITDA/P=Earnings Before Interest, Taxes, Depreciation and Amortization to Price
6
Performance with a value hedge
♦ Futures or total return swaps
Cummulative Performance since Sept- 98
0.6
Hedged Portfolio
– Hedging costs = negative alpha
Un-Hedged Portfolio
0.4
♦ Invest in a value fund
0.2
– Positive alpha
– Diversification of asset
selection risk
0
-0.2
-0.4
-0.6
-0.8
Jan98
Jan00
Jan02
Jan04
Jan06
Un- Hedged Portfolio: U.S. Large Cap Equity Model; Hedged Portfolio: Simulated performance of a portfolio mixing U.S. Large Cap Equity Model
and Large Cap Growth Model Portfolio. Active returns versus the Russell 1000 Growth
Source : UBS Global Asset Management
♦ In this example, the “value hedge” consists of an allocation to a growth portfolio.
The Growth and Core portfolios were rebalanced on a monthly basis in order to
maintain a B/P ratio equal to that of the Russell 3000 Index.
Un-Hedged Portfolio: U.S. Large Cap Equity Model; Hedged Portfolio: Simulated performance of a portfolio mixing U.S. Large Cap Equity Model and Large Cap Growth
Model Portfolio. The weight of the Growth portfolio changes so that the Value tilt of the overall portfolio=0. Active returns are computed versus the Russell 1000 Growth
(i.e. portfolio returns – performance of the Russell 1000). The performance shown is a back-tested; not realized performance. The model assumes constant rebalancing
guidelines throughout the period. The model performance does not take into account transaction costs linked to the rebalancing of the portfolios between the two
7
underlying portfolios. 1. The model assume adjustments for risk capital on a monthly basis after the close of business on a spec ific day. Live portfolios may not adjust
allocations at exactly the same time.
Does it make sense to fully hedge?
♦ A fully hedged strategy performed better than an unhedged strategy.
♦ This does necessarily mean that we need to completely remove the Value bias.
♦ If we maintain a value-neutral profile at all times, we would give up the value
premium.
♦ We need to measure the risk and return of the value premium to assess how much
of the value tilt we want to keep in our portfolio .
We can’t implement forward-looking strategies
the tech-bubble period as a reference point.
using
8
SECTION 2
Assessing the value premium
The academic perspective
♦ Fama and French (1992) find that B/P is the most important factor in explaining the
cross-section of asset returns.
– The results of Fama and French were replicated numerous times in different markets over
different horizons.
♦ Does the value premium reflect higher risk of value stocks?
– “If asset pricing is rational, size and BE/ME must proxy for risk. ... If stock prices are
irrational, however, the likely persistence of the results is more suspect”
Fama
& French (1992)
♦ We have two competing explanations:
– Behavioral finance: Cognitive bias and agency costs are the source of the value premium.
– Efficient market hypothesis: Value stocks are riskier since they are more prone to financial
distress.
10
The academic perspective
♦ What is the magnitude of the value premium?
♦ Chan and Lakonishok (2004) refined the definition of value and growth based on
BV/MV (Book Value to Market Value), CF/P, E/P, and the sales-to-price ratio
Average Annual Return
Standard Deviation of Annual Returns
Information Ratio
Spread Decile 10
Spread Decile 9 & 10 Spread Decile 9 minus
Decile 2
minus Decile 1
minus Decile 1 & 2
27.10%
21.07%
15.95%
10.14%
9.31%
8.48%
0.37
0.44
0.53
Decile 10 contains Value stocks and decile 1 contains Glamour or Growth stocks
Source : Chan and Lakonishok (2004) and UBS Global Asset Management
♦ However, we cannot invest in deciles portfolios; we are considering portfolios
benchmarked against value, growth or core indices.
11
Performance of Russell value and growth indices
♦ When using market cap-weighted indices instead of a deciles portfolio the value
premium is less attractive.
♦ Nevertheless we still observe an IR between 0.15 and 0.20.
