Portfolio Construction

Ugo Pomante

Agenda
1.
2.
3.
4.
5.
6.
7.
8.

Introduction to the Portfolio Construction

Analysis
y of Financial Markets
Strategic Asset Allocation: Nave Portfolio Formation Rule
Strategic
g Asset Allocation: A Q
Quantitative Approach
pp
Nave versus Markowitz
Putting
g Markowitz at work
Heuristic techniques
Bayesian techniques

Agenda
1. Introduction to the Portfolio Construction

Portfolio construction: Is it easy?

At a first g
glimpse
p it might
g appear
pp
that
building a portfolio is easy:
you have
h
j t to
just
t aggregate
t different
diff
t
assets
Asset 1
Asset 2
Asset 3
Asset 4
Asset 5
Asset 6

Portfolio construction: Is it easy?

The answer is NO
..Unfortunately building a GOOD portfolio
that is able to satisfy the needs/expectations of
an investor is HARD.

In fact, in order to construct a good portfolio,

you have to make many difficult decisions.

Portfolio Construction: Problems to be solved

Selection of Financial Markets where to invest the
money
y ((Liquidity,
q
y, Bonds,, Equities,
q
, etc)) : Which?
How many?
Estimation of future trend of the markets selected.
Construction of an optimization model that returns
the optimal portfolio (that returns the optimal
weights of the Financial Market) in the long run.
run
Development of an evaluation model that it is able
to verify
y that the p
portfolio selected satisfies the
needs of the investor.
Development of a market timing model, useful in
order
d
t make
to
k tactical
t ti l changes
h
t the
to
th portfolio
tf li
composition, in order to anticipate bull/bear trends.
Selection of the best financial products for every
financial market.
6

A mistake? Very dangerous!

Adverse selection of Financial Markets (too much risky
or poorly performing markets)
Error in Estimation of future trend of the markets
selected or overconfidence in the ability of prediction.
Construction of a weak optimization model.
p
y to verify
y that the p
portfolio selected satisfies
Incapacity
the needs of the investor.
Errors in market timing decisions (increase of the
e uity weight
equity
ei ht at the beginning
be i i of a bear
bea trend).
t e d)
Adverse selection of products for every financial market
(poor performance or high costs).
costs)

Dont under-estimate the process of portfolio

construction
Mistakes can be deadly
deadly .
So, it is necessary to have:
o skilled human resources;
o good IT procedures;
o consistent models of Portfolio Construction.

Well-Organized
Well
Organized procedures
Institutional investors (pension funds,
funds mutual
funds, etc) organize the process of portfolio
construction on stages...
stages
where every phase is able either to create
(extra-performance) or destroy value (underperformance);
There are three main stages.

Stages of the Portfolio

Portfolio Construction
Construction
1. Strategic Asset Allocation

+
2 Tactical Asset Allocation
2.

+
3. Stock-Bond/Fund Selection

10

Stage 1: Strategic Asset Allocation (SAA)

Strategic Asset Allocation is:
the portfolio composed by financial markets (or
asset classes)..
.. that the investor must hold in the long run (all
the investment horizon).

11

Strategic Asset Allocation: Example

The investor has a 5-years investment horizon.
The Asset Manager builds a portfolio, composed by
financial markets (that is supposed to be coherent with the
risk tolerance of the investor).

10%
32%
Domestic Money Market
Domestic Bond Market
Foreign Bond Market

8%

50%

Equity Market

5 years

On average the portfolio composition is expected to be this one in the next

12
five years.

Stage 2: Tactical Asset Allocation (TAA)

Tactical Asset Allocation is:
The change made to the strategic composition in
order to anticipate bull/bear trends.
.. that the investor must hold in the short run
(next 13 months).

13

Tactical Asset Allocation: Example

The SAA is the following:
10%
32%
Domestic Money Market
Domestic Bond Market
Foreign Bond Market

50%

8%

Equity Market

But the Asset Manager has the expectation that in the

next 3 months the Equity Market will decrease.
decrease
So, for the next 3 months he suggests the following
changes in the portfolio composition:
20%

22%

10%
32%
Domestic Money Market

Domestic Money Market

Domestic Bond Market

Foreign Bond Market

8%

50%

Domestic Bond Market

8%

Foreign Bond Market

Equity Market

50%

Equity Market

After three months, the tactical portfolio will be dismantled (and the strategic
14
portfolio will be resumed)..may be we will create e new tactical solution.

Stage 3: Stock-Bond/Fund Selection

Stock-Bond/Fund Selection is:
the process to select the best product for every
market in the portfolio.
You can (alternatively):
o directly select stocks & bonds (stockbond
(stock bond
selection);
o indirectly
i di tl select
l t stocks
t k & bonds,
b d identifying
id tif i the
th
best fund managers (fund selection).

15

Stock-Bond/Fund Selection : Example (1/2)

A French
F
h Pension
P i Fund
F d has
h the
th following
f ll i SAA:
SAA
10%
32%
Domestic Money Market
Domestic Bond Market
Foreign Bond Market

8%

50%

Equity Market

Th
The board
b d off directors
di t
d
does
nott have
h
th ability
the
bilit off
directly selecting the stocks/bonds.
so the Pension Fund identifies,
identifies for every market,
market the
fund managers that are supposed to be the best ones:
M k t (Asset
Markets
(A t Cl
Classes))
Domestic Money Market
Domestic Bond Market
o e g Bond
o d Market
a et
Foreign
Equity Market

Funds
F
d selected
l t d
MS Euro Liquidity Fund
Parvest Euro Gov. Bonds
JPM Global Bonds
16
Fidelity International

Stock-Bond/Fund Selection : Example (2/2)

10%
32%
Domestic Money Market
Domestic Bond Market

From
Markets....

Foreign Bond Market

50%

8%

Equity Market

10%
32%
MS Euro Liquidity Fund
Parvest Euro Gov. Bonds

..to
to
Products.

JPM Global Bonds

8%

50%

Fidelity International

17

The pillars of Asset Allocation

In estor:
Investor:

A t Manager:
Asset
M

Investors preferences

Asset Managers expectations

about the future trend of
Financial Markets

Optimization Model

OPTIMAL PORTFOLIO
18

Agenda
2. Analysis
y of Financial Markets

19

From Investors Preferences to

Fi
Financial
i l Markets
M k Evaluation
E l i
It is
i well
ell known
k o
that Investors:
I e to
o love return;
o hate risk (are risk adverse ).
So, if we want to build a p
portfolio that best suit
the investors preferences, we need to know
the riskreturn p
profile of Financial Markets and
Market Portfolio.
Ri k R
RiskReturnanalysisofFinancialMarkets
l i f Fi
i lM k

20

The Financial Markets

Analysis of the following Asset Classes.

