EDHEC Institutional Days

Monaco, December 8th, 2010, 14:15-15:30

Alternatives to Cap-Weighted Indices

Lionel Martellini

Professor of Finance, EDHEC Business School

Scientific Director, EDHEC Risk Institute

lionel.martellini@edhec.edu
www.edhec-risk.com

2
 Outline

„ Introduction: Beyond Cap-Weighting

„ In Search of Representative Indices

– Cap-Weighting
– Fundamental Weights

„ Designing Efficient Investment Benchmarks

– Ad-Hoc Diversification: De-concentrating Portfolios
– Scientific Diversification: Towards the Efficient Frontier

„ Alternative Weighting Schemes: Conditions for Optimality?

„ Conclusion: Concept Selection vs. Concept Diversification

3
„ Introduction: Beyond Cap-Weighting

„ In Search of Representative Indices

– Cap-Weighting
– Fundamental Weights

„ Designing Efficient Investment Benchmarks

– Ad-Hoc Diversification: De-concentrating Portfolios
– Scientific Diversification: Towards the Efficient Frontier

„ Alternative Weighting Schemes: Conditions for Optimality?

„ Conclusion: Concept Selection vs. Concept Diversification

4
 Beyond Cap Weighting
Comparing Alternatives

„ A number of (index or fund) providers have recently designed

and launched non cap-weighted indices.

„ The (non-exhaustive) list includes:

– fundamental indices
– equally-weighted indices
– minimum variance indices
– efficient indices
– equal-risk contribution (a.k.a. risk parity) indices
– maximum diversification indices

„ This presentation provides a summary of the objectives of, and

assumptions behind, the various indexing concepts.
5
 Beyond Cap Weighting
Concepts versus Figures

„ We will not focus so much on past performance; track records (by

definition) all look pretty good!

„ Instead, we propose to provide an academic perspective on the

conceptual assumptions underpinning the different methods.
– (Even out-of-sample) track records are sample-dependent and thus
performance figures rely on the data and time period at hand.
– For long-term benchmarks, it is important that performance is driven
by a sound concept that relies on reasonable assumptions rather than
by exploiting anomalies in past returns data.
– If achieving higher risk-adjusted performance is not the focus of a
methodology, achieving it is at best a collateral benefit.

In Seneca’s words (circa 30 BC):

“If one does not know to which port one is sailing, no wind is favorable.”6
 Beyond Cap-Weighting
Which Port do we want to Sail to: Indices versus Benchmarks

„ The words « index » and « benchmark » are often used

interchangeably; yet they define a priori very different concepts.

„ Market perspective: an index is a portfolio that should

represent the performance of a given segment of the market.
=> focus on representativity

„ Investor perspective: a benchmark is a reference portfolio that

should represent the fair reward expected in exchange for risk
exposures that an investor is willing to accept.
=> focus on efficiency

„ CW portfolios have long been portrayed as representative and

efficient, but have faced increased criticism on both fronts.
7
„ Introduction: Beyond Cap-Weighting

„ In Search of Representative Indices

– Cap-Weighting
– Fundamental Weights

„ Designing Efficient Investment Benchmarks

– Ad-Hoc Diversification: De-concentrating Portfolios
– Scientific Diversification: Towards the Efficient Frontier

„ Alternative Weighting Schemes: Conditions for Optimality?

„ Conclusion: Concept Selection vs. Concept Diversification

8
 In Search of Representative Indices
Cap-Weighting for Representativity?

„ A market cap weighted scheme is the obvious default option

when it comes to representing a given segment of the market.
– Market cap weighted indices provide by construction a fair
representation of the stock market;
– In the end, cap-weighting is nothing but an ad-hoc weighting
scheme that achieves some form of representativity.

„ Cap-weighted indices, however, may not provide a fair

representation of the underlying economic fundamentals.

„ Some have argued that they represent well the stock market but
not the economy.