Distribution of the Monthly Returns Differences between the Russell 3000 Value and the Russell 3000 Growth
(From January 1979 to September 2006)
40
35
Number of Obervations
30
25
Monthly Returns
Russell
3000
Russell 3000
Value
Mean
Median
Standard Deviation
Min
Max
5th Percentile
Number of Observations
1.14%
1.49%
4.38%
-22.4%
12.8%
-5.9%
333
1.22%
1.42%
4.02%
-20.8%
13.2%
-5.0%
333
Annual Returns
Russell
3000
Russell 3000
Value
Mean
Median
Standard Deviation
Min
Max
5th Percentile
Number of Observations
14.33%
17.27%
15.52%
-21.5%
36.8%
-12.5%
28
15.33%
18.30%
13.44%
-15.2%
37.0%
-9.5%
28
Russell 3000
Russell 3000
Value - Russell
Growth
3000 Growth
1.05%
0.16%
1.26%
0.19%
5.14%
2.89%
-24.0%
-12.8%
14.2%
14.1%
-7.4%
-4.0%
333
333
20
15
10
5
0
-10%
-5%
0
5%
10%
15%
Russell 3000
Russell 3000
Value - Russell
Growth
3000 Growth
13.29%
2.04%
13.13%
1.07%
19.60%
12.85%
-28.0%
-27.2%
41.7%
30.5%
-23.0%
-22.1%
28
28
Source : FactSet and UBS Global Asset Management
12
Impact of the strategy’s beta
♦ From an asset allocation standpoint, we need to pay attention to the fact that a
value tilt will affect other parameters; for example beta:
♦ Historical beta of Russell 3000 Value – Russell 3000 Growth is -0.27
Value-Growth Spread vs. Russell 3000
0.15
Dec-1991
Nov-1980
0.1
Aug-1982
May-1990
Jan-1985
Apr-2001
Dec-1999
Jan-1987
Oct-1982
Aug-1984
Dec-1998
0.05
Apr-1999
Jun-2000
Jan-2001
Nov-1999
Oct-2001
Dec-2000
Feb-2000
Russell 3000
0
Oct-2000
Jan-1984
Sep-2000
-0.05
Mar-2001
Aug-1990
Sep-2002
-0.1
Sep-2001
Nov-2000
Feb-2001
Mar-1980
-0.15
Aug-1998
-0.2
Oct-1987
-0.25
-0.2
-0.15
-0.1
-0.05
0
Russell 3000 Value - Russell 3000 Growth
0.05
0.1
0.15
Source : FactSet and UBS Global Asset Management
13
Is the value premium only true for indices?
♦ According to some studies the value premium is less important for active
managers.
♦ Median IR across value managers was 0.13, while it was 0.33 across growth
managers.
Distribution of Large Cap Growth Managers Information Ratios (12/2000 - 12/2005)
Distribution Large Cap Value of Information Ratios (12/2000 to 12/2005)
30
20
18
25
16
14
20
12
15
10
8
10
6
4
5
2
0
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
0
-1.5
-1
-0.5
0
0.5
1
1.5
2
Source: UBS Global Asset Management and PSN database
14
Not if we account for survivorship bias…
♦ Survivorship bias is the most likely explanation.
Source: John C. Bogle “The Mutual Fund Industry 60 Years Later: For Better or Worse?” Financial Analyst Journal,
January/February 2005
15
Principles behind our allocation to value
♦ The value premium does not reflect higher risk for value stocks.
♦ The value premium does not disappear if we allocate to active managers instead
of indices.
♦ The value premium is not risk free, we need to balance the active return and
active risk it brings to our portfolio.
♦ If we consider indices instead of portfolio of deep value and deep growth stocks,
the risk-adjusted return of the value tilt is less important.
♦ From an asset allocation perspective we need to base our decision process on
the performance of indices.
16
SECTION 2
Measuring the impact of the value tilt on
the information ratio
Measuring the impact of the value tilt on the portfolio IR
♦ How much of our alpha is due to our exposure to the Russell Growth Index and the
Russell Value Index?