ASSET CLASSES:
- Money Market EMU
- Bond
B d Market
M k t EMU
- Bond Market World
- Equity Market Europe
- Equity Market North America
- Equity Market Japan
- Equity
q y Market Pacific ex Japan
p
- Equity Emerging Markets

MARKET INDEXES:
- JPM Euro 3 Months
- Citygroup
Cit
EMU Aggr.
A
all
ll maturities
t iti
- JPM Global
- MSCI Europe
- MSCI North America
- MSCI Japan
- MSCI Pacific ex Japan
p
- MSCI Emerging Markets

21

Historical series of annual returns

Thanks to the market indexes we have a set of historical
returns of financial markets:
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
998
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008

JPM Euro 3
months
7,28%
9,16%
11,55%
10,41%
11 11%
11,11%
9,03%
6,30%
6,58%
4,83%
4,42%
4,46%
3,15%
4,32%
4,74%
3,53%
2,54%
2,18%
2,20%
3,02%
4,42%
5,75%

Citygroup EMU All

Maturities
4,30%
1,40%
3,10%
11,37%
12 80%
12,80%
14,44%
-1,84%
16,27%
7,29%
6,16%
10,94%
-2,97%
8,39%
6,25%
8,49%
3,77%
7,56%
5,67%
-0,28%
0,97%
9,97%

JPM Global MSCI Europe

17,89%
1,84%
-0,96%
17,13%
11 33%
11,33%
20,41%
-9,60%
10,18%
12,41%
18,31%
6,82%
10,67%
10,49%
4,75%
0,31%
-4,92%
2,09%
7,93%
-5,11%
-0,86%
18,47%

26,59%
19,57%
-17,12%
11,58%
-1 35%
-1,35%
35,53%
-10,63%
9,77%
27,59%
41,85%
17,21%
33,06%
-2,46%
-16,83%
-32,86%
11,92%
9,28%
23,01%
16,65%
-0,73%
-45,21%

MSCI North
America
25,56%
20,59%
-17,05%
27,60%
9 64%
9,64%
15,27%
-11,72%
23,41%
30,93%
52,36%
17,75%
42,14%
-5,84%
-8,77%
-35,81%
6,09%
1,40%
21,23%
1,53%
-5,46%
-35,73%

MSCI Japan
51,41%
-3,38%
-43,67%
9,86%
-17 04%
-17,04%
33,65%
7,76%
-7,64%
-9,54%
-11,52%
-3,41%
3,41%
87,20%
-22,85%
-25,97%
-25,17%
11,76%
6,38%
43,29%
-5,87%
-15,38%
-26,51%

MSCI Pacific ex
Japan
40,98%
6,05%
-24,16%
32,86%
10 01%
10,01%
88,14%
-25,11%
1,77%
26,84%
-21,47%
-16,21%
16,21%
62,47%
-10,91%
-7,26%
-23,53%
17,29%
15,57%
27,27%
14,68%
13,77%
-49,45%

MSCI Emerging
Markets
51,45%
51,78%
-23,58%
58,26%
16 11%
16,11%
83,69%
-18,48%
-14,07%
11,89%
1,03%
-32,85%
32,85%
90,86%
-26,37%
0,40%
-22,66%
25,87%
13,54%
50,45%
15,72%
22,10%
-51,85%

((in Euro))

22

 Investors aim to maximise the return (the

final value) of their investments

23

Average Return (1/2)

Investors love financial markets with higher average returns:

R1 R2 ..... RT 1 RT
R
T

R
Excel:
l
=average(Historical series)

Rt
t 1

(1988-2008)

JPM Euro 3
months

Citygroup EMU
All Maturities

JPM
Global

MSCI
Europe

MSCI North
America

MSCI
Japan

MSCI Pacific
ex Japan

MSCI
Emerging
Markets

Average off A
A
Annuall
Returns

5,76%

6,38%

7,12%

7,45%

8,34%

1,59%

8,55%

14,44%
24

Average of Annual Returns (2/2)

Average of Annual Returns (1988-2008)
16 00%
16,00%
14,00%
12,00%
10,00%
8,00%
6,00%
4,00%
2,00%
0,00%
JPM Euro Citygroup
3 months EMU All
Maturities

JPM
Global

MSCI
Europe

MSCI
North
America

MSCI
Japan

MSCI
MSCI
Pacific ex Emerging
Japan
Markets

25

Average Return of a Portfolio (1/3)

If we known:
- the
h Portfolio
P f li Weight
W i h off eachh market
k (w
( i);
)
- the Average Returns of each market ( R)i
The estimation of the Portfolio Average Return is straightforward:
k

RPort wi Ri
i 1

Excel:
=sumproduct(Weights,
d t(W i ht Average
A
Returns)
R t
)

26

Average Return of a Portfolio (2/3)

Using Matrices:

RPort w1

w2 wi

R1

R2


wk
R
i

R
k

27

Average Return of a Portfolio (3/3)

JPM Euro 3
months

Citygroup EMU
All Maturities

JPM
Global

MSCI
Europe

MSCI North
America

MSCI
Japan

MSCI Pacific
ex Japan

MSCI
Emerging
Markets

W i ht
Weights

5 00%
5,00%

40 00%
40,00%

5 00%
5,00%

17 00%
17,00%

23 00%
23,00%

3 00%
3,00%

2 00%
2,00%

5 00%
5,00%

Average of Annual
Returns

5,76%

6,38%

7,12%

7,45%

8,34%

1,59%

8,55%

14,44%

5 00%
5,00%
2,00%

5,00%

3,00%

23,00%
40,00%

JPM Euro 3 months

Citygroup EMU All Maturities
JPM Global
MSCI Europe
MSCI North America
MSCI Japan
MSCI Pacific ex Japan
MSCI Emerging Markets

RPort wi Ri 7.32%
i 1

17,00%
5,00%

Investors want to maximise the average (or expected) return of

the portfolio.
28

 Investors are risk

risk adverse
adverse:: given a targeted
return they try to minimise the risk.

29

What is risk?
risk ?

(1/2)

The financial literature has formulated many mathematical &

statistical
t ti ti l indicators
i di t
useful
f l in
i order
d to
t capture

t risk.
i k
Examples:
Standard Deviation;
Semi Standard Deviation;
Downside risk;
Beta;
Duration/Modified Duration;
Value
V l at Risk
Ri k (VaR);
(V R)
Shortfall probability;
Tracking Error Volatility (TEV).
(TEV)

Manyindicators.maybeaphenomenon
Many
indicators
maybe a phenomenon
30
difficulttomeasure.

What is risk?
risk ?

(2/2)

Commonly
Co
o ly in
i the financial
fi a ial environment
e i o e t risk
i k is
i
interpreted as the uncertainty of returns;
So markets with volatile, unstable returns are
considered risky.
y
Graphical evidences
Graphicalevidences

31

Very Low volatility: Money Market EMU

Annual Return of JPM Euro 3 months (Money Market EMU) [1988-2008]
100%

80%

60%

40%

20%

5.76%
0%
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
-20%

-40%

-60%

- Low Interest Rate Risk

32

Low volatility: Bond Market EMU

Annual Return of JCitygroup EMU All Maturities (Bond Market EMU) [1988-2008]
100%

80%

60%

40%

20%

6.38%
0%
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
-20%

-40%

- Higher Interest Rate Risk ( maturity)

-60%

- Credit Risk ((corporate

p
bonds))
33

Middle volatility: International Bond Market

Annual Return of JPM Global (International Bond Market) [1988-2008]
100%

80%

60%

40%

20%

7.12%
0%
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
-20%

-40%

- High Interest Rate Risk ( maturity)

-60%

- Credit Risk (corporate bonds)

- Exchange Risk

34

High volatility: European Equity Market

Annual Return of MSCI Europe (European Equity Market) [1988-2008]
100%

80%

60%

40%

20%

7.45%
0%

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
-20%

-40%

- High Equity Risk

-60%

- Very Low Exchange Risk (if domestic currency is )

35

Very High volatility: Emerg. Mkts Equity

Annual Return of MSCI EM (Emerg. Mkts Equity) [1988-2008]
100%

80%

60%

40%

20%

7.45%
0%
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
-20%

-40%

- Very
V Hi
High
hE
Equity
it Risk
Ri k

-60%

- High Exchange Risk (if domestic currency is )