9
 In Search of Representative Indices
Fundamental Weighting for Representativity?

„ Rather than using the market cap, fundamental indices use

firm attributes such as book value, dividends, sales or cash
flows as measures of size.

„ These indices aim at better representing the economy.

Arnott (2007): “The Fundamental Index weights companies in
accordance to their footprint in the broad economy […] you wind up
with a portfolio that mirrors the economy”.

„ Whether or not fundamentally weighted indices better

represent the economy is actually an open question, if only
because representativity is not a concept that is linked to clear
measures.

10
„ Introduction: Beyond Cap-Weighting

„ In Search of Representative Indices

– Cap-Weighting
– Fundamental Weights

„ Designing Efficient Investment Benchmarks

– Ad-Hoc Diversification: De-concentrating Portfolios
– Scientific Diversification: Towards the Efficient Frontier

„ Alternative Weighting Schemes: Conditions for Optimality?

„ Conclusion: Concept Selection vs. Concept Diversification

11
 Designing Efficient Investment Benchmarks
Efficiency is Related to Diversification

„ In any case, it is not clear why investors would care about their
portfolios representing the economy.

„ From the investor’s perspective, the focus should be on

efficiency: obtaining fair rewards for given risk budgets.

„ Efficiency is intimately related to diversification: it is by

constructing well-diversified portfolios that one can achieve a
fair reward for a given risk exposure.

„ CW portfolios in fact appear to be rather inefficient and poorly

diversified portfolios, and several approaches have been
developed so as to improve diversification compared to cap-
weighting.

12
 Designing Efficient Investment Benchmarks
Cap-Weighting for Efficiency?

„ Cap-weighting is often believed to lead to risk/reward efficient

portfolios, but that belief is not really based on firm grounds.

„ The belief in efficiency of CW is based on a naïve interpretation

of W. Sharpe’s Capital Asset Pricing Model (CAPM):
– No need to gather any information on risk & return parameters
to find optimal portfolios ... because everybody else does!
– When relaxing the highly unrealistic assumptions of the CAPM,
financial theory does not predict that the market portfolio is efficient
(Sharpe (1991), Markowitz (2005)).
– If there are multiple risk factors, the mean-variance optimal portfolio
is no longer CW (Merton (1971), Cochrane (1999)); in a post-CAPM
multi-factor world, CW is just an arbitrary weighting scheme.

13
 Designing Efficient Investment Benchmarks
CW leads to High Concentration

„ Cap-weighting is particularly inefficient because it leads to high

concentration: the effective number of stocks in the index is low.

Index Nominal Effective

number number
S&P 500 94
NASDAQ 100 37
FTSE 100 (UK) 100 28
The effective number of
stocks is the reciprocal of the FTSE Eurobloc 300 104
Herfindhal index, a measure FTSE Japan 500 103
of portfolio concentration.
Average effective number based on quarterly assessment for the time
period 01/1959 to 12/2008 for the S&P, 01/1975 to 12/2008 for the
NASDAQ, and 12/2002 to 12/2008 for the other indices .

„ Some index construction approaches simply avoid this

concentration; such simple “de-concentration” strategies do not
aim for “optimality” and are not grounded in portfolio theory. 14
„ Introduction: Beyond Cap-Weighting

„ In Search of Representative Indices

– Cap-Weighting
– Fundamental Weights

„ Designing Efficient Investment Benchmarks

– Ad-Hoc Diversification: De-concentrating Portfolios
– Scientific Diversification: Towards the Efficient Frontier

„ Alternative Weighting Schemes: Conditions for Optimality?

„ Conclusion: Concept Selection vs. Concept Diversification

15
 Ad-Hoc Approach to Well-Diversified Portfolios
Equal Weighting and Equal Risk Contribution

„ Naïve de-concentration:
– Equal-weighting simply gives the same weight to each of N stocks
in the index (“1/N rule”).
– Equal-weighting is the naïve route to constructing well diversified
portfolios.