♦ Approach 1: Brinson based decomposition of returns:
g
g
v
v
α = wPORT
+ wPORT
− R1000
R1000
R1000
g
g
v
v
) wg + ( RPORT
) wv
+ ( RPORT
− R1000
− R1000
Growth stocks
selection
Value stocks
+
α V / G =Style rotation alpha
α S =Security selection
+cross-product
selection
Source : UBS Global Asset Management
18
Measuring the impact of the value tilt on the portfolio IR
v
∆
♦ Let
LCC be the value-stock weight in excess of 50% of the fund:
w g =0.5-∆vPORT and wv =0.5+∆ vPORT
♦ The style rotation alpha of the portfolio is:
v
g
Style Rotation Alpha = ∆ v ( R1000
− R1000
)
♦ The style rotation active risk is:
Style Rotation Active Risk=∆vσ
g
−R
1000 1000
Rv
Source : UBS Global Asset Management
19
Measuring the impact of the value tilt on the portfolio IR
♦ Now we can quantify the elements that will enter in our decision:
IR =
v
g
 w1∆1v + (1 − w1) ∆v2  ( R1000
− R1000
) + w1α1S + (1 − w1 )α 2S


2
 w1∆1v + (1 − w1 ) ∆v  σ 2
2

Rv
g
−R
1000 1000
+ w12σ 12 + (1 − w1 ) 2σ 22 + 2 w1 (1 − w1 )σ 1σ 2 ρ12
– The value tilt of each fund
– The value premium
– The risk associated with the value premium
– The pure security selection active return of each fund
– The pure security selection risk of each fund
– The correlation between the pure security selection active returns of the two funds
♦ Second alternative, measure the historical delta or use a multi-factor model:
v
g
Cov( Fund , R1000
− R1000
)
∆=
v
g
σ 2 ( R1000
− R1000
)
Source : UBS Global Asset Management
20
Measuring the impact of the value tilt
♦ Using Brinson approach or model approach leads to similar results.
Brinson vs. Beta based measure of Value Tilt for Large-Cap Core
0.5
0.4
0.2
0.1
0
M
ar
-9
Se 0
p9
M 0
ar91
Se
p9
M 1
ar9
Se 2
p9
M 2
ar
-9
Se 3
p9
M 3
ar
-9
Se 4
p9
M 4
ar9
Se 5
p9
M 5
ar96
Se
p9
M 6
ar
-9
Se 7
p9
M 7
ar98
Se
p9
M 8
ar9
Se 9
p9
M 9
ar0
Se 0
p0
M 0
ar
-0
Se 1
p0
M 1
ar0
Se 2
p0
M 2
ar0
Se 3
p0
M 3
ar
-0
Se 4
p0
M 4
ar
-0
Se 5
p0
M 5
ar06
Value exposure
0.3
-0.1
-0.2
Barra Beta vs. VG Spread
Method 3 Brinson Framework
Source : UBS Global Asset Management
21
Measuring the impact of the value tilt
♦ However, when we consider a portfolio with a large tilt, the Brinson-based
methodology shows its limits.
Brinson vs. Beta based measure of Value Tilt for Large-Cap Growth
-0.1
Ju
l-0
2
O
ct02
Ja
n03
Ap
r-0
3
Ju
l-0
3
O
ct03
Ja
n04
Ap
r-0
4
Ju
l-0
4
O
ct04
Ja
n05
Ap
r-0
5
Ju
l-0
5
O
ct05
Ja
n06
Ap
r-0
6
Ja
n00
Ap
r-0
0
Ju
l-0
0
O
ct00
Ja
n01
Ap
r-0
1
Ju
l-0
1
Oc
t-0
1
Ja
n02
Ap
r-0
2
0
-0.2
Value exposure
-0.3
-0.4
-0.5
-0.6
-0.7
-0.8
-0.9
Barra Beta vs. VG Spread
Method 3 Brinson Framework
Source : UBS Global Asset Management
22
SECTION 2
Finding the optimal exposure
The parameter of the problem
♦ We estimate all the parameters of the problem:
The value tilt of each fund
Beta with respect to the spread between
the value and growth indices
Value premium & risk associated with the
premium
Historical analysis
The pure security selection active return of
each fund
Due diligence
The pure security selection risk of each
fund
Due diligence + risk model + historical
analysis
The correlation between the active returns of
the two funds
Due diligence + historical analysis
24
Impact on active returns
SB – second bullet – the steeper the
decline in returns?
