Factor Investing with Black–

Litterman–Bayes: Incorporating
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

Factor Views and Priors in Portfolio

Construction
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

Petter N. Kolm and Gordon Ritter

Petter N. Kolm
is a clinical professor and
KEY FINDINGS
director of the Mathematics
in Finance Master’s Program n The authors propose a general framework referred to as Black–Litterman–Bayes (BLB)
in the Courant Institute of
for constructing optimal portfolios for factor-based investing.
Mathematical Sciences at
New York University in New n The framework allows for the incorporation of investor views and priors on factor risk
York, NY. premiums, including data-driven and benchmark priors.
[email protected]
n The authors provide computationally efficient closed-form formulas for the (posterior)
Gordon Ritter expected returns and return covariance matrix.
is an adjunct professor at
the Courant Institute of n In a realistic empirical example, using a number of well-known cross-sectional US equity
Mathematical Sciences, factors, they demonstrate that the BLB approach can add value to mean–variance-op-
NYU Tandon School of timal multi-factor risk premium portfolios.
Engineering, Baruch
College, Columbia University ABSTRACT
in New York, NY and General
Partner/CIO at Ritter Alpha, The authors propose a general framework referred to as Black–Litterman–Bayes (BLB)
LP in New York, NY. for constructing optimal portfolios for factor-based investing. In the spirit of the classical
[email protected]
Black–Litterman model, the framework allows for the incorporation of investor views and
priors on factor risk premiums, including data-driven and benchmark priors. Computationally
efficient closed-form formulas are provided for the (posterior) expected returns and return
covariance matrix that result from integrating factor views into an arbitrage pricing theory
multi-factor model. In a step-by-step procedure, the authors show how to build the prior
and incorporate the factor views, demonstrating in a realistic empirical example and using
a number of well-known cross-sectional US equity factors, that the BLB approach can add
value to mean–variance-optimal multi-factor risk premium portfolios.

TOPICS
Factor-based models, factors, risk premia, portfolio construction, portfolio theory*

R
isk premiums and smart beta investing have attracted considerable attention
from institutional and individual investors for their simplicity, transparency, and
*All articles are now
categorized by topics
low cost. Johnson (2018) estimated that at the end of 2018 almost 1,500 smart
and subtopics. View at beta products were traded on the public exchanges alone, amounting to about $800
PM-Research.com. billion of assets under management worldwide. These funds represent a hybrid of
 114 | Factor Investing with Black–Litterman–Bayes: Incorporating Factor Views and Priors Quantitative Special Issue 2021

active and passive investing, often deploying rules-based investment strategies and
portfolio construction methodologies. Nevertheless, the main goal of these invest-
ment vehicles is to outperform the market, just like that of active approaches. For
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

this purpose, the majority of these funds rely on combining well-known risk premiums
into a portfolio, perhaps by a simple equally weighted approach or more sophisticated
portfolio optimization schemes. Numerous academic studies provide support for
this form of factor investing, confirming the presence of factor premiums in different
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