„ Semi-naïve de-concentration:
– Equal risk contribution (ERC) takes into account contribution to risk.
– Contribution to risk is not proportional to dollar contribution.
– Find portfolio weights such that contributions to risk are equal
(Maillard, Roncalli and Teiletche (2010)):
∂σ p ∂σ p
wi = wj
∂wi ∂w j
16 16
 Ad-Hoc Approach to Well-Diversified Portfolios
Maximum Diversification Benchmarks/Anti-Benchmark

„ Statistical de-concentration:
– Define a diversification index and try and maximize it by utilizing the
correlations that drive the magic of diversification: “The whole is
better than the sum of its parts”.
– Maximum Diversification (also known as anti-benchmark) aims at
generating portfolios with the highest possible diversification index
(Choueifaty and Coignard (2008)):
⎛ n ⎞
⎜ ⎟
⎜ ∑ w σ
i i ⎟
DI = Max ⎜ i =1 ⎟
w ⎜ n ⎟
⎜
⎜
∑ wi w j σ ij ⎟
⎟
⎝ i , j =1 ⎠

The weighted average risk (in the numerator) will be high compared
to portfolio risk (in the denominator) and thus DI will be high if the
portfolio weights exploit well the correlations.
17
„ Introduction: Beyond Cap-Weighting

„ In Search of Representative Indices

– Cap-Weighting
– Fundamental Weights

„ Designing Efficient Investment Benchmarks

– Ad-Hoc Diversification: De-concentrating Portfolios
– Scientific Diversification: Towards the Efficient Frontier

„ Alternative Weighting Schemes: Conditions for Optimality?

„ Conclusion: Concept Selection vs. Concept Diversification

18
 Scientific Approach to Well-Diversified Portfolios
Towards the Efficient Frontier

ƒ Scientific diversification is based on reaching a high risk/return

objective through portfolio construction techniques.

ƒ In practice, to get a decent proxy for efficient portfolios, one needs

to use careful risk and return parameter estimates; practical
approaches to scientific diversification make different choices
regarding the challenge of risk and return estimation.

ƒ Technology is available to generate reliable risk parameter

estimates:
– Suitably designed factor models to mitigate the curse of
dimensionality (see also statistical shrinkage techniques).
– Accounting for non-stationarity: e.g., GARCH and Regime Switching
models.

ƒ On the other hand, statistics is close to useless in terms of

expected return estimation (Merton (1980)).
19 19
 Scientific Approach to Well-Diversified Portfolios
GMV vs. MSR

Expected Maximum Sharpe

Return Ratio (MSR) Portfolio
Global
Minimum
Variance
(GMV)
Portfolio ●
●

Volatility

ƒ The MSR provides the highest reward per unit of portfolio volatility:
needed optimization inputs are expected returns, correlations and
volatilities.
ƒ The GMV provides the lowest possible portfolio volatility: needed
optimization inputs are correlations and volatilities. 20
 Scientific Approach to Well-Diversified Portfolios
Minimum Variance Benchmarks (GMV)

ƒ If you feel comfortable about estimating risk parameters the variance-

covariance matrix, but not about estimating expected return
parameters, the global minimum variance (GMV) benchmark is the
way to go (e.g., Amenc and Martellini (2003)).

ƒ This approach provides a low 18

volatility portfolio but also a low 16
Efficient frontier
Tangency line
performance portfolio: ex-ante, 14
MSR

Annualized expected return

MSR+cash is better than GMV. 12
MSR + cash
10

ƒ Ex-post, MV portfolios tend to be 8

concentrated portfolios with 6

overweighting of low volatility 4

GMV
stocks, with a Sharpe ratio lower 2

than that of EW (Garlappi et al. 0

0 5 10 15 20 25
(2007)). Annualized volatility

21 21
 Scientific Approach to Well-Diversified Portfolios
Efficient Indexation (MSR)

ƒ Efficient Indexation is about maximizing the Sharpe ratio.

ƒ Just like in the Minimum Variance approach, Efficient

Indexation exploits information on the covariance matrix of
stock returns; the approach uses suitably designed factor
models to mitigate the curse of dimensionality.

