Preface
CHAPTER 1 An introduction to asset pricing models1.1 Historical asset pricing models
1.2 The beginning of modern asset pricing models
1.2.1 Markowitz portfolio optimization
1.2.1.1 Portfolio return and risk
1.2.1.2 The optimization problem
1.2.1.3 The Lagrangian formulation
1.2.1.4 The minimum risk hyperbola
1.2.1.5 The global minimum variance
(global minimum risk) portfolio
1.2.1.6 The efficient frontier
1.2.1.7 Efficient portfolios and their zero-correlation
“matching” portfolios
1.2.1.8 Deriving the equation for a tangent
line to the minimum standard
deviation hyperbola
1.2.1.9 Correlation between portfolios
1.3 The Two-Fund theorem
1.4 The CAPM with a riskfree asset
1.5 Sharpe’s capital asset pricing model
1.5.1 Estimating the CAPM model
1.6 Efficient markets
1.7 Studies that attack the CAPM
1.8 Does proving the CAPM wrong=market inefficiency?
or, do efficient markets=the CAPM is correct?
1.9 Small capitalization and value stocks
1.9.1 Momentum stocks
1.10 The asset pricing models of today
1.10.1 Introduction to multifactor models
1.10.2 Multifactor models of stock returns
1.10.2.1 Regression-based models
1.10.3 Multifactor models of bond returns
1.11 The q-factor pricing model
1.11.1 Economic intuition
1.11.1.1 Marginal q and investment
1.11.1.2 Expected stock returns and investment
1.11.1.3 Investment and profitability channels
1.11.2 q-Model factor construction
1.11.3 Operationalizing the q-model
1.12 Summary
1.13 Limitations
1.14 Chapter-end problems
1.15 Appendix: calculus operations using matrix algebra
References
CHAPTER 2 An introduction to returns-based performance
evaluation and potential biases inits econometric application2.1 Introduction
2.2 Goals, guidelines, and perils of performance evaluation
2.2.1 Benchmarks
2.2.2 Performance measures
2.2.3 Manipulation-proof performance measures
2.2.4 Type I or Type II error (which would you prefer?)
2.2.5 The confounding role of risk-aversion
2.3 Returns-based performance assessment
2.3.1 Baseline models
2.3.2 Models with timing factors
2.3.3 Return smoothing
2.3.4 Non-normal alphas
2.3.5 Nonstable regression parameters
2.3.6 Unpriced benchmarks
2.3.7 Bayesian methods
2.3.8 Conditional returns-based performance measurement
2.3.9 Stochastic discount factors
2.3.10 False-discovery rate approach to measuring
performance of a group of funds
2.3.11 The frequency of return reporting
2.3.11.1 High-frequency returns data
2.3.11.2 Low-frequency returns data
2.3.12 Overlapping observations corrected standard errors
2.3.12.1 An example using overlapping
quarterly returns
2.5 Chapter-end Problems
References
CHAPTER 3 Returns-based performance measures3.1 Introduction
3.2 Luck versus skill
3.3 The ultimate goal of performance measures
3.4 Two nonregression approaches
3.4.1 The Sharpe Ratio
3.4.2 Tracking error
3.5 Regression-based performance measures
3.5.1 Single-factor alpha (“Jensen alpha”)
3.5.2 Multiple-factor alpha
3.5.3 Timing and selectivity performance measures
3.5.4 Conditional regression models
3.5.5 The Information Ratio as a performance measure
3.6 Chapter-end problems
References
CHAPTER 4 Portfolio holdings–based performanceevaluation4.1 Introduction
4.2 Unconditional holdings–based performance measurement
4.2.1 The self-benchmarking method of performance
evaluation
4.2.1.1 Statistical foundations
4.2.1.2 Empirical evidence
4.2.1.3 Relation to the Brinson, Hood, and
Beebower attribution approach
4.2.2 The DGTW method of performance evaluation
for equity portfolios
4.2.2.1 The characteristic selectivity measure
4.2.2.2 The characteristic timing measure
4.2.2.3 The average style measure
4.2.2.4 Summing the components
4.2.2.5 Comparison of DGTW measures with
factor-based regression approaches
4.2.2.6 Extensions
4.2.2.7 Empirical evidence
4.2.2.8 The correlation between performance
measures
4.2.3 The Cici and Gibson method of performance
evaluation for bond portfolios
4.2.4 The Cohen, Coval, and Pastor method of
performance evaluation
4.3 Conditional holdings–based performance measurement
4.3.1 The Ferson-Khang conditional portfolio
holdings approach
4.3.1.1 Description
4.3.1.2 Estimation
4.3.1.3 Empirical evidence
4.4 Chapter-end problems
References
Further reading
CHAPTER 5 Combining portfolio holdings-based and
returns-based performance evaluation
(and the “return gap”)5.1 Introduction
5.2 Performance decomposition methodology
5.2.1 The characteristic selectivity measure
5.2.1.1 Examples of style drift
5.2.1.2 Benchmarking stocks (computing
abnormal returns)
5.2.1.3 Example of the CS measure for US mutual
funds (a return to our famous managers)
5.2.1.4 Benchmarking bonds (computing abnormal
returns)
5.2.2 The characteristic timing measure
5.2.3 The average style measure
5.2.4 Trade execution costs
5.2.5 Measuring net return selectivity
5.3 Application to US domestic equity mutual funds
5.4 Empirical results for US domestic equity mutual funds
5.4.1 Overall mutual fund returns
5.4.2 Benchmark-adjusted mutual fund returns
5.4.3 The correlation between performance measures
5.4.4 Baseline mutual fund return decomposition
5.4.5 A comparison of the average mutual fund to the
Vanguard Index 500 fund
5.4.6 Do funds that trade more frequently generate
better performance?