36

Standard Deviation of Returns

Finally,
Fi ally wee need
eed a statistical
tati ti al indicator
i di ato able to
synthesise the volatility.
The most common parameter is the:

Standard deviation ()
Standarddeviation()
T

Ri R
i 1

T 1

Excel:
=stdev(Historical series)
37

 an easy interpretation

(1/2)

Annual Return of MSCI Europe (European Equity Market) [1988

[1988-2008]
2008]
100%

80%

60%

40%

20%

7.45%
0%

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
-20%

-40%

-60%

Standarddeviationcanbeseenasthe
averageofgapsbetweentheaveragereturn
g
gap
g
andeveryannualreturn.
38

 an easy interpretation

(2/2)

Annual Return of MSCI Europe (European Equity Market) [1988

[1988-2008]
2008]
100%

80%

60%

40%

20%

7.45%
0%

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
-20%

-40%

-60%

The standard deviation of MSCI Europe annual

return is 22.01%;
We can say
y that the annual return is likely
y to have an
average deviation from the average return of 22.01%.
39

Standard Deviations of Asset Classes

(1988-2008)

JPM Euro 3
months

Citygroup EMU
All Maturities

JPM
Global

MSCI
Europe

MSCI North
America

MSCI
Japan

MSCI Pacific
ex Japan

MSCI
Emerging
Markets

2,88%
,
1

5,11%
,
2

8,55%
,
3

22,01%
,
4

22,53%
,
5

29,95%
,
6

31,29%
,
7

37,93%
,
8

of Annual
Returns
RANKING

Standard Deviation of Annual Returns (1988-2008)

40,00%
35,00%
30,00%
25,00%
20 00%
20,00%
15,00%
10,00%
5,00%
0,00%
JPM Euro Citygroup
3 months EMU All
Maturities

JPM
Global

MSCI
Europe

MSCI
North
America

MSCI
Japan

MSCI
MSCI
Pacific ex Emerging
Japan
Markets
40

Standard deviation of a Portfolio: NOT a

weighted average (1/2)
If we known:
- the
h Portfolio
P f li Weight
W i h off eachh market
k (w
( i)
- the standard deviations of each market (i)
The estimation of the Portfolio standard deviation is NOT the
following:
k

Port wi i
i 1

That is, The portfolio standard deviation is NOT the

weighted average of the standard deviation of the markets.

41

Standard deviation of a Portfolio: NOT a

weighted average (2/2)
JPM Euro 3
months

Citygroup EMU
All Maturities

JPM
Global

MSCI
Europe

MSCI North
America

MSCI
Japan

MSCI Pacific
ex Japan

MSCI
Emerging
Markets

of Annual
Returns

2 88%
2,88%

5 11%
5,11%

8 55%
8,55%

22 01%
22,01%

22 53%
22,53%

29 95%
29,95%

31 29%
31,29%

37 93%
37,93%

Weights

5,00%

40,00%

5,00%

17,00%

23,00%

3,00%

2,00%

5,00%

5 00%
5,00%
2,00%

5,00%

3,00%

23,00%
40,00%

JPM Euro 3 months

Citygroup EMU All Maturities
JPM Global
MSCI Europe
MSCI North America
MSCI Japan
MSCI Pacific ex Japan
MSCI Emerging Markets

Port wi i 14.96%
i 1

17,00%
5,00%

42

NOT a weighted average: 1st empirical evidence

Using historical series of MSCI market indices on the time
horizon 2001-2008,, we measure the standard deviation of the
following equity market sectors:

- MSCI Europe Pharmaceutical, Pharm

=12,54%;
12,54%;
Ph
- MSCI Europe Biotechnology, Biotech= 30,32%;

Whichisthestandarddeviationofthe
MSCI Europe Pharma/Biotech?
MSCIEuropePharma/Biotech?

Pharma / Biotech 12.44%

It ca
cant be thee
weighted
average!
43

NOT a weighted average: 2nd empirical evidence

St d d Deviation
Standard
D i ti
off Annual
A
l Returns
R t
(1988-2008)
(1988 2008)

MSCI World
MSCI Emerging
Markets
MSCI Pacific ex Japan

MSCI Japan

MSCI North
N th America
A
i

MSCI Europe
0,00%

5,00% 10,00% 15,00% 20,00% 25,00% 30,00% 35,00% 40,00%

The world
Th
ld equity
it market
k th
has a standard
t d dd
deviation
i ti
off returns
t
that
th t is
i
lower than the standard deviation of all the country markets.
44
Again, risk cant be the weighted average.

The diversification effect

Since 1952 is well known that it is possible to reduce risk
avoiding concentration.
Proverb:
P
b Dont
D put your eggs in
i the
h same basket
b k
Fi
Financial
i l history
hi t
shows
h
th t markets
that
k t have
h
th tendency
the
t d
t
to
move one each other in a different way:

year 1998: MSCI Europe (+17.21%)

(+17 21%) vs MSCI EM ((32.85%)
yyear 1995: MSCI North America ((+23.41%)) vs MSCI
Japan (+1.77%)
Thanks
Th
k to
t the
th diversified
di
ifi d behaviour
b h i
off financial
fi
i l markets,
k t the
th
portfolio standard deviation is lower than the weighted
average.
g
45

We need to Capture the diversification effect

In order to measure the diversification effect ((that is,, the
power of diversification in reducing risk) we must
measure:

The Correlation ()

46

Correlation (): characteristics

(1/2)

o The correlation is calculated for a couple of markets;

o -1 +1
o If > 0, markets move in the same direction (both gain or

o
o
o
o

both lose)
If = +1,
+1 markets
k t perfectly
f tl move in
i the
th same direction
di ti
If < 0, markets move in opposite direction (one gains, the
other loses)
If = -1, markets perfectly move in opposite direction (they
move pperfectly
f y synchronised,
y
, but in opposite
pp
direction))
If = 0, markets are independent (no tendency to move in
the same or in the opposite direction)
(follows)

47

Correlation (): characteristics

(2/2)

o If =+1, no diversification
o If <+1, yes diversification
o The lower the correlation, the higher the diversification (the
risk
i k reduction)
d i )

Excel:
=correl(Historical series mkt1, Historical series mkt2)

48

Correlation: the scatter graph

Case 1: Positive correlation
MSCI North America
60,00%

1997

50,00%

40,00%

30,00%

20,00%

10 00%
10,00%

MSCI Europe
-50,00%

-40,00%

-30,00%

-20,00%

-10,00%

0,00%
0,00%

10,00%

20,00%

30,00%

40,00%

50,00%

-10,00%

-20,00%

2008

= +0.919

-30,00%

-40,00%

-50,00%

St
Strong
tendency
t d
to
t move in
i the
th same direction
di ti (20 times
i
on 21)
49

Correlation: the scatter graph

Case 2: Zero correlation
MSCI North America
60,00%

= +0.012
+0 012

50,00%

40,00%

30,00%

20,00%

10,00%

-20,00%

-15,00%

-10,00%

-5,00%

0,00%
0,00%

Citygroup EMU
All Maturities
5,00%

10,00%

15,00%

20,00%

10 00%
-10,00%

-20,00%

-30,00%

-40,00%

-50,00%

No tendency (12 times in the same direction - 9 times in opposite direction)