ƒ While direct estimation of expected returns from past returns is

useless, all hope on expected returns estimation is not lost!

ƒ Common sense suggests that expected return parameters

should be positively related to risk parameters (risk-return
tradeoff ).

ƒ Efficient Indexation uses indirect estimation of expected returns

through a stock’s riskiness.
22
 Scientific Approach to Well-Diversified Portfolios
On the Risk-Return Relationship

„ Theory unambiguously confirms the existence of a positive

risk/return relationship:
– Systematic risk is rewarded (APT);
– Specific risk is also rewarded (Merton (1987)) (*);
– Total volatility (model-free) should therefore be rewarded;
– Higher moment risk is also rewarded (many references).

„ Use the risk-return relationship to build efficient portfolios: magic

of diversification is about mixing high-risk-and-therefore-high-
return stocks in a smart way so as to generate low risk portfolios!

23
(*) See also Barberis and Huang (2001) Malkiel and Yu (2002), Boyle, Garlappi, Uppal and Wang (2009) .
Scientific Approach to Well-Diversified Portfolios
iv Puzzle – VW Portfolios over Short Horizons
Value Weighted Portfolios: Short Horizon (iVol)
25.0%
•Ang, Hodrick, Xing and Zhang (2006,
2009): “iv puzzle” 20.0%

Average Portfolio Return and

•12 Month idiosyncratic volatility

Standard Error Bounds

15.0%
•1 Month realized return
10.0%
•10 VW Portfolios
•Value Weighted Portfolio returns 5.0%
•Negative Relationship
0.0%
•High-Low returns mainly driven by high 0% 5% 10% 15% 20% 25% 30%
iVol portfolio -5.0%

-10.0%
Average Risk over Cross-Section
Value Weighted Portfolios: Short Horizon (iVol)
1000.00

100.00
Value of 1$ invested in 1964

10.00 Ten VW portfolios containing an equal number of

stocks (extracted from the CRSP data base) are built
1.00 every month after sorting the stocks based on some
64' 67' 70' 73' 76' 79' 82' 85' 88' 91' 94' 97' 00' 03' 06' 09' risk measure, here idiosyncratic volatility w.r.t. FF
0.10
model (calculated using daily data for last 12
months); the returns of each of these portfolios are
0.01
calculated subsequent one-month periods and
0.00 averaged across the portfolio formation date.

Low 2 3 4 5 6 7 8 9 High
 Scientific Approach to Well-Diversified Portfolios
No iv Puzzle – EW Portfolios over Short Horizons
Equally Weighted Portfolios: Short Horizon (iVol)
25.0%
• Return reversal : Huang, Liu, Rhee, and Zhang (2009)
• Extreme winners and losers (over the past month) 20.0%

Average Portfolio Return and

typically have high iVol over the last 1 month

Standard Error Bounds

15.0%
• In high iVol portfolios: # past winners is almost equal
to # past losers, but average weight of past winners is 10.0%
substantially larger.
• Short-term return reversal effect: past-month winners 5.0%

tend to under perform in subsequent month . 0.0%

• So, VW lowers the portfolio return compared to other 0% 5% 10% 15% 20% 25% 30%
portfolios and EW does not. -5.0%

-10.0%
Average Risk over Cross-Section
Equally Weighted Portfolios: Short Horizon (iVol)
10000.00
Value of 1$ invested in 1964

1000.00

•Negative relationship disappears when

100.00
EW used.
10.00
•Extremely low return of High-Volatility
portfolio disappears.
1.00 •We still do not have a positive relationship.
64' 67' 70' 73' 76' 79' 82' 85' 88' 91' 94' 97' 00' 03' 06' 09'

0.10

Low 2 3 4 5 6 7 8 9 High
 Scientific Approach to Well-Diversified Portfolios
What iv Puzzle ? – EW Portfolios over Long Horizons
Equally Weighted Portfolios: Long Horizon (iVol)
30.0%
•Positive risk-return relationship across all 25.0%

Average Portfolio Return and

portfolios.