asset classes, domestically and internationally, and across various themes, such as
betting against beta, carry, low volatility, momentum, quality, return seasonality, size,
and value (see, e.g., Fama and French 1993; Carhart 1997; Subrahmanyam 2010;
Moskowitz, Ooi, and Pedersen 2012; Asness, Moskowitz, and Pedersen 2013; Ang
2014; Frazzini and Pedersen 2014; Keloharju, Linnainmaa, and Nyberg 2016; Koijen
et al. 2018; and Baltussen, Swinkels, and Van Vliet 2019).
Although a number of studies have addressed portfolio construction for risk pre-
miums (see, e.g., Bass, Gladstone, and Ang 2017; Bergeron, Kritzman, and Sivitsky
2018: Dopfel and Lester 2018; Bender, Le Sun, and Thomas 2018; and Aliaga-Diaz
et al. 2020), most of the work in this area has focused on factor models from the
frequentist’s perspective.
First introduced in two Goldman Sachs Fixed Income Research Notes in the early
1990s (Black and Litterman 1990, 1991b) and later published in academic journals
(Black and Litterman, 1991a, 1992), perhaps the most well-known Bayesian portfolio
construction methodology is the Black–Litterman (BL) model. The purpose of BL is
to improve the inputs in portfolio construction. A key insight of Black and Litterman
is that the unreasonable and unstable portfolio holdings frequently encountered in
classical mean–variance optimization (MVO) is the result of feeding noisy and incon-
sistent risk and return forecasts into an optimizer. They realized that risk and return
forecasts have to be consistent with one another to be effective in optimization. Their
solution resulted in the BL model, which combines the market-implied views with
investor views, producing consistent risk and return forecasts (see, e.g., Litterman
and He 1999; Satchell and Scowcroft 2000; Fabozzi, Focardi, and Kolm 2006, 2010;
Fabozzi et al. 2007; and Meucci 2009 for detailed expositions and discussions of
the BL model). The BL method itself is often described as “being Bayesian,” but the
original authors do not elaborate directly on its connections with Bayesian statis-
tics. Kolm and Ritter (2017) introduced the Black–Litterman–Bayes (BLB) model, a
most general Bayesian portfolio optimization procedure. They elucidated the direct
link between BL and Bayesian regression, highlighting its relationship to Bayesian
networks (Pearl 2014). In addition, they provided a detailed Bayesian treatment of
views on factor risk premiums in the context of multifactor models. Specifically, they
showed that the arbitrage pricing theory (APT) of Ross (1976), today a cornerstone in
the practice of quantitative investing, has a natural Bayesian extension. They derived
the associated BL optimal factor portfolio and discussed two common types of priors
in these settings, benchmark-optimal and data-driven priors.
The authors of several related studies have explored BL and more general Bayes-
ian approaches to portfolio construction, including Jones, Lim, and Zangari (2007);
Zhou (2009); Avramov and Zhou (2010); Figelman (2017); Rubesam and Hwang
(2019); and Melas et al. (2019).
In this article, we explore the BLB model for factor portfolios and factor views, one
of the models introduced by Kolm and Ritter (2017). In particular, we take as a premise
that some asset managers and hedge funds employ systematic investment processes
in which the principal goal is to optimally harvest various factor premiums. In other
words, these managers base their investment decisions on the view that certain factor
risk premiums are likely to deliver risk-adjusted returns that are attractive or cannot
be obtained through the capital asset pricing model (CAPM) market portfolio alone.
 Quantitative Special Issue 2021 The Journal of Portfolio Management | 115

In the original BL framework, all views were on portfolios of securities. Factors

are not portfolios; rather, they are unobservable random variables that represent
sources of risk that are common to all securities trading in the market. Hence, the
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

BL model is not directly applicable to views about factor performance. However, an

extension of the same method used by Black and Litterman allows a portfolio man-
ager to express views on factors directly.
In this article, we make several contributions. First, we show how to incorporate
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

views on factor risk premiums into APT models. Second, we provide concrete closed-
form formulas for the BLB expected returns and covariances that result from inte-
grating an investor’s factor views into an APT model. Third, with a realistic empirical
example, we demonstrate how the incorporation of factor views can add value to
multifactor risk premium portfolios. Specifically, we describe a step-by-step procedure
to build the prior and incorporate the factor views and demonstrate that, in the event
the views are prescient, the BLB portfolio strategy outperforms an analogous MVO
strategy in which the latter uses the views to build expected returns but does not
benefit from the BLB prior or likelihood.

THE BLACK–LITTERMAN MODEL

The BL model blends market-implied expected returns with investor views, taking
into account the investor’s confidence about these inputs (Black and Litterman 1991a,
1992). In other words, the so-called BL expected returns are a confidence-weighted
average of the market-implied expected returns and the investor’s views.
We consider a market with n securities whose returns rt are multivariate and
normally distributed with an expected return vector, μ, and return covariance matrix,
S; that is

rt ~ N (µ, Σ ) (1)