Alpha of the combination of LCG and LCC vs. weight invested in
LCG for different Value Tilts of LCC
Expected Active Return
2.50%
2.00%
Value Tilt=0%
Value Tilt=5%
Value Tilt=10%
1.50%
Value Tilt=15%
Value Tilt=20%
Value Tilt=25%
Value Tilt=30%
Value Tilt=35%
1.00%
Value Tilt=40%
0.50%
0%
4.0
0%
8.0
0%
12
.00
%
16
.00
%
20
.00
%
24
.00
%
28
.00
%
32
.00
%
36
.00
%
40
.00
%
44
.00
%
48
.00
%
52
.00
%
56
.00
%
60
.00
%
64
.00
%
68
.00
%
72
.00
%
76
.00
%
80
.00
%
84
.00
%
88
.00
%
92
.00
%
96
.00
10 %
0.0
0%
0.00%
Weight of Growth
Source : UBS Global Asset Management
♦ As we increase our exposure to the growth fund, our alpha decreases since:
– We are losing the value premium.
– We assumed the selection alpha of the growth fund to be equivalent to the selection alpha
of the initial portfolio.
♦ The higher the value tilt in the initial portfolio, the steeper the decline.
25
Impact on active risk
Sigma of the combination of LCG and LCC vs. weight invested in
LCG for different Value Tilts of LCC
6.00%
Expected Active Risk
5.00%
Value Tilt=0%
4.00%
Value Tilt=5%
Value Tilt=10%
Value Tilt=15%
3.00%
Value Tilt=20%
Value Tilt=25%
Value Tilt=30%
Value Tilt=35%
Value Tilt=40%
2.00%
1.00%
0%
4.0
0%
8.0
0
12 %
.00
%
16
.00
%
20
.00
%
24
.00
%
28
.00
%
32
.00
%
36
.00
%
40
.00
%
44
.00
%
48
.00
%
52
.00
%
56
.00
%
60
.00
%
64
.00
%
68
.00
%
72
.00
%
76
.00
%
80
.00
%
84
.00
%
88
.00
%
92
.00
%
96
.00
10 %
0.0
0%
0.00%
Weight of Growth
Source : UBS Global Asset Management
♦ Active risk first declines as investing in the growth fund brings diversification benefits and hedges the value
exposure.
♦ Then risk increases as further investment in the growth fund creates more growth exposure.
♦ The risk decrease is more pronounced when the initial portfolio has a strong value tilt since adding growth
hedges this tilt.
♦ On the other hand, when there is no initial tilt, the diversification effects bring a small risk reduction; very
quickly the impact of taking a negative value exposure dominates the diversification benefits.
26
Impact on information ratio
Information Ratio of the combination of LCG and LCC vs. weight invested in
LCG for different Value Tilts of LCC
0.70
0.60
Expected IR
0.50
Value Tilt=0%
Value Tilt=5%
Value Tilt=10%
Value Tilt=15%
Value Tilt=20%
Value Tilt=25%
Value Tilt=30%
Value Tilt=35%
Value Tilt=40%
0.40
0.30
0.20
0.10
96
%
10
0%
92
%
88
%
84
%
80
%
76
%
72
%
68
%
64
%
60
%
56
%
52
%
48
%
44
%
40
%
36
%
32
%
28
%
24
%
20
%
16
%
8%
12
%
4%
0%
0.00
Weight of Growth
Source : UBS Global Asset Management
♦ The information ratio first increases under the impact of the diversification benefits
and the hedging of the value tilt, then decreases because of the negative value tilt.
♦ The diversification benefits are greater than the alpha loss due to a negative value
tilt.
27
How exact should the implementation be?
♦ The exact rebalancing of the portfolio results in small IR gains.