Standard Error Bounds

20.0%
•Not only the extreme portfolios.
15.0%
•Intuition: long-horizon realized returns are
less susceptible to local events and hence 10.0%

better proxies for expected returns. 5.0%

compared to short-horizon realized returns. 0.0%

0% 5% 10% 15% 20% 25%
-5.0%

-10.0%
Average Risk over Cross-Section
Equally Weighted Portfolios: Long Horizon (iVol)
10000.00
Value of 1$ invested in 1964

1000.00

Ten EW portfolios containing an equal number of

100.00 stocks (extracted from the CRSP data base) are built
every month after sorting the stocks based on some
10.00 risk measure, here idiosyncratic volatility w.r.t. FF
model (calculated using daily data for last 12
1.00 months); the returns of each of these portfolios are
64' 67' 70' 73' 76' 79' 82' 85' 88' 91' 94' 97' 00' 03' 06' 09' calculated subsequent 24-months periods and
0.10
averaged across the portfolio formation date.

Low 2 3 4 5 6 7 8 9 High
 Scientific Approach to Well-Diversified Portfolios
tv Puzzle – VW Portfolios over Short Horizons
Value Weighted Portfolios: Short Horizon (tVol)
25.0%

• Blitz and Vliet (2007) 20.0%

Average Portfolio Return and

•12 Month total volatility

Standard Error Bounds

15.0%
•1 Month realized return
10.0%
•10 Portfolios
•Value Weighted Portfolio returns 5.0%
•Negative risk-return relationship
0.0%
• High-Low returns mainly driven by 0% 5% 10% 15% 20% 25% 30%
low tVol portfolio -5.0%

-10.0%
Average Risk over Cross-Section
Value Weighted Portfolios: Short Horizon (tVol)
1000.00
Value of 1$ invested in 1964

100.00

Ten VW portfolios containing an equal number of

10.00 stocks (extracted from the CRSP data base) are built
every month after sorting the stocks based on some
1.00 risk measure, here total volatility (calculated using
64' 67' 70' 73' 76' 79' 82' 85' 88' 91' 94' 97' 00' 03' 06' 09' daily data for last 12 months); the returns of each of
0.10 these portfolios are calculated subsequent one-
month periods and averaged across the portfolio
0.01 formation date.

Low 2 3 4 5 6 7 8 9 High
 Scientific Approach to Well-Diversified Portfolios
No tv Puzzle – EW Portfolios over Short Horizons
Equally Weighted Portfolios: Short Horizon (tVol)
25.0%

•12 Month total volatility 20.0%

Average Portfolio Return and

Standard Error Bounds
•1 Month realized return 15.0%
•10 Portfolios
10.0%
•Equally Weighted Portfolio returns
5.0%

0.0%
0% 5% 10% 15% 20% 25% 30%
-5.0%

-10.0%
Average Risk over Cross-Section
Equally Weighted Portfolios: Short Horizon (tVol)
1000.00
Value of 1$ invested in 1964

100.00
•Again, Negative relationship disappears
when EW used.
10.00
• Extremely low return of High-Volatility
portfolio disappears.
1.00 •We still do not have a positive relationship.
64' 67' 70' 73' 76' 79' 82' 85' 88' 91' 94' 97' 00' 03' 06' 09'

0.10

Low 2 3 4 5 6 7 8 9 High
 Scientific Approach to Well-Diversified Portfolios
What tv Puzzle? – EW Portfolios over Long Horizons
Equally Weighted Portfolios: Long Horizon (tVol)
•12 Month total volatility 30.0%

•24 Month realized return 25.0%

Average Portfolio Return and

•10 EW Portfolios

Standard Error Bounds

20.0%
•Positive risk-return relationship 15.0%
across all portfolios.
10.0%
•Not only the extreme portfolios.
5.0%
•Results are valid even tVol is
calculated using larger period. 0.0%
0% 5% 10% 15% 20% 25%
-5.0%