A mean–variance investor will choose portfolio holdings, h = (h1, …, hn)′, by solving

the MVO problem

λ 2
max µ p − σp (2)
h 2

where mp := E[h′rt] = h′µµ is the expected portfolio return, σ 2p := V[h′ rt ] = h′ Σh is the

portfolio variance, and l is the investor’s risk aversion parameter that determines
the trade-off between expected portfolio return and risk. The portfolio holdings h are
in units of dollars, or whatever currency the investor is using. By solving Equation 2,
we obtain the optimal holdings

h∗ = λ −1 Σ −1 µ

In the BL model, investors can have their own views. A view might be that “the
German equity market will outperform a capitalization-weighted basket of the rest of
the European equity markets by 5%” (Litterman and He 1999). To express this view,
we let p = (p1, …, pn)’ denote the portfolio that is long one unit of the DAX index and
short a one-unit basket that holds each of the other major European indexes (e.g.,
UKX, CAC40, AEX) in proportion to their respective aggregate market capitalizations
so that Si pi = 0. Denoting the expected return of this portfolio by q = 0.05, we can
express the investor’s view as E[p′r] = q, where r is the vector of security returns over
the considered time horizon. If the investor has k such views
 116 | Factor Investing with Black–Litterman–Bayes: Incorporating Factor Views and Priors Quantitative Special Issue 2021

E [pi′r ] = qi , for i = 1, …, k (3)

the portfolios pi are more conveniently arranged as rows of the k × n matrix, P. Then
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

the investor’s views can be compactly expressed as

E[Pr] = q (4)
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

where q = (q1, …, qk)′. However, the expectations (Equation 4) alone do not convey
how confident investors may be in their views. For this purpose, in the BL model, the
investor specifies a level of confidence via the k × k matrix W such that

q = Pµ + ε q , ε q ~ N (0, Ω) (5)

In other words, an investor’s views are represented as the (1) mean and (2)
covariance matrix of linear combinations of the expected returns m. This information
is partial and indirect because the views are statements about portfolio returns
rather than about individual security returns directly. Furthermore, this information
is noisy, with the noise eq assumed to be multivariate normally distributed, because
the future is uncertain.
Black and Litterman were motivated by the principle that, in the absence of any
sort of investor views and constraints, the mean–variance-optimal portfolio holdings
should be the global CAPM equilibrium portfolio, heq. Consequently, in the absence
of any views, the investor’s model for security returns is given by Equation 1 with

µ ~ N ( π, C) (6)

where p is the global CAPM expected return in excess of the risk-free rate, and the
inverse of the covariance matrix C represents the amount of confidence an investor
has in the estimate of p. The distributional assumption (Equation 6) is referred to
as the CAPM prior.
Given the security return distribution (Equation 1), the prior (Equation 6) and the
investor’s views (Equations 5), the BL model yields the (posterior) expected returns

µ BL := (P ′Ω −1P + C −1 ) −1 (P ′Ω −1 q + C −1 π ) (7)

and (posterior) return covariance matrix

Σ BL := V[µ BL ] = (P ′Ω −1P + C −1 ) −1 (8)

frequently referred to as the Black–Litterman expected returns and covariance matrix,

respectively. Kolm and Ritter (2017) provided a full Bayesian derivation of the BL
model.

BLACK–LITTERMAN FOR FACTOR INVESTING

Linear Factor Models

Taking APT (Ross 1976; Roll and Ross 1980) as a starting point, we consider a
market in which the return in excess of the risk-free rate of the ith security on day t
is given by

rt,i = xt,i,1 ft,1 + xt,i,2 ft,2 +  + xt,i,k ft,k + ε t,i (9)

 Quantitative Special Issue 2021 The Journal of Portfolio Management | 117

where ft,j, j = 1, …, k are k factors, xt,i,j denotes the jth factor loading of the ith security,
and et,i is the residual return (or idiosyncratic return) of the ith security. We assume
the residual returns et,i are white noise processes that are mutually uncorrelated with
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

each other and across time and are uncorrelated with all factors; that is

E[ε t,i ] = 0, for all i and t,

Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

cov[ft, j , ε s,i ] = 0, for all j, i, t, and s,

 σ 2 , if i = j and t = s,
cov[ε t,i , ε s, j ] =  t,i
 0, otherwise

Additionally, we assume for simplicity that the residual returns are normally dis-
tributed

ε t,i ~ N (0, σ 2t,i ) (10)