50

Correlation: the scatter graph

Case 3: Negative correlation
MSCI Japan
100 00%
100,00%

= -0.26

80,00%

60,00%

40,00%

20,00%

-20,00%

-15,00%

-10,00%

-5,00%

0,00%
0
00%
0,00%

Citygroup EMU
All Maturities
5,00%

10,00%

15,00%

20,00%

-20,00%

-40,00%

-60,00%

(7 times in the same direction - 14 times in opposite direction)

51

Correlation Matrix
The Correlation Matrix shows the correlations between all
the couples of markets:

(1988-2008)
Correlations

JPM Euro 3
months

Citygroup EMU
All Maturities

JPM
Global

MSCI
Europe

MSCI North
America

MSCI
Japan

MSCI Pacific
ex Japan

JPM Euro 3
months

Citygroup EMU
All Maturities

0,30

JPM Global

0,27

0,58

MSCI Europe

-0,09

-0,07

0,31

MSCI North
America

0,02

0,01

0,44

0,92

MSCI Japan

-0,21
0 21

-0,26
0 26

0 24
0,24

0 63
0,63

0 57
0,57

MSCI Pacific ex
Japan

0,05

0,01

0,31

0,70

0,55

0,75

MSCI Emerging
Markets

0,10
,

-0,20
,

0,25
,

0,68
,

0,60
,

0,78
,

0,91
,

MSCI
Emerging
Markets

1
52

Correlation Matrix with Excel

Insert here the

return series of all
the markets

53

TheGiftofglobalization
As showed by the correlation matrix, globalization has
strongly increased the correlation amoung equity markets.

Today traditional risky assets are not able to produce big

benefits of diversification.
This is the main reason why many institutional investors
suggest not to limit the investment to the classical asset classes
(bonds and listed stocks)
They suggest to invest money also in alternative
alternative investments
investments::
Hedge funds;
Commodities;
Pay attention: They do not
Private Equity;
show negative correlation!
54
Real Estate.

Standard Deviation of a Portfolio (1/2)

If we known:
- the portfolio weight of each market (wi)
- the standard deviations of each market (i)
- the correlations between couples
p of markets (i,j
i j)
The estimation of the Portfolio standard deviation is the following:

Port

w w
i

i j i, j

i 1 j 1

A two markets portfolio:

port ( w1 1 ) 2 ( w2 2 ) 2 2 w1 w2 1 2 12
55

Standard Deviation of a Portfolio (2/2)

Using Matrices:

Port

w1 1

w2 2 wi i

1,1

2,1
...

wk k
i ,1

k ,1

1, 2
2, 2
...

1,3 1, j 1,k w1 1
2,3 2, j 2,k w2 2

...

...

i,2

Corr
C

i, j

k ,2

i ,3

k ,3 k , j

...

i ,k wi i

k ,k wk k

=SQRT(MMULT(MMULT(rw,corrM),cw))
cw=transpose(rw)

56

Standard deviation of a Portfolio:

Numerical example
JPM Euro 3
months

Citygroup EMU
All Maturities

JPM
Global

MSCI
Europe

MSCI North
America

MSCI
Japan

MSCI Pacific
ex Japan

MSCI
Emerging
Markets

of Annual
Returns

2,88%

5,11%

8,55%

22,01%

22,53%

29,95%

31,29%

37,93%

Weights

5,00%

40,00%

5,00%

17,00%

23,00%

3,00%

2,00%

5,00%

Port wi w j i j i , j 11.44%
i 1 j 1

Port wi i 14.96%
i 1

57

Standard deviation: is it a good measure of risk?

Annual Return of MSCI Europe (European Equity Market) [1988-2008]
100%

80%

60%

40%

20%

7.45%
0%

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
-20%

-40%

-60%

Risk means bad returns, so we should focus only on volatility that has
58
negative consequences.

Semi-standard deviation (semi-)

This statistical indicator is perfect when you want to measure
the downside risk of a market
B t this
But
thi measure is
i rarely
l used.
d

Why?
It is difficult to measure the semi of a portfolio
It is difficult to build a model of portfolio optimization in
which
hi h the
h risk
i k is
i measured
d using
i semi-
i

Refusing the semi is a convenient solution!

59

From volatility to potential loss

Is it easy to interpret a measure of volatility?
Financial experience suggests that investors are not
able to interpret the meaning of standard deviation.
For investors risk means losses, not volatility..
so it can be useful to capture risk estimating the
portfolio
f li potential
i l loss.
l
Value at Risk (VaR)

60

Introduction to VaR models

An investor want to invest money in the European Equity
Market. His holding period is 1 year.
H wants to know
He
k
which
h h is the
h riskk off this
h equity market.
k
The standard deviation of annual returns is 22.01%.
Therefore the annual return is likely to have an average
Therefore,
deviation from the average return of 22.01%.

This statistical indicator is not able to tell

him which is the risk he would incur in case of
sizeable and exceptional losses (year:2008)
If he want to explore the darkest
darkest side
side of the risk,
risk
he need a VaR methodology.

61

VaR models: the aim

VaR models are able to estimate the potential losses.
For example, thanks to them we can say:
Do you want to invest 100,000 on European Equity Market?
Well, you have to know that in case of terrible financial events
you can lose 35,000!

A capital loss of 35% is very easy to understand!

62

VaR models: definition

Given a time horizon (=1 year), Value at
Ri k is
Risk
i the
h potential
i l loss
l
( 35%) where
(=-35%)
h
the confidence level (=98%) means that the
probability
b bili off higher
hi h losses
l
i 1-conf.level
is
fl l
(=2%).
Key elements:
1 Time horizon;
1.
2. Potential loss (not maximum loss);
3. Confidence level (1-c.l. is the prob. of higher losses).
63

VaR models: Calculation

(1/2)

We analyse a parametrical methodology named variancecovariance .

covariance
The statistical assumption of this method is the following:
Returns
Returns are normally distributed
distributed
180
160
140
120
100
80
60
40
20
0
64

VaR models: Calculation

(2/2)

Given this assumption, VaR is estimated as follows:

VaR R k
Va
AVERAGE RETURN

STANDARD
S
N
DEVIATION
V
ON

THE K VALUE IS RELATED TO THE CONFIDENCE LEVEL WE CHOOSE:

- IF c.l = 95% k = 1.65
- IF c.l = 98% k = 2.05
Statistical Tables
- IF c.l = 99% k = 2.33
65

VaR models: Example

1 year VaR of European Equity Market

REqu
7.45%
E . Europe
E

Equ
q . Europe
p 22.01%
l.c. 98% k 2.05
VaR R k 7.45% 2.05 22.01% 37,7%
Given a 1 year time horizon, the potential loss is -37.7%.
The pprobability
y of higher
g
losses is 2%.
%
If an investor wants to invest money on European Equity
Market he has to tolerate a -37.7% annual loss.
66

In the following
g analysis
y
we make the
assumption that we are an Asset Management
Committee involved in a Strategic Asset
Allocation process.
Our objective
j
is to identify
yag
good model to
build a SAA.

67

Agenda

3. Strategic Asset Allocation: Nave Portfolio Formation Rule

68

This first solution follows a

qualitative approach

Nave

portfolios
p

refuses

mathematical solutions.

69

NavePortfolioFormationRule:
A primitive approach
Aprimitiveapproach
Nave strategies:
are mathematics/statistics
th
ti / t ti ti free;
f
dont need optimization models;
don
dontt need numericalquantitative estimations; estimations
can be qualitativejudgemental (European Equity Market will
beat the Japanese Equity Market).