Information Ratio of the combination of LCG and LCC vs. weight invested in
LCG for different Value Tilts of LCC
Optimal Value Tilt and Optimal Weight in LCG for different Value Tilts in LCC
0.70
50%
0.60
40%
Weight of the Growth Portfolio
0.50
Value Tilt=0%
Value Tilt=5%
30%
Value Tilt=10%
Value Tilt=15%
Value Tilt=20%
20%
Value Tilt=25%
0.30
Value Tilt=30%
Value Tilt=35%
Value Tilt=40%
0.20
10%
Overall Value Tilt
0.10
0%
Total Value Tilt
84
%
88
%
92
%
80%
76
%
96
%
10
0%
Value Tilt of LCC
Weight LCG
72
%
0.00
60
%
64
%
68
%
-10%
56%
Initial Portfolio Tilt
40%
52
%
35%
48
%
30%
36
%
40
%
44
%
25%
32%
20%
28%
15%
24
%
10%
8%
12
%
16
%
20
%
5%
4%
0%
0%
Weight of LCC
0.40
Weight of Growth
Source : UBS Global Asset Management
28
Impact of the negative value tilt in the growth fund
♦ Variation in the tilt in the growth portfolio does not significantly impact the optimal
allocation.
Information Ratio assuming Correlation=0.1 and IR=0.4 for LCG
Information Ratio assuming Correlation=0.1 and IR=0.4 for LCG
Information Ratio assuming Correlation=0.1 and IR=0.4 for LCG
0.70
0.70
0.70
Value Tilt=0%
0.60
0.50
0.40
0.30
0.20
0.10
Value
Value
Value
Value
Value
Value
Value
Tilt=5%
Tilt=10%
Tilt=15%
Tilt=20%
Tilt=25%
Tilt=30%
Tilt=35%
0.60
Value
Value
Value
Value
Value
Value
Tilt=40%
Tilt=0%
Tilt=5%
Tilt=10%
Tilt=15%
Tilt=20%
0.40
Value
Value
Value
Value
Tilt=25%
Tilt=30%
Tilt=35%
Tilt=40%
0.60
0.50
0.50
Value Tilt=0%
Value Tilt=5%
Value Tilt=10%
Value Tilt=15%
Value Tilt=20%
Value Tilt=25%
Value Tilt=30%
Value Tilt=35%
Value Tilt=40%
0.30
0%
15%
30%
45%
60%
75%
0.10
0.00
0.00
90%
0%
15%
30%
Weight of Growth
45%
60%
75%
Information Ratio assuming Correlation=0.1 and IR=0.4 for LCG
0.50
Value Tilt=0%
Value Tilt=5%
Value Tilt=10%
Value Tilt=15%
Value Tilt=20%
Value Tilt=25%
Value Tilt=30%
0.40
0.30
45%
60%
75%
90%
90%
Value Tilt=0%
Value Tilt=5%
Value Tilt=10%
Value Tilt=15%
Value Tilt=20%
Value Tilt=25%
Value Tilt=30%
Value Tilt=35%
Value Tilt=40%
Value Tilt=0%
Value Tilt=5%
Value Tilt=10% 0.40
Value Tilt=15%
Value Tilt=20%
Value Tilt=25%
0.30
Value Tilt=30%
Value Tilt=35%
Value Tilt=40%
0.40
0.30
0.20
0.10
0.10
0.00
75%
0.50
0.50
0.20
0.10
60%
0.60
Value Tilt=35%
Value Tilt=40%
0.20
45%
0.70
0.60
Weight of Growth
30%
Information Ratio assuming Correlation=0.1 and IR=0.4 for LCG
Information Ratio assuming Correlation=0.1 and IR=0.4 for LCG
0.60
30%
15%
Weight of Growth
0.70
15%
0%
90%
Weight of Growth
0.70
0%
0.30
0.20
0.20
0.10
0.00
Value Tilt=0%
Value Tilt=5%
Value Tilt=10%
Value Tilt=15%
Value Tilt=20%
Value Tilt=25%
Value Tilt=30%
Value Tilt=35%
Value Tilt=40%
0.40
0.00
0.00
0%
15%
30%
45%
Weight of Growth
60%
75%
90%
0%
15%
30%
45%
60%
75%
90%
Weight of Growth
Source : UBS Global Asset Management
29
Backtesting the strategy
♦ A simple hedge of the Value exposure results in higher IR during the bubble.