Typically, factor loadings xt,i,j are known, exogenously given, and nonstochastic.
Some examples of factors include size and value. For a size factor, the loadings are
frequently some nonlinear transformation of the market capitalization of each stock,
which are known and exogenous. For a value factor, the loadings are frequently taken
to be some measure of the book value or earnings of each company, divided by market
capitalization. The value factor is also often passed through a nonlinear transforma-
tion to normalize it cross sectionally. The factors f t,j, j = 1, …, k are unobservable
(or hidden) random variables that collectively account for all co-movement of security
returns. Because they cannot be observed directly, they must be obtained via sta-
tistical inference. In contrast to the residuals, different factors are not necessarily
independent.
Using matrix/vector notation, we write Equations 9 and 10 as

rt = X t ft + ε t , ε t ~ N (0, Dt ) (11)

where rt := (r t,1, …, r t,n)′ is an n-dimensional random vector of cross-sectional secu-

rity returns in excess of the risk-free rate, Xt is the n × k matrix of factor loadings
known before time t, ft is the k-dimensional random vector process of factors, et is
the n-dimensional random vector process of residual return, and

Dt := diag(σ 2t,1,…, σ 2t,n ), where all σ t, j > 0 (12)

is the (strictly positive) n × n diagonal matrix of residual variances.

We note that rt represents close-to-close total returns from t - 1 to t, when t
denotes a day or, when more precision is required, the exact time of the equity market
close on day t. The subscript t has the same meaning for the other random processes.
We assume the ft process has finite first and second moments such that

E[ft ] = µ f , and V[ft ] = Ft (13)

where the elements of mf are referred to as the factor risk premiums. From Equations
11 and 13, we immediately obtain that the expected security returns and their cova-
riance matrix are given by

E[rt ] = X t µ f (14)
 118 | Factor Investing with Black–Litterman–Bayes: Incorporating Factor Views and Priors Quantitative Special Issue 2021

V[rt ] = Dt + X tFt X t′ =: Σ t (15)

where X t′ denotes the transpose of Xt. We note that Equation 15 and the positivity
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

assumption in Equation 12 together imply that the security-level inverse Σ t−1 exists.
Classical MVO uses the security-level covariance matrix and expected returns. As is
well known, the historical sample covariance matrix is a poor estimator and should not
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

be used for MVO. For instance, if we consider a universe that is similar to the Russell
3000, then one has n = 3,000 securities and n(n + 1)/2 × 4,500,000 independent
elements in the covariance matrix, which is too many independent parameters to esti-
mate from daily security returns. In this situation, APT models are especially useful in
that they provide the covariance matrix estimator (Equation 15) that, unlike the sample
covariance matrix, leads to stable, well-diversified optimal portfolios (Ritter 2016).
We state without proof that the inverse of Equation 15 can be obtained by the
Woodbury matrix inversion lemma, given by

Σ t−1 = Dt−1 − Z t [Ft−1 + X t′Z t ]−1 Z t′ (16)

where Z t := Dt−1 X t . The key to making computations (Equation 16) run quickly is to
systematically avoid the bottleneck of creating and manipulating a full n × n matrix
(as opposed to n × k or k × k matrixes, where k is much smaller than n). For example,
in the Markowitz case, one should compute

Σ t−1 µ t = Dt−1 µ t − Z t [Ft−1 + X t′Z t ]−1 Z t′µ t (17)

observing that the right-hand side can be calculated without having to build a full
n × n matrix.

Black–Litterman for Factor Models

Leveraging the general BLB approach of Kolm and Ritter (2017), we describe a
BL-type model for factor models of the kind in Equations 9 and 10. For simplicity,
we treat Ft as known. Therefore, once Xt,Dt are given, St is also known via Equation
15, just as in the original BL model. It follows that the source of uncertainty is in the
expected factor premiums mf. Next, we discuss two kinds of priors for factor premiums.