Nave strategies:
are easy to put into practice;
can generate good solution, never optimal ones;
generate portfolios that are usually diversified
reasonable.

and

70

NavePortfolioFormationRule:
Example
Example
(1/7)
We need to identify the SAA of a pension fund.
As members al the Asset Allocation Committee,
Committee we need to
manage a procedure able to identify the portfolio of asset
classes.
1. First, we select the asset classes where putting money:

Money Market EMU

Bond Market EMU
Equity Market Europe
q y Market North America
Equity
Equity Market Japan
Equity Market Pacific ex Japan
E it Emerging
Equity
E
i Markets
M k t

Safe
Assets
Risky
Assets
71

NavePortfolioFormationRule:
Example
Example
(2/7)
2. We identify the risk profile of the portfolio selected:

Given the expected risk tolerance of investors that will put

money in the pension fund we make the following decision:

Money Market
M
M k t EMU
Bond Market EMU
Equity Market Europe
Equity Market North America
Equity Market Japan
Equity Market Pacific ex Japan
Equity Emerging Markets

Safe
Assets

= 70%

Risky
Assets

= 30%

Risk profile can be captured with quantitative methdology (see

72
Cucurachi). But Nave portfolios refuses quantitative approach.

NavePortfolioFormationRule:
Example
Example
(3/7)
3. We select the weights inside the SafeAssets Group

In the time horizon of the investment we forecast a general

increase of interest rates in EMU area.
area
So, in order to maximise the expected return, we need to
reduce the maturity of the Safe
Safe-Assets
Assets Group.
Safe-Assets
Safe
Assets

Money Market EMU

Bond Market EMU
Sum

Weights
55%
15%
70%
73

NavePortfolioFormationRule:
Example
Example
(4/7)
4. We select the weights inside the RiskyAssets Group
If we dont have a view about the future trends of the Stock
Markets, we should replicate the composition of the World Market.
This solution is named Market Neutral: it is loyal to the Market.
Risky-Assets
y

Equity
q y Markets
Capitalisation

Portfolio Weights
g

EquityMarketEurope

31%

(31%30%)=9.3%

Equity Market North America

EquityMarketNorthAmerica

48%

(48% 30%)=14.4%
(48%
30%) 14 4%

EquityMarketJapan

10%

(10% 30%)=3.0%

Equity Market Pacific ex Japan

EquityMarketPacificexJapan

4%

(4% 30%)
(4%
30%)=1.2%
1 2%

EquityEmergingMarkets

7%

(7% 30%)=2.1%

100%

30%

Sum

No valueadded for investors: we just replicate the market!

74

NavePortfolioFormationRule:
Example
Example
(5/7)
4. We select the weights inside the RiskyAssets Group
But we try to beat the market, so we depict the future:
Europe
p will over p
perform North America
EM will over perform Japan
Pacific ex Japan NEUTRAL

E
Europe
NorthAmerica
Japan
PacificexJapan
Em.Mkts
Sum

Equity Markets
Capitalisation

abanddoned

Risky-Assets

New Group
Weights

Portfolio Weights

31%

40%

(40% 30%) 12 0%
(40%30%)=12.0%

48%

39%

(39% 30%)=11.7%

10%

5%

(5% 30%)=1.5%
(5%
30%)= 1 5%

4%

4%

(4% 30%)=1.2%

7%

12%

((12% 30%)=3.6%
)

100%

30%

75

NavePortfolioFormationRule:
Example
Example
(6/7)
Th final
The
fi l portfolio
tf li
Assets

Portfolio
Weights

MoneyMarketEMU

55.0%

BondMarketEMU

15.0%

EquityMarketEurope

12.0%

EquityMarketNorthAmerica

11.7%

Equity Market Japan

EquityMarketJapan

1 5%
1.5%

EquityMarketPacificexJapan

1.2%

EquityEmergingMarkets
q y
g g

3,6%

Money Market EMU

1,2%

3.6%

Bond Market EMU

1,5%
,
11,7%

Sum

100.0%

Equity Market Europe

12,0%

Equity Market North America

55,0%

Equity Market Japan

15,0%

Equity Market Pacific ex

J
Japan
Equity Emerging Markets

76

NavePortfolioFormationRule:
Example
Example
(7/7)
The final
Th
fi l nave
portfolio:
tf li
is diversified;
has a reasonable composition.
but at its best:
it is a good solution.
it is not the optimal one.
Ifyouwantmore,youneed
MODERNPORTFOLIOTHEORY(MPT) 77

Agenda

4. Strategic
g Asset Allocation: A Q
Quantitative Approach
pp

78

Quantitative Approach:
The Markowitz Model
Harry
y Markowitzs Portfolio
f
selection is the
father of portfolio optimization
..and
d his
hi model
d l (even
(
if it is
i 50 years old)
ld) is
i
widely used in portfolio construction.
N
No
d bt there
doubt,
th
are other
th
mathematical
th
ti l
approach. But no one has the Markowitzs
model
d l aptitude
tit d to
t be:
b
rigorous from a mathematical point of view;
easy to be
b implemented.
i l
d
79

The Markowitz Model: The hypothesises

Given a unique
q time horizon
..investors want to maximise the expected
return
t
(They love
lo e returns).
etu )
Investors are risk adverse ((Theyy hate risk))
The statistical parameter used to measure risk is
the standard
ta da d deviation.
de iatio
One of the measures considered, the semi-standard deviation, produces
efficient portfolios some what preferable to those of the standard deviation.
Those produced by the standard deviation are satisfactory, however, and the
standard deviation itself is easier to use, more familiar to many, and perhaps
easier
i to interpret
i
than
h the
h semi-standard
i
d d deviation.
d i i
(
(Markowitz,
k i 1959).
)
80

TheExpectedReturn StandardDeviation
Principle
Risk is bad variable:
Therefore, investors are willing to increase risk only if higher
risk produces higher return.
E(R)

.
. .
.
A

Solutions B & C are inefficient

Solutions D is efficient: it is an optimal solution for high risk
81
tolerance investors

The Markowitz Model:

A very scheduled process

E(R)

Asset Classes
MKT1
MKT2
MKT3
MKT4
MKT5
MKT6
MKT7
MKT8
MKT9
MKT10
MKT11
MKT12

MKT1
MKT2
MKT3
MKT4
MKT5
MKT6
MKT7
MKT8
MKT9
MKT10
MKT11
MKT12

MKT1
MKT2
MKT3
MKT4
MKT5
MKT6
MKT7
MKT8
MKT9
MKT10
MKT11
MKT12

Optimization
1
1
1
Rendimento

1
1
1
1
1
1
1
1
Rischio

82

The Markowitz Model:

A few remarks
1. 5 stages to be performed.
2 The
2.
Th process is
i timeexpensive:
ti
i
you need
d to
t estimate
ti t many
parameters.
3. For example: with 8 asset classes selected, it is necessary to
estimate:
8 expected return;
8 standard deviation;
28 correlation.
4 Unfortunately asset managers don
4.
dontt like to produce
quantitative and numerical estimation (European equity
market is expected to perform 7.0%).they prefer to
produce qualitative estimation (European equity market
will beat North American equity market ).