-10.0%
Average Risk over Cross-Section
Equally Weighted Portfolios: Long Horizon (tVol)
10000.00
Value of 1$ invested in 1964

1000.00

Ten EW portfolios containing an equal number of

100.00 stocks (extracted from the CRSP data base) are built
every month after sorting the stocks based on some
10.00 risk measure, here total volatility (calculated using
daily data for last 12 months); the returns of each of
1.00 these portfolios are calculated subsequent 24-month
64' 67' 70' 73' 76' 79' 82' 85' 88' 91' 94' 97' 00' 03' 06' 09' periods and averaged across the portfolio formation
0.10
date.

Low 2 3 4 5 6 7 8 9 High
Scientific Approach to Well-Diversified Portfolios
Downside Risk & Expected Returns
Evidence that stock downside risk is related to expected returns

Authors Risk Measure Moments

Zhang (2005) Skewness Skew
Boyer, Mitton and Skewness Skew
Vorkink (2009)
Tang and Shum (2003) Skewness Skew

Connrad, Dittmar and Skewness Skew

Ghysels (2009)
Ang et al. (2006) Downside correlation Vol, Skew, Kurt
Huang et al (2009) Value-at-Risk (EVT) Vol, Skew, Kurt
Bali and Cakici (2004) Value-at-Risk Vol,Skew, Kurt
(Historical)
Chen et al. (2009) Semi-deviation Vol, Skew, Kurt
Estrada (2000) Semi-deviation Vol, Skew, Kurt
 Scientific Approach to Well-Diversified Portfolios
Total Semi-Deviation – EW Decile Portfolios Long Horizon
Equally Weighted Portfolios: Long Horizon (sem i-deviation)
30.0%

25.0%
•12 Month Total Semi-Deviation

Average Portfolio Return and

Standard Error Bounds
20.0%
•24 Month realized return
•10 Portfolios 15.0%

•Equally Weighted Portfolio returns 10.0%

5.0%

0.0%
0% 5% 10% 15% 20%
-5.0%

-10.0%
Average Risk over Cross-Section
Equally Weighted Portfolios: Long Horizon (sem i-deviation)
10000.00
Value of 1$ invested in 1964

1000.00 •Positive risk-return relationship

across all portfolios.
100.00 •Not only the extreme portfolios.
•Results are valid even semi-
10.00
deviation is calculated using larger
period.
1.00
64' 67' 70' 73' 76' 79' 82' 85' 88' 91' 94' 97' 00' 03' 06' 09'

Low 2 3 4 5 6 7 8 9 High
Scientific Approach to Well-Diversified Portfolios
Downside Risk and Expected Returns

The average cumulative return for portfolios sorted on semi-deviation.

80%
Port Low

70% Port 2
Port 3
60% Port 4
Port 5
50% Port 6
Port 7
40% Port 8
Port 9
30% Port High

20%

10%

0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Month after portfolio formation

Ten portfolios containing an equal number of stocks (extracted from the CRSP data base) are built every month after sorting the stocks
based on their semi-deviation (calculated using daily data for last 30 months); the cumulative returns of each of these portfolios are
calculated for various holding periods and averaged across the portfolio formation date.

32
Scientific Approach to Well-Diversified Portfolios
Long-Term Results

Ann.
Ann. std. Sharpe Information Tracking
Index average
Deviation Ratio Ratio Error
return
Efficient Index 11.63% 14.65% 0.41 0.52 4.65%
Cap-weighted 9.23% 15.20% 0.24 0.00 0.00%
Difference (Efficient
2.40% -0.55% 0.17 - -
minus Cap-weighted)
p-value for difference 0.14% 6.04% 0.04% - -
The table shows risk and return statistics portfolios constructed with using the same set of constituents as the cap-weighted S&P 500 index.
Rebalancing is quarterly subject to an optimal control of portfolio turnover (by setting the reoptimisation threshold to 50%). Portfolios are
constructed by maximising the Sharpe ratio given an expected return estimate and a covariance estimate. The expected return estimate is
set to the median total risk of stocks in the same decile when sorting on total risk. The covariance matrix is estimated using an implicit factor
model for stock returns. Weight constraints are set so that each stock's weight is between 1/2N and 2/N, where N is the number of index
constituents. P-values for differences are computed using the paired t-test for the average, the F-test for volatility, and a Jobson-Korkie test
for the Sharpe ratio. The results are based on weekly return data from 01/1959. We use a calibration period of 2 years and rebalance the
portfolio every three months (at the beginning of January, April, July and October).