Data-Driven Priors
If the ft process is stationary, such that mf and ft are approximately constant over
time, then we can obtain a prior, pf, from a (Bayesian) time-series model of the factor
returns ft. One possible choice is the mean (or rolling mean) of a time series of the
ordinary least squares estimates

ˆft = (X t′X t )−1 X t′rt +1 (18)

More sophisticated approaches, such as that of hierarchical or mixed-effects

models, can also be used (Gelman et al. 2003).
An advantage of data-driven priors is that they do not require a benchmark
portfolio (such as the capitalization-weighted benchmark in the original BL model).
Consequently, data-driven priors may be preferred in absolute return strategies in
which the effective benchmark is cash. Below we provide an empirical example that
uses a data-driven prior in a hypothetical absolute-return strategy.
 Quantitative Special Issue 2021 The Journal of Portfolio Management | 119

Benchmark Priors
If there is a benchmark portfolio, hB, then close in spirit to the original BL model
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

is to find a benchmark-optimal prior. For a general discussion of benchmark-optimal

priors in this setting, we refer the reader to Kolm and Ritter (2017). For instance, we
could use the prior
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

π f ~ N (ξ, V ) (19)

where x is a k-dimensional parameter vector, and V is a k × k matrix that is symmet-

rical and positive definite.
Kolm and Ritter (2017) showed that by combining this prior with the factor model
of Equations 9 and 10, one obtains the a priori expected returns and return covari-
ance matrix

E π [r] = ( Σ −1 + Σ −1 XH−1 X ′Σ −1 )−1 Σ −1 XH−1 V −1 ξ (20)

Vπ [r] = ( Σ −1 + Σ −1 XH−1 X ′Σ −1 )−1 , with (21)

H : = V −1 + X ′Σ −1 X

Here all quantities are assumed to be sampled at the same time t. However, for read-
ability purposes, in this section, we omit the explicit time index, allowing the time depen-
dence to be implicit. Hence, the a priori mean–variance-optimal portfolio is given by

hprior = (λVπ [r])−1 ⋅ E π [r] = λ −1 Σ −1 XH−1 V −1 ξ (22)

From Equation 22, it is clear that unlike the original BL, any arbitrary benchmark
portfolio cannot be realized as an a priori optimal portfolio. Those that can are nec-
essarily of the form l-1S-1Π, where Π is some linear combination of the columns of X.
These are precisely the portfolios that are optimal with respect to a set of individual
risk premiums that are in the factor model. Because not every possible portfolio is
realizable as a priori optimal, the market portfolio may also not be optimal. Never-
theless, if (1) the market is in a CAPM equilibrium and (2) one of the columns of X
contains the CAPM betas, then the individual risk premiums will be proportional to
that column of X and the market portfolio is realizable as a priori optimal.

Factor Views
We are free to express views on factor risk premiums in the same way we would
express views in the original BL model. In particular, we can express views on any
(linear) portfolios of risk premiums. However, it is less likely for portfolio managers
to have views on linear combinations of factors. Perhaps the most common and
parsimonious situation is one in which we have a view on each factor risk premium
that is independent of our views on other factors. Using the value and momentum
factors as an example, a portfolio manager might have two independent views: (1)
a view on the value premium and (2) a view on the momentum premium. We can
express views of this kind as

q = µ f + ε q, ε q ~ N (0, Ω) (23)

where Ω = diag(ω 21,…, ω 2k ) is a diagonal k × k matrix with all wj > 0.

 120 | Factor Investing with Black–Litterman–Bayes: Incorporating Factor Views and Priors Quantitative Special Issue 2021

The Black–Litterman Expected Returns and Covariance Matrix

for Factor Portfolios
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

By combining the factor model (Equations 9 and 10), the factor prior
(Equation 19) and the investors factor views (Equation 23), Kolm and Ritter (2017)
showed that the (posterior) expected returns and return covariance matrix are given by
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

µ BL-fac = Σ BL-fac ⋅ Σ −1 X (V −1 + X ′Σ −1 X )−1 V −1 ξ (24)

Σ BL-fac = ( Σ −1 + Σ −1 X (V −1 + X ′Σ −1 X )−1 X ′Σ −1 )−1 (25)

where all quantities are sampled at time t, and V := (V −1 + Ω −1 )−1, and ξ := V (V −1 ξ + Ω −1q)
are the posterior hyperparameters. We emphasize that S is the covariance matrix of
security returns from the factor model (Equation 15). Consequently, from the expected
returns (Equation 24) and covariance matrix (Equation 25), we obtain the MVO portfolio

h* = λ −1 Σ −1 Π (26)

where

Π := Xµ f (27)