5 N
5.
No way, if we want to use the
h Markowitz
M k i model,
d l
83
quantitative estimations are required.

Stage 1: Selection of Asset Classes

Asset Classes
MKT1
MKT2
MKT3
MKT4
MKT5
MKT6
MKT7
MKT8
MKT9
MKT10
MKT11
MKT12

(1/2)

From a theoretical point of view, we should

not narrow the number of investment
opportunities.
Therefore we could select dozens of asset
classes
l
( h may be
(they
b hundreds).
h d d )
But this theoretical position cannot be
performed because of p
p
practical p
problems:
Increase of parameters to be estimated;
Reduction of Asset Under Management
(AUM) for every asset class Increase
of management fees

84

Stage 1: Selection of Asset Classes

Asset Classes
MKT1
MKT2
MKT3
MKT4
MKT5
MKT6
MKT7
MKT8
MKT9
MKT10
MKT11
MKT12

(2/2)

Asset Managers usually select not more than

10 12 asset classes
1012
Marginal players (ex: Japanese Money
Market ) are ignored;
Similar (highly positively correlated)
markets are aggregated.

85

Stage 2: Expected Returns [E(R)]

Exp. Returns
MKT1
MKT2
MKT3
MKT4
MKT5
MKT6
MKT7
MKT8
MKT9
MKT10
MKT11
MKT12

(1/2)

MY SUGGESTION:
Expected Returns shouldnt be the
historical average returns.
Empirical studies say that the Rear
Rearview
view
mirror strategy doesnt work.
Future is different from the past (returns
probability
b bili
di ib i
distribution
are
not
stationary).

Awrongbelief:
Among
g financial p
practitioners is widely
y spread
p
the idea that Harry
y
Markowitz suggested to use historical estimators. This is wrong:
The procedures, I believe, should combine statistical techniques and the
judgment of practical men. [] One suggestion is to use the observed
parameters for some period of the past. I believe that better methods, which
86
take into account more information, can be found (Markowitz, 1952)

Stage 2: Expected Returns [E(R)]

Exp. Returns
MKT1
MKT2
MKT3
MKT4
MKT5
MKT6
MKT7
MKT8
MKT9
MKT10
MKT11
MKT12

(2/2)

Empirical studies suggest that historical

average return are not good predictor of
the future return.
Empirical studies suggest that estimation
error in
i E(R) are deadly.
d dl
Asset Managers must forecast the future,
not trust the p
predictive p
power of the p
past.
Expected returns must be forward looking,
not backward looking.

Statisticaltechniquescanbeuseful:

Macroeconomic models: based on the connection between

future return and macroeconomics factor;
Autoregressive models: based on the study of trend of the
87
historical series of returns.

Stage 3: Standard Deviations ()

MKT1
MKT2
MKT3
MKT4
MKT5
MKT6
MKT7
MKT8
MKT9
MKT10
MKT11
MKT12

(1/2)

Empirical studies suggest that historical

standard deviations are good predictor of the
future standard deviation.
Estimation error in standard deviation are
not deadly.
d dl

MY SUGGESTION:
You can apply the classical
classical rule
rule, using the observed
observed for
some period of the past.
We save time and focus our efforts on Expected Return
prediction.
88

Stage 3: Standard Deviations ()

MKT1
MKT2
MKT3
MKT4
MKT5
MKT6
MKT7
MKT8
MKT9
MKT10
MKT11
MKT12

(2/2)

If you want, you can use more sophisticated

technical models:

Implied volatility;
Econometric
E
i models
d l (ARCH,
(ARCH GARCH).
GARCH)

89

Stage 4: Correlations ()

(1/2)

Empirical studies suggest that historical

correlations are good predictor of the future
correlation.
Estimation error in correlation are not
d dl
deadly.

1
1
1
1
1
1
1
1
1
1
1
1

MY SUGGESTION:
You can apply the classical
classical rule
rule, using the observed
observed for
some period of the past.
90

Stage 5: Correlations ()

(2/2)

If you want, you can use more sophisticated

technical models:

1
1
1
1
1
1
1
1
1
1
1
1

Econometric models (ARCH, GARCH).

91

Final Stage: Optimization

If we have: Asset Classes, E(r), and

We can optimize (Quadratic Programming).

E(R)
Rendimento

Optimization

(1/3)

Risk
Rischio

Fi d the
Find
th weights
i ht (w
( i) able
bl tto:

Objective function

MIN Portfoglio

Constraints:
1st constraint:

Exp.
p Return = E(R)*
( )

2nd constraint:

w1 ++ .. wi + ..wn =1

3rd constraint:
i

wi 0
92

Final Stage: Optimization

Mathematical structure of the Markowitz

optimization:

E(R)
Rendimento

Optimization

(2/3)

Risk
Rischio

2
Mi Port
Min
W

Constraints :
k

wi E ( Ri ) E ( R*)
i 1
n

wi 1
i 1

wi 0 with i 1,..., k
93

Final Stage: Optimization

2
Mi Port
Min
W

Constraints :
k

wi E ( Ri ) E ( R*)
i 1
n

wi 1
i 1

wi 0 with i 1,..., k

(3/3)

We run this optimization for a targeted expected return

[E(R)*]
The optimization returns:
o the portfolio composition
o that is efficient as, given the targeted E(R), it is able
to minimise the standard deviation
Running the optimization for different targeted E(R) we
obtain a range of efficient portfolio

Optimization

E(R))

Rendimentto

Efficient Frontier

Rischio

94

MarkowitzOptimization:
An application
Anapplication
(1/8)
Asset Classes selected:

95

MarkowitzOptimization:
An application
Anapplication
(2/8)
Expected Returns estimated:

96

MarkowitzOptimization:
An application
Anapplication
(3/8)
Standard deviations estimated:

97

MarkowitzOptimization:
An application
Anapplication
(4/8)
Correlations estimated:

98

MarkowitzOptimization:
An application
Anapplication
(5/8)
Output: Efficient Frontier
E(R)

99

MarkowitzOptimization:
An application
Anapplication
(6/8)
Output: Portfolio composition

100

MarkowitzOptimization:
An application
Anapplication
(7/8)
All the other portfolios are inefficient

N P
Nave
Portfolio
f li

101

MarkowitzOptimization:
An application
Anapplication
(8/8)
A better portfolio exists:

102

MarkowitzOptimization:Excel(1/2)
Markowitz Optimization can be easily processed using Excel

=SUM (D2:D8)
=SUMPRODUCT(B2:B8;D2:D8)
=SQRT(MMULT(MMULT(TRANSPOSE(N2:N8);F2:L8);(N2:N8)))
SQRT(MMULT(MMULT(TRANSPOSE(N2 N8) F2 L8) (N2 N8)))

Then click all together:

g
Ctrl+Shift+Enter
103

MarkowitzOptimization:Excel(2/2)

Constraints

104

Weights

Risk

Markowitzversus Nave

Which is your choice to build a SAA?

MARKOWITZ

NAVE

105

Agenda

5. Nave versus Markowitz

106

It gglitters
e s but.
bu .
Markowitz optimization seems to be the best solution.
N
Nevertheless
h l
fi
financial
i l literature
li
h showed
has
h
d that
h this
hi model
d l has
h
some problems:

1 Effi
1.
Efficient
i t portfolios
tf li
are often
ft
unreasonable
bl (Portfolios
(P tf li
highly concentrated and/or big weights to marginal
markets ).
markets)
2. Efficient Portfolios are unstable (small changes in
expected returns can strongly affect the portfolio
composition).
3 Estimations are supposed to be perfect (Asset managers
3.
are clairvoyant! Estimation error doesnt exist).
4 Efficient portfolios are estimation
4.
estimation error maximizers
maximizers
107

Extra-argument 1:
Efficient Portfolios are unstable

108

Extra-argument 1:
Efficient Portfolios are unstable

(1/5)

Asset Management Committee Alfa has the

following estimation

109

Extra-argument 1:
Efficient Portfolios are unstable

(2/5)

Optimal Portfolio are the following:

Only European
E it
Equity

110

Extra-argument 1:
Efficient Portfolios are unstable

(3/5)

Asset Management Committee Beta has the

same expectations. The only difference is the
following:

7.4%
7%

Very homogenous forecast..similar views

about the future trend of the markets, but111

Extra-argument 1:
Efficient Portfolios are unstable

(4/5)

The portfolio composition is the opposite:

Only North American
Equity

112

Extra-argument 1:
Efficient Portfolios are unstable

(5/5)

It is not encouraging/reassuring to realize that

very small changes in expected returns can
strongly affect the portfolio composition.