33
„ Introduction: Beyond Cap-Weighting

„ In Search of Representative Indices

– Cap-Weighting
– Fundamental Weights

„ Designing Efficient Investment Benchmarks

– Ad-Hoc Diversification: De-concentrating Portfolios
– Scientific Diversification: Towards the Efficient Frontier

„ Alternative Weighting Schemes: Conditions for Optimality?

„ Conclusion: Concept Selection vs. Concept Diversification

34
 Conditions for Optimality
Keep it Simple… But Not Too Simple
ƒ Each of the aforementioned weighting methods makes different
methodological choices.

ƒ However, portfolio theory tells us that there is only one optimal

portfolio: the tangency (MSR) portfolio.

Question: Under which conditions would the portfolio construction

choices of different index weighting schemes be truly optimal?

KIS(BNTS) principle: robustness of a method may justify simple

assumptions but is important that assumptions also remain
reasonable; if the conditions are too restrictive, we are unlikely to
obtain optimal portfolios.

35
 Conditions for Optimality
Assumptions about Parameter Values

● wMSR = f (μi ,σ i , ρij )

The true tangency portfolio is a function of the
(unknown) true parameter values

Expected Optimal
Return Portfolio

●
●●
●
● ●
Cap-weighted index

Volatility
● wˆ MSR = f (μˆ i , σˆ i , ρˆ ij )
Implementable proxies depend on
assumptions about parameter values

36
 Conditions for Optimality
Indices aiming at Representativity

ƒ Cap-weighting:
– One simply turns to the market, and hope that everyone else has
done a careful job at estimating risk and return parameters and
designing efficient benchmarks so we simply do not have to it
ourselves!
– This would be a very naïve belief in the CAPM.

ƒ Fundamental weighting:
– Conditions under which this weighting scheme would be optimal
are not clear.
– As an example, it would be optimal if risk parameters are
identical and expected return is proportional to the four
fundamental variables used for the weighting.

37
 Conditions for Optimality
De-Concentration Approaches

ƒ Equal-weighting:
– Optimal if and only if one assumes all stocks have the same
expected return and…
– … the same volatility and…
– … the same pairwise correlations!

ƒ Equal Risk Contribution (Maillard et al. (2010)):

– Optimal if and only if one assumes all stocks have same
Sharpe ratios and…
– … the same pairwise correlations.

ƒ Maximum Diversification (Choueifaty and Coignard (2008)):

– Optimal if and only if one assumes all stocks have same
Sharpe ratios.
38
 Conditions for Optimality
Efficient Frontier Approaches

ƒ Minimum Variance:
– Only optimal if one assumes that all stocks have the same
expected returns, hardly a neutral/reasonable choice.

ƒ Efficient Indexation:
– Optimal if one assumes that expected returns between stocks
are different, and positively related to downside risk.

39
„ Introduction: Beyond Cap-Weighting

„ In Search of Representative Indices

– Cap-Weighting
– Fundamental Weights

„ Designing Efficient Investment Benchmarks

– Ad-Hoc Diversification: De-concentrating Portfolios
– Scientific Diversification: Towards the Efficient Frontier

„ Alternative Weighting Schemes: Conditions for Optimality?