µ f := (V −1 + Ω −1 + X ′Σ −1 X)−1 (V −1 ξ + Ω −1q) (28)

Equations 26–28 represent the solution to BLB optimization in the context of factor
models. We observe that the security-level risk premiums Π = Xµ f are linear combi-
nations of the factors that form the columns of X. Intuitively, we can think of µ f as a
set of factor risk premiums adjusted to take account of the views. These adjustments
give more weight to factors that have high prior mean–variance ratios, xi/Vii, and/or
high expected return-uncertainty ratios, qi /ω 2i .

Discussion
The original work of Black and Litterman was done under normality assumptions
for the return distributions. This is mathematically convenient because the normal
prior is a conjugate prior for the normal likelihood, so the posterior is also normal with
all means and covariances obtainable in closed form. Nevertheless, it is of clear inter-
est to investigate the applicability of similar methods beyond the normal distribution.
Chamberlain (1983) showed that under elliptically distributed return distributions,
expected utility—for any concave utility function—is a function of only the portfolio’s
return and variance. Elliptical distributions are a large family of distributions, including
many fat-tailed distributions such as the Student’s t-distributions. In the context of
BLB optimization, this implies that we are free to use any form for the prior and the
views—including models in which the unknown parameter(s) need not simply be the
mean return—as long as the final posterior distribution p(rt+1|q) of the cross-sectional
security return vector is a member of the elliptical family; then any risk-averse expect-
ed-utility maximizer will end up maximizing the posterior expected portfolio return
minus a constant multiple of the posterior portfolio variance. Of course, with a dif-
ferent (e.g., nonnormal) model, the investor need not arrive at precisely the same
expected return and variance formulas as did Black and Litterman. Additionally, if the
 Quantitative Special Issue 2021 The Journal of Portfolio Management | 121

prior is not conjugate to whatever likelihood is chosen, then computing the posterior
moments may be difficult or even impossible.
One of the early successes of the Black–Litterman model was that it provided a
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

form of regularization of the fundamentally unstable inversion of the historical sample

covariance matrix. However, the Black–Litterman technique is not primarily meant as
a regularization technique. It has that property as a by-product because its likelihood
and prior tend to pull the portfolios toward known, reasonable portfolios.
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

Importantly, with common usages of APT, regularization is, strictly speaking, no

longer necessary because the resulting covariance matrix is already stably invertible.
For example, when the mean–variance portfolio is computed via Equation 17, the
phenomenon of corner portfolios does not occur. The resulting portfolios are typically
well diversified, as long as the factor model itself is reasonable.
We stress that, even with a stable covariance matrix, the BL procedure is still
tremendously valuable—perhaps even more so, because even the classical BL model
does still use S as an input, while being less sensitive to its instabilities. However,
as compared to MVO, the BL model gives the portfolio manager additional degrees of
freedom to control the portfolio, in the form of wi, which quantifies the uncertainty in
the ith view. A key difference is that the wi parameters do not appear in the classical
Markowitz formulation.