♦ We cannot implement forward-looking strategies, using the technology bubble as a
reference point.
Source : UBS Global Asset Management
Un-Hedged Portfolio: U.S. Large Cap Equity Model; Hedged Portfolio: Simulated performance of a portfolio mixing U.S. Large Cap Equity Model and Large Cap Growth
Model Portfolio. The weight of the Growth portfolio switches between 15% and 25% depending on the Value tilt of the Large Cap equity model value exposure changes.
Active returns are computed versus the Russell 1000 Growth (i.e. portfolio returns – performance of the Russell 1000). The performance shown is the result of a backtest; not realized performance. The model assumes constant rebalancing guidelines throughout the period. The model performance does not take into account transaction
30 close of
costs linked to the rebalancing of the portfolios between the two underlying portfolios. The model assume adjustments for risk capital on a monthly basis after the
business on a specific day. Live portfolios may not adjust allocations at exactly the same time.
Conclusions
♦ A simple rebalancing process works just as well as complex rebalancing one.
♦ Automatic rebalancing may result in overriding decision from our bottom-up
process.
♦ The keys to our investment decision are:
– To keep the active risk coming from our value exposure in line with the premium it
generates;
– To take advantage of diversification benefits;
– To stick to a strong due diligence process; and
– To learn from past events to implement forward looking policies.
31
Additional disclosures
Past performance is no guarantee of future results. There is no guarantee that investment objectives, risk or return targets
discussed in this presentation will be achieved.
The opinions expressed in this presentation are those of the UBS Global Asset Management Business Group of UBS AG and are
subject to change. No part of this presentation may be reproduc ed or redistributed in any form, or referred to in any publication,
without express written permission of UBS Global Asset Management. This material supports the presentation(s) given on the
specific date(s) noted. It is not intended to be read in isolation and may not provide a full explanation of all the topics that were
presented and discussed.
Information contained in this presentation has been obtained from sources believed to be reliable, but not guaranteed. Furthermore,
there can be no assurance that any trends described in this presentation will continue or that forecasts will occur because economic
and market conditions change frequently.
The information contained in this presentation should not be considered a recommendation to purchase or sell any particular security.
There is no assurance that any securities discussed herein will remain in an account’s portfolio at the time you receive this information
or that securities sold have not been repurchased. The securities discussed do not represent an account’s entire portfolio over the
course of a full market cycle.
It should not be assumed that any of the securities transactions or holdings referred to herein were or will prove to be profitable, or
that the investment recommendations or decisions we make in the future will be profitable or will equal the investment performance of
the securities referred to in this presentation.
A client's returns will be reduced by advisory fees and other expenses incurred by the client. Advisory fees are described in Part II of
Form ADV for UBS Global Asset Management (Americas) Inc.
This presentation does not constitute an offer to sell or a solicitation to offer to buy any securities and nothing in this presentation shall
limit or restrict the particular terms of any specific offering. Offers will be made only to qualified investors by means of a prospectus or
confidential private placement memorandum providing information as to the specifics of the offering. No offer of any interest in any
product will be made in any jurisdiction in which the offer, solicitation or sale is not permitted, or to any person to whom it is unlawful to
make such offer, solicitation or sale.
Any statements made regarding investment performance expectations, risk and/or return targets shall not constitute a representation
or warranty that such investment objectives or expectations will be achieved. The achievement of a targeted ex-ante tracking error
does not imply the achievement of an equal ex-post tracking error or actual specified return. According to independent studies, exante tracking error can underestimate realized risk (ex-post tracking error), particularly in times of above-average market volatility and
increased momentum. Different models for the calculation of ex-ante tracking error may lead to different results. There is no
guarantee that the models used provide the same results as other available models.
Copyright © 2006 UBS Global Asset Management (Americas) Inc.
US-I