„ Conclusion: Concept Selection vs. Concept Diversification

40
 Conclusion

„ Cap-weighted indices are not efficient or well-diversified

portfolios because they were never meant to be.
„ While alternative weighting schemes typically improve
performance, they have different objectives and more or less
strong assumptions need to be made before one can
conclude that they are truly optimal portfolios.
„ Investors – beyond assessing performance – need to
consider whether assumptions and objectives behind each
concept are compatible with their views and needs.

„ An outstanding question, which we do not address in this

presentation, is that of concept diversification versus concept
selection.

41
 References

ƒ Amenc, N., F. Goltz, L. Martellini, and P. Retkowsky, 2010, Efficient Indexation: An Alternative to
Cap-Weighted Indices," Journal of Investment Management, forthcoming.
ƒ Bali, Turan G., and Nusret Cakici, 2004, Value at Risk and Expected Stock Returns. Financial
Analysts Journal, 60(2), 57-73.
ƒ Barberis, N., and M. Huang, 2001, Mental Accounting, Loss Aversion and Individual Stock Returns,
Journal of Finance, 56, 1247-1292.
ƒ Barberis, N. and M. Huang, Stocks as lotteries: The implications of probability weighting for
security prices, 2007, working paper.
ƒ Boyer, B., and K. Vorkink, 2007, Equilibrium Underdiversification and the Preference for Skewness,
Review of Financial Studies, 20(4), 1255-1288.
ƒ Boyer, B., T. Mitton and K. Vorkink, 2009, Expected Idiosyncratic Skewness, Review of Financial
Studies, forthcoming.
ƒ Chen, D.H., C.D. Chen, and J. Chen, 2009, Downside risk measures and equity returns in the
NYSE, Applied Economics, 41, 1055-1070.
ƒ Connrad, J., R.F. Dittmar and E. Ghysels, Ex Ante Skewness and Expected Stock Returns, 2008,
working paper.
ƒ Choueifaty, Y., and Y. Coignard, 2008, Toward Maximum Diversification, The Journal of Portfolio
Management, 35, 1, 40-51.
ƒ Cochrane, John H., 2005, Asset Pricing (Revised), Princeton University Press
ƒ Estrada, J, 2000, The Cost of Equity in Emerging Markets: A Downside Risk Approach, Emerging
Markets Quarterly, 19-30.
ƒ Grinold, Richard C. “Are Benchmark Portfolios Efficient?”, Journal of Portfolio Management, Fall
1992.
42
 References

ƒ Haugen, R. A., and Baker N. L., “The Efficient Market Inefficiency of Capitalization-weighted Stock
Portfolios”, Journal of Portfolio Management, Spring 1991.
ƒ Malkiel, B., and Y. Xu, 2002, Idiosyncratic Risk and Security Returns, working Paper, University of
Texas at Dallas.
ƒ Maillard,, S., T. Roncalli and J. Teiletche, 2010, The Properties of Equally Weighted Risk
Contribution, Journal of Portfolio Management.
ƒ Markowitz, H. M., “Market efficiency: A Theoretical Distinction and So What?”, Financial Analysts
Journal, September/October 2005.
ƒ Merton, Robert, 1987, A Simple Model of Capital Market Equilibrium with Incomplete Information,
Journal of Finance, 42(3).
ƒ Schwartz, T., 2000, How to Beat the S&P500 with Portfolio Optimization, DePaul University,
working paper.
ƒ Sharpe, W.F., 1991 , “Capital Asset Prices with and without Negative Holdings”, Journal of
Finance, 46.
ƒ Tang, Y., and Shum, 2003, The relationships between unsystematic risk, skewness and stock
returns during up and down markets, International Business Review.
ƒ Tinic, S., and R. West, 1986, Risk, Return and Equilibrium: A revisit, Journal of Political Economy,
94, 1, 126-147.
ƒ Tobin, J., 1958, Liquidity Preference as Behavior Towards Risk, Review of Economic Studies, 67,
65-86.
ƒ Zhang, Y., 2005, Individual Skewness and the Cross-Section of Average Stock Returns, Yale
University, working paper.

43