AN EMPIRICAL CASE STUDY

In this section, we provide an empirical example of the BLB model for factor mod-
els and investigate whether our methodology provides reasonable empirical results.
Specifically, we construct an instance of BLB model (Equations 26–28) suitable for
quantitatively oriented institutional portfolio managers interested in systematically
trading single-name US equities in a factor-based fashion.
Our universe of stocks is chosen with certain liquidity thresholds in place. In
particular, stocks must be priced above $5, bid–offer spreads must be less than
100 bps, and they must trade at least 10,000 shares and $1 million of volume per
day. We restrict attention to only common stock and only the most liquid share class
of each company. This typically results in a universe of 1,800–2,000 stocks, but in
the period around the 2008 financial crisis, the size of our universe briefly dropped
to around 1,400 stocks.
Our factor model includes a Global Industry Classification Standard–based indus-
try classification with 58 industries that are either single Global Industry Classification
Standard subindustries or several closely related subindustries merged together.
Our model also includes a set of continuous numerical factors known colloquially as
style factors because they are thought to represent investment styles such as value,
contrarian, or dividend capture.
Roughly speaking, factor timing is defined as attempting to predict which factors
will have superior factor returns over a forthcoming time period, with the goal of
monetizing those views in the event they materialize. Any such predictions may be
naturally translated into views on factors or combinations of factors.
In the following example, we simulate a hypothetical portfolio manager who pos-
sesses definite skill at factor timing. We then illustrate how to use the factor-based
BLB to construct a sequence of portfolios that successfully monetizes the portfolio
manager’s factor-timing views.
The views are expressed through (1) the k-dimensional vector q representing the
expected returns on the respective factors and (2) the k × k matrix W that represents
the uncertainties in those views. Specifically, we take W to be diagonal, where Ω ii = ω 2i
 122 | Factor Investing with Black–Litterman–Bayes: Incorporating Factor Views and Priors Quantitative Special Issue 2021

EXHIBIT 1 and wi can be thought of as the width of a confidence

Cumulative Factor Return Performance interval around qi, for each i = 1, …, k.
To simulate a manager who is skilled at factor tim-
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

ing, we assume that, at the very beginning of our his-

torical data, this portfolio manager made a single call
0.2 regarding the expected returns and uncertainties on a
handful of factors and has no views on the remaining
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

factors. In particular, the portfolio manager chooses q

0.0 and W at the beginning of 2007 and remains steadfast
in those views until the present time. With the benefit
of hindsight, we simulate the performance of this hypo-
–0.2 thetical portfolio manager’s strategy, in which some of
the factor bets turned out to be correct. We strongly
stress that this study is meant to simulate what hap-
ShortInterest
–0.4
pens in the event the portfolio manager made the right
Size
call at the beginning, as a way of illustrating the effi-
Liquidity
cacy of the BLB portfolio construction with factors. In
Quality
this article, we neither simulate nor provide guidance
2008 2010 2012 2014 2016 2018 2020 on the mechanism of factor timing itself.
Specifically, we assume that the portfolio manager
NOTES: Cumulative factor return performance for the four
factors that we selected for inclusion in the prescient views
decided, at the beginning of the sample period (January
over the sample period, 2007–2020. The y-axis is the cumu- 2007), to harvest the size (the small-cap effect) and
ˆ
lative sum of ft for all dates t between January 1, 2007 and liquidity premiums. In addition, the portfolio manager
November 1, 2020. decided to assign negative premiums to firms with high
short interest and a positive premium to a quality factor
based on accounting metrics. The directions of these
four factor bets align with well-known results from behavioral finance. In Exhibit 1,
we depict the cumulative factor return performance of these four factors over the
2007–2020 sample period. Furthermore, we assume that the portfolio manager was
prescient in deciding to take no explicit view on the well-known and much-studied value
and momentum factors, neither of which performed well in the US equity market over
the sample period of 2007–2020. We refer to this portfolio manager’s views as the
prescient views because they benefit from having taken no explicit view on value and
momentum (within the US equity market) over the period.
The procedure for our empirical study is as follows:

1. For each day t, construct the matrix of factor loadings Xt by loading the data
as a pandas data frame, transforming the loadings, and applying the function
patsy.dmatrix with the appropriate model formula. The transformation entails
reshaping the raw loading (e.g., market cap) to be Gaussian and centered
within the estimation universe.
2. For each day t, estimate the factor returns f t,j using ordinary least squares as
ˆft = (X t′X t )−1 X t′rt +1 .
3. Choose an in-sample period for the factor covariance matrix. We choose
2007–2015 to be in sample with respect to the factor covariance, F.
4. Estimate the factor covariance F as the sample covariance of ˆft with t
restricted to the in-sample period. We separate the calculation of the factor
covariance into volatility estimation and correlation estimation. The factor
covariance is given by F = SRS, where S is a diagonal matrix containing the
factor volatilities, and R is the factor correlation matrix. Note that R was esti-
mated in Python by applying the method numpy.corrcoef to the coefficients
 Quantitative Special Issue 2021 The Journal of Portfolio Management | 123