Question: Can I trust a model that

Q
give importance to basis points?

113

Efficient portfolios are

estimation
estimation error maximizers
maximizers
Because of the estimation error, efficient portfolios are
lik l to
likely
t have
h
very bad
b d performance.
f
Example:
- The Return of Emerg.Mkts Equity is expected to be

g
)
8.0% ((the highest)
- Efficient portfolios with high risk are concentrated
on Emerging Market Equity
Eq it
- Ex-post we discover that the estimation is wrong, as
this market collapses
- The concentration pproduces a blood bath.

114

Efficient portfolios are

estimation
estimation error maximizers
maximizers

Since estimation error is often large,

portfolios selected according to the
Markowitz criterion are likely not more
efficient than a Nave portfolio.

MARKOWITZ

NAVE

115

Agenda

6. Putting
g Markowitz at work

116

PuttingMarkowitzatwork
InordertoputMarkowitzatwork,weneedto
removetheperfectestimationhypothesis
e o e the e fe t e ti atio hy othe i

Asset Manager are not clairvoyant. They make mistakes.

Estimation error must be managed.
managed
It is better to have portfolios with lower expected return,
but with lower exposition to estimation error.
The problem is the portfolio concentration.

We need to modify
fy the model,, in order to
promote a greater diversification

117

Twotechniques
Heuristic Approaches

Bayesian Approaches

They adjust the

Optimization Process

Theyy adjust
j
the inputs
p
estimated (above all
expected returns)

118

Agenda

7. Heuristic techniques

119

TwoHeuristicApproach
Constrained Optimization

ResamplingTM

Easy

Difficult

120

ConstrainedOptimization

It is necessary to add supplementary

constraints to the Markowitz optimization
Find the weights (wi) able to:

Objective function

MIN Portfoglio

Constraints:
1st constraint:

Exp. Return = E(R)*

2nd constraint:

w1 ++
+ + .. wi + ..w
wn =11

3rd constraint:

wi 0

4th constraint:
5th constraint:

wi Ki

>

wi Hi

These Constraints drive a larger diversification 121

ConstrainedOptimization
Example
(1/3)
Asset Classes selected:

Expected Returns estimated:

Standard deviations estimated:

Correlations estimated:

Supplementary constraints

122

ConstrainedOptimization
Example
(2/3)
Output: Constrained Frontier
E(R)

C
Constrained
i d frontier
f
i is
i lower
l
than
h the
h efficient
ffi i frontier
f
i
123

ConstrainedOptimization
Example
(3/3)
Output: Portfolio composition

Di
Diversification
ifi i increases
i
124

Extra-argument 2:
Improving the Constrained
Optimization

125

Extra-argument 2:
Improving the Constrained Optimization

(1/4)

We observed that using traditional

constrained it is hard to build well
diversified portfolios;
May-be,
May be a few of the possible portfolios are
well diversifiedbut others are still very
concentrated;
t t d
In order to well diversify all the possible
portfolios we can use different constraints
called: : Infra
Infra-group
group constraints
126

Extra-argument 2:
Improving the Constrained Optimization

(2/4)

Examples of infra-group constrains:

At best,
best Emerging Market Equity is 18% of
the equity composition (upper bound);
North America Equity Market can not be
less than 25% of the equity composition.

I use together upper & lower bounds,

bounds
so every
y risky
y asset is free to move
inside a reasonable range.
127

Extra-argument 2:
Improving the Constrained Optimization

(3/4)

Example: Input

Asset Classes selected:

Expected Returns estimated:

Standard deviations estimated:

Example of Optimization with infra-group constraints:

Correlations estimated:

Infra group constraints

Infra-group

128

Extra-argument 2:
Improving the Constrained Optimization

(4/4)

Example: Output

All portfolios are well-diversified

well diversified

129

ResamplingTM

(1/)

Resampling is a methodology that force a certain

level of portfolio diversification.
Resampling is based on:
1 The simulation of a large number of statistically
1.
statistically
consistent investment scenarios
2 The simulated E(R),
2.
E(R) and are used as input of a
new Markowitz Optimization.
3. After repeating steps 2. thousands of time the final
portfolios
(Resampled
Portfolios)
have
the
composition of the average efficient portfolio

130

Resampling:Example
Asset Classes selected:

Expected Returns estimated:

(1/3)
Standard deviations estimated:

Correlations estimated:

131

Resampling:Example

(2/3)

Output: Resampled Frontier

0.1

0.09

Expected Retturn

0.08

0.07

0.06

0.05

0.04

0.03

0.02

Unconstrained
Resampled
0

0.05

0.1
0.15
Standard Deviation

0.2

0.25

R
Resampled
l d frontier
f
i is
i lower
l
than
h the
h efficient
ffi i frontier
f
i
132

Resampling:Example

(3/3)

Output: Portfolio composition

Weights
100%

80%

60%

40%

20%

0%
1

12

23

34

45

56

67

78

89

100

Portfolios

Diversification increases
133

Extra-argument 3:
A deeper analysis of ResamplingTM

134

Extra-argument 3:
A deeper analysis of Resampling

(1/7)

In order to process the resampling technique

we need:

Markowitz
Optimization

Simulation
P
Process
135

Extra-argument 3:
A deeper analysis of Resampling

(2/7)

The need to simulate returns:

We know that our expectations can be wrong;
So in order to incorporate uncertainty, we can
run a simulation process that return behaviours
of market returns that are different from our
expectation.
t ti

136

Extra-argument 3:
A deeper analysis of Resampling

(3/7)

Si l ti
Simulation:
A graphical
hi l representation
t ti
L confidence
Low
fid

Hi h confidence
High
fid

E(R)

E(R)
( )

Expectation
Simulation

137

Extra-argument 3:
A deeper analysis of Resampling

(4/7)

Whatt d
Wh
do we need
d iin order
d tto simulate?
i l t ?
Forecasts ( E(R), , )
Confidence on estimations
Random process that is able to make
deviations from the expectation.

Simulation
138

Extra-argument 3:
A deeper
d
analysis
l i off Resampling
R
li

(5/7)

Example:

139

Extra-argument 3:
A deeper analysis of Resampling

(6/7)

If we are able to simulate, we can estimate the

Resampled
p
p
portfolios

140

Extra-argument 3:
A deeper analysis of Resampling
Stage 1: Expectations

E(R)

MSCI
Europe

MSCI
USA

MSCI
Japan

MSCI
EM

7,0%

6,0%

4,5%

8,0%

20 0%
20,0%

21 0%
21,0%

22 8%
22,8%

E(R)
Ca
ap

MSCI EM

0,76

0,76

0,65

MSCI
USA

MSCI
Japan

MSCI
EM

8,0%

6,9%

6,2%

2,3%

MSCI
Japan

0,91

0,54

0,64

MSCI EM

0,90

0,85

0,62

MSCI
USA

7,1%

8,6%

MSCI
Japan
p

MSCI
EM

10,1% 11,6%

O i i
Optimiz.