5.5 Results for US domestic corporate bond mutual funds
5.6 Appendix A
5.6.1 Description of matching process for LSEG 12
and CRSP mutual fund databases
5.7 Appendix B
5.7.1 Description of execution cost estimation
procedure
5.8 Chapter-end problems
References
Further reading
CHAPTER 6 Fund manager selection using macroeconomic
information6.1 Introduction
6.2 A dynamic model of managed fund returns
6.2.1 The “Dogmatist”
6.2.2 The “Skeptic”
6.2.3 The “Agnostic”
6.2.4 Optimal portfolios of managed funds
6.3 Empirical example: US domestic equity fund data
6.4 Empirical example: results for US domestic equity funds
6.4.1 Optimal portfolios of equity mutual funds
6.4.2 Out-of-sample performance
6.4.3 The determinants of the superior
predictability-based performance
6.4.3.1 Attributes of portfolio strategies
6.4.3.2 Industry allocation analysis
6.4.3.3 Industry attribution analysis
6.4.4 Survivorship bias
6.5 Chapter-end problems
6.6 Appendix A: Description of mutual fund database
A.1 Investment objectives
A.2 Net returns
A.3 Turnover and expenses
A.4 Flows
6.7 Appendix B: Investments when fund risk loadings
and benchmark returns may be predictable
B.1 Prior beliefs
B.2 The likelihood function
B.3 The predictive moments
6.8 Appendix C: Investments when skills may be predictable
C.1 The Agnostic
C.2 The Skeptic
References
CHAPTER 7 Performance evaluation of market timers:
a new approach7.1 Introduction
7.2 Methodology
7.2.1 Cashflow and discount rate components
of market returns
7.2.2 Construction of fund beta
7.2.3 A differential return timing measure
7.3 Data and variable construction
7.4 Empirical analysis of timing performance
7.5 Identifying funds with timing skills
7.5.1 Characteristics of funds ranked on past-year
total timing
7.5.2 Timing performance for funds sorted on
past-year total timing
7.5.3 Strategic shifts in fund betas
7.5.4 Other dimensions of fund portfolio performance
7.6 Further characterizing cashflow versus discount rate timing
7.6.1 Industry rotation
7.6.2 Large-cap versus small-cap and value versus
growth mutual funds
7.6.3 Negative discount rate timing and aggregate
fund net flows
7.7 Additional analyses and robustness tests
7.7.1 Initiating buys and terminating sells
7.7.2 Time-varying cash positions
7.7.3 A placebo test using index funds
7.7.4 Using different approaches to decompose the
market return
7.7.5 Timing ability over the business cycle
7.7.6 Other tests
7.8 Chapter-end problems
References
Further reading
CHAPTER 8 Performance evaluation of non-normal portfolios8.1 Introduction
8.2 Bootstrap evaluation of fund alphas
8.2.1 Rationale for the bootstrap approach
8.2.1.1 Individual mutual fund alphas
8.2.1.2 The cross-section of mutual fund alphas
8.2.2 Implementation example: US domestic equity
mutual funds
8.2.2.1 The baseline bootstrap procedure:
residual resampling
8.2.2.2 Bootstrap extensions
8.3 Data
8.4 Results for US equity funds
8.4.1 The normality of individual fund alphas
8.4.2 Bootstrap analysis of the significance of
alpha outliers
8.4.2.1 Baseline bootstrap tests: residual
resampling
8.5 Sensitivity analysis
8.5.1 Time series dependence
8.5.2 Residual and factor resampling
8.5.3 Cross-sectional bootstrap
8.5.4 Length of data records
8.5.5 Bootstrap tests for stockholdings-based alphas
8.6 Performance persistence
8.7 Chapter-end problems
References
Further reading
CHAPTER 9 Multiple fund performance evaluation:
the false discovery rate approach9.1 Introduction
9.2 The impact of luck on managed fund performance
9.2.1 Overview of the approach
9.2.1.1 Luck in a multiple fund setting
9.2.1.2 Measuring luck
9.2.1.3 Estimation procedure
9.2.2 Comparison of our approach with existing methods
9.2.3 Cross-sectional dependence among funds
9.3 An empirical example: US domestic equity mutual funds
9.3.1 Asset pricing models
9.3.2 Data
9.4 An empirical example: results for US domestic
equity funds
9.4.1 The impact of luck on long-term performance
9.4.2 The impact of luck on short-term performance
9.4.3 Performance persistence
9.4.4 Additional results
9.4.4.1 Performance measured with pre-expense
returns
9.4.4.2 Performance measured with other asset
pricing models
9.4.4.3 Bayesian interpretation
9.5 Chapter-end problems
References
Further reading
CHAPTER 10 Holding Horizon: a new measure of active
investment management10.1 Introduction
10.2 Empirical methodology
10.2.1 The measure of holding horizon
10.2.2 Risk models
10.3 Data and summary statistics
10.3.1 Summary statistics