20,9% 23,0% 23,3% 31,8%

MSCI
USA

MSCI
Japan

MSCI
Europe
MSCI
USA
MSCI
Japan

0,87

0,76

0,77

MSCI EM

0,78

0,87

0,62

Time

E(R)

MSCI
EM

MSCI
Europe

Path Simulations

Cap

MSCI
Europe
p

MSCI
Europe

Time

MSCI
EM

Cap

MSCI
USA

MSCI
Europe
MSCI
USA
MSCI
Japan

E(R)

Simul.

Optimiz.

16,0%
,
21,5%
,
29,2%
,
23,5%
,

Path Simulations

Stage 2+3000:

0,65

MSCI
Europe

MSCI
Europe
Time

3kst

0,60

MSCI
EM

MSCI
USA

7,6% 10,6%

MSCI
Japan

MSCI
EM

6,0%

11,0%

23 7% 25,4%
23,7%
25 4% 21,6%
21 6% 33,3%
33 3%

E(R)

MSCI
Europe

MSCI
USA

MSCI
Japan

MSCI
EM

2,7%

2,0%

4,6%

4,3%

Optimiz
Optimiz.

21,5% 21,2% 20,3% 36,5%

Average composition

0.09

100%

0.08

80%
0.07

0.06

60%

E(R)

Simul.

Stage 2+2: 2st Simul.

MSCI
Japan

0,85

29 0%
29,0%

Path Simulations

Stage
g 2+1:

MSCI
USA

Very low-Low-Medium-High-Very high

Stage 2: Confidence
1st

MSCI
Europe
MSCI
Europe
MSCI
USA
MSCI
Japan
p

(7/7)

0.05

141

40%
0.04

0.03

0.02
0

20%
0.05

0.1

0.15
0.2
Sigma

0.25

0.3

0.35

0%
1

12

23

34

45

56

Portfolios

67

78

89

10

Agenda

8. Bayesian Techniques

142

Bayesian Techniques

The most common and widely used

Bayesian technique is:
Th Bl
The
Black-Litterman
k Litt
Model
M d l

143

The Black-Litterman Model

The Black
Black-Litterman
Litterman Model creates
better expected
p
return forecasts to be
used with
the Markowitz optimization

144

How does it work?

Start with the
Expected Returns.

Market

Neutral

Apply your views of how certain

markets are going to behave.
behave
The end result is a set of return
forecasts that give rise to
diversified portfolios when used
with the Markowitz Optimization.
145

Portfolio Market Neutral

The Market Neutral Portfolio is the

capitalization-weighted portfolio of
the assets.
The Market Neutral expected
Returns are the expected returns
th t are implied
that
i li d by
b the
th Market
M k t
Neutral Portfolio.

146

Add your own views

Investors generally have opinions,
opinions or
views, about how certain markets
will behave in the future.
Each view includes a measure of
certainty.

147

Example of an Absolute View

Opinion: II think that European Equity
Market is going to do well.
View: European Equity Market will
have a return of 11%
Confidence of View: 55%

148

Example of an Relative View

Opinion: II believe that Europe is
going to outperform Japan.
View: European Equity Market will
outperform Jananese Equity Market
by 3%.
Confidence of View: 80%

149

Combine the Market Returns with

your Views
Market Neutral
Expected Returns

Views

Merging
Black-Litterman
Bl
k Litt
Expected Returns

150

Final Result
Thanks to the Black-Litterman Forecast
Return efficient portfolio are much
Return,
more diversified..
and the composition reflects our views.
Peso
100%

80%

60%

40%

20%

0%
1

12

23

34

45 56
Portafogli

67

78

89

100

151

Market Neutral Expected Returns:

example
l
Asset Class

Rend. Market Neutral

MKT1
MKT2
MKT3
MKT4
MKT5
MKT6
MKT7
MKT8
MKT9
MKT10
MKT11
MKT12

MKT1
MKT2
MKT3
MKT4
MKT5
MKT6
MKT7
MKT8
MKT9
MKT10
MKT11
MKT12

6,41%
,
7,12%
8,67%
7,32%
,
8,95%

Ottimizzazione
1
1
Azionario Pac ex Japan
6% 4%
12%

Azionario Europa
28%

Azionario America
Azionario Giappone

50%

Azionario EM

1
1
1
1
1
1
1
1

1152
1

Market Neutral Expected Returns:

example
l
Azionario Pac ex Japan
Azionario Europa
Azionario America
Azionario Giappone
Azionario EM

Pesi market neutral

4,0000%
28 0000%
28,0000%
50,0000%
12,0000%
6,0000%

Matrice Varianze-Covarianze
0,0408
0,0266
0,0215
0 0266
0,0266
0 0278
0,0278
0 0303
0,0303
0,0215
0,0303
0,0475
0,0171
0,0227
0,0321
0,0329
0,0345
0,0443

0,0171
0,0227
0
0227
0,0321
0,0413
0,0357

0,0329
0,0345
0
0345
0,0443
0,0357
0,0690

Misura di avvversione al rischio

1,4310

Rf

Matrice Varianze-Covarianze
0,0408
0,0266
0,0215
0,0266
0,0278
0,0303
0,0215
0,0303
0,0475
0,0171
0,0227
0,0321
0,0329
0,0345
0,0443

0,0171
0,0227
0,0321
0,0413
0,0357

0,0329
0,0345
0,0443
0,0357
0,0690

Pesi market neutral

4,0000%
28,0000%
50,0000%
12,0000%
6,0000%

Rend. Market Neutral

Azionario Pacifico ex Giappone

Azionario Europa
Azionario America
Azionario Giappone
Azionario EM

6,41%
,
7,12%
8,67%
7,32%
,
8,95%

153

Black & Litterman: the views (1/4)

The
Th mostt interesting
i t
ti feature
f t
off the
th B&L model
d l is
i that
th t
this model does not impose that the Asset Managers
produce detailed estimates for all the markets
involved.
In practical terms,
terms the Asset Managers might simply
express estimates for only a few of the markets under
observation (e.g.
(e g Those which they know better).
better)
Estimates might be either absolute or relative:

It is mandatory that every estimate is accompanied by a

percentage representing the degree of confidence vis-a-vis
the estimates (either in % or in relative terms)
154

Black & Litterman: views

(segue)

Wi
Withh the
h purpose off determining
d
i i the
h final
fi l returns, it
i is
i necessary
to build a matrix (P) so as to make possible to identify which are
the asset classes involved by the views.
views

1
Azionario
Pacifico ex
Giappone

p1
p2

0
0

Azionario
Europa

Azionario
America

Azionario
Giappone

Azionario EM

0
1

-1
1
0

1
0

0
0

P
155

Black & Litterman: views

(segue)

A vector
t column
l
(Q),
(Q) on the
th contrary,
t
identifies
id tifi the
th returns
t
which characterise either the absolute and/or the relative views.

4%
9%

156

Black & Litterman: views

22% c1
22% c2

(segue)

157

Black & Litterman: le views

(segue)

Th
The last
l t input
i t is
i identified
id tifi d by
b a matrix
t i which
hi h represents
t
the confidence of the analysts (Asset Managers) have in
their views.
How to build this matrix is one of the most widely debated
issues.
We determine the matrix with the following method:
1

1
p1 p1T

c1

1 p 2 p T2
c2
0

1 p K
c
K

Elements ci of vector C

T
p K

158

Synthesis

BL ()

P
T

()
1

PT 1 Q

159