10.3.2 The persistence of fund holding horizon
10.4 Empirical results on fund performance
10.4.1 Fund performance using a sorted portfolio
approach
10.4.2 Fund performance using Fama–MacBeth
regressions
10.4.3 Value added from financial markets
10.5 The horizon–performance relation at the stock level
10.5.1 Informativeness of fund holdings
10.5.2 Economic source
10.6 Comparison of H-H with portfolio turnover
10.7 The demand side
10.8 Additional analyses and robustness tests
10.8.1 Being a new dimension of active fund
management
10.8.2 Illiquidity
10.8.3 Out-of-sample test of H-H’s ex ante predictability
10.8.4 Fund performance conditional on benchmarks
10.8.5 Other tests
10.9 Chapter-end problems
10.10 Appendix
References
CHAPTER 11 Target date funds: an analysis of strategies
and performance11.1 Introduction
11.2 Related literature
11.3 The evolution of target date funds’ role in US retirement
savings
11.3.1 Description of target date funds
11.3.2 The rise of target date funds to become a
fundamental pillar of US retirement savings
11.4 Heterogeneity across target date fund suites
11.4.1 Variation in glide paths
11.4.1.1 Description and evolution of glide paths
11.4.1.2 To be “to” or to be “through”:
that is the question!
11.4.2 Variation in asset allocations
11.4.3 Glide path design
11.4.4 Variation in the use of tactical asset allocation
11.4.5 Active versus passive underlying funds:
yet another choice to be made
11.4.6 Variation in investment fund structure
11.5 Evaluating the quality of target date fund suites
11.5.1 Overview of the evaluation of investment funds
11.5.2 Complications in evaluating target date funds
11.5.3 Framework for evaluating target date funds
11.5.3.1 Evaluating a target date fund’s investment
objectives and associated profile of risk
and expected returns
11.5.3.2 Evaluating a target date fund manager’s
ability to implement the target date fund’s
investment objectives
11.5.3.3 Do multiple comparisons over multiple
periods
11.5.3.4 Performance attribution to evaluate the
underlying drivers of target date funds
performance
11.5.3.5 Comparison of ex post asset class returns
to ex ante expected returns
11.6 Future Developments in the TDF Marketplace
11.7 Conclusion
11.8 Chapter-end problems
References
CHAPTER 12 Fund rating systems12.1 Introduction
12.1.1 Ideal properties of a rating system
12.1.2 Objectives of a fund rating system
12.2 Quantitative versus qualitative performance metrics
12.3 Handling historical fund data
12.3.1 Backfilling of returns of liquidated funds and
splicing of returns of merged or acquired funds
12.3.2 Assessment of managed funds in the presence
of overlapping data observations
12.4 The Morningstar rating systems
12.4.1 The style box classification system
12.4.2 The Morningstar Star Rating system
12.4.2.1 Computation of the Morningstar
risk–adjusted return
12.4.2.2 Within-category ranking on the
Morningstar risk–adjusted return
12.4.3 Is the Star Rating system predictive of
risk-adjusted performance?
12.4.4 The Morningstar Analyst Rating system
12.4.4.1 The Morningstar quantitative rating
system and the Morningstar Medalist
Rating system
12.4.5 Are the Morningstar Medalist Ratings predictive
of fund risk–adjusted returns?
12.4.6 The Morningstar Sustainability Rating system
12.5 The Lipper rating systems
12.5.1 The Lipper fund classification system
12.5.2 The Lipper Leader rating system
12.5.3 Is the Lipper Leader rating system predictive
of risk-adjusted performance?
12.6 The Zacks rating system
12.7 Standard and Poor’s rating systems
12.8 Other investment rating services
12.9 Do the fund ratings services conform to the ideals
proposed by peer-reviewed academic literature?
12.10 The future of fund ratings systems
12.11 Chapter-end problems
References
Further reading
CHAPTER 13 Active management in mostly efficient markets:
a survey of the academic literature13.1 Introduction
13.2 Some caveats
13.3 Does active management add value?
13.4 Active management and “mostly efficient markets”
13.5 Identifying superior active managers
13.5.1 Past performance
13.5.2 Macroeconomic forecasting
13.5.3 Fund/manager characteristics
13.5.4 Portfolio holdings analysis
13.5.4.1 Active share
13.5.4.2 News sensitivity and interpretation
13.5.5 Value added performance
13.5.6 Active risk budgeting
13.6 Conclusions
13.7 Chapter-end problems
References
Further reading
A complete solutions manual for all chapter-end problems in this volume
is available from the author,
[email protected]