ˆft estimated in a previous step. We mitigate the effect of estimation error in

F-1 by applying a lower bound of 0.01 to the factor volatilities S.
For a market-neutral study, we set our prior to x = 0, and V = SRsS, where Rs
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

5.
denotes the Ledoit–Wolf shrinkage estimator for the correlation matrix (see
Ledoit and Wolf 2004). Under this prior, an all-cash benchmark portfolio is a
priori optimal. Practitioners may wish to choose a prior under which a different
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

benchmark, such as the S&P 500, is optimal; details of this procedure are
given in the discussion of Equation 29.
6. We set q and W to be consistent with the prescient views outlined in the pre-
ceding. Specifically, we specify one view per factor. For all industry and style
factors not explicitly represented in the prescient views, we set qi = 0 and
wi = 0.05, representing no view on their expected return, with a high level of
uncertainty. For the factors involved in the prescient views, we set qi equal
to -3 bps for size, -1 bp for liquidity, -0.5 bp for short interest, and +1 bp
for quality. We also set a premium of +1 bp for the intercept factor.
7. For each t, produce forward-looking estimates of Dt using historical residual
vols blended with implied vols.
8. For each t, compute Z t := Dt−1 X t without expanding Dt−1 as a full n × n matrix
by multiplying each ith row of Xt by Dii−1 . Note that Zt has the same dimensions
as Xt; both are n × k matrixes. Hence, as discussed previously, Zt is easier
to compute with than an n × n matrix.
9. For each t, define the intermediate calculation result Kt := Σ t−1 X t and compute

Kt = Z t − Z t [F −1 + X t′Z t ]−1 Z t X t
µ f = (V −1 + Ω −1 + X t′Kt )−1 (V −1 ξ + Ω −1q)
h*t = λ −1Kt µ f

10. Finally, compute the portfolio return rp,t +1 = (h∗t )′rt +1 .

Although the BLB method for factor models as presented here is not much more
complicated than a standard MVO approach to the same problem, we believe that
any added complexity should be justified through added performance. Therefore, to
gauge the value-add of the portfolio construction technique, we also include results
for an MVO baseline, which is proportional to Σ t−1 µ t where µ t = Xtq. Note that the
baseline includes q and hence also benefits from the prescient views, but it does
not benefit from the additional Bayesian portfolio construction of BLB. Of course, for
the baseline computation, we use Equation 16; hence, our method does not entail
the (notoriously unstable) security-level sample covariance matrix.
In Exhibit 2, we show a comparison between the BLB and MVO approaches. We
have scaled the two portfolios so that they realize the same volatility over the sample
period. The BLB portfolio sequence achieves a Sharpe ratio of 1.16 whereas the MVO
portfolio achieves about 0.9. There are several reasons for the difference, but perhaps
the most important is the presence of W in the BLB portfolio. Furthermore, the BLB
portfolio has a higher gross market value but achieves lower exposure to factors on
which the manager has no views, in part because of the shrinkage toward V-1x. In
any case, this illustrates that the BLB method provides the manager with additional
degrees of freedom that have the potential to reduce risk relative to MVO portfolios
that also benefit from the same prescient factor bets.
The average turnover of the two strategies we consider is comparable: Each turns
over an average of about 10%–15% of gross market value per day. Such high turn-
over is unsurprising because the portfolio formation process is unaware of costs.
 124 | Factor Investing with Black–Litterman–Bayes: Incorporating Factor Views and Priors Quantitative Special Issue 2021

EXHIBIT 2 The turnover is mostly driven by changes in the factor

Comparison of BLB and MVO Simulations exposures themselves and changes in the residual vari-
ance forecasts. This is likely too much turnover to be
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

1e7 directly implementable, but the transaction cost–aware

50 optimization that we discuss in the conclusions section
blb
mvo would be implementable.
Although we shall not do so here, it is possible to
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

4 run this procedure with a benchmark that is approxi-

mately the market-capitalization-weighted benchmark
Cumulative Wealth

with holdings vector hB. This is achieved by choosing a

3
normal prior N(x, V) having the property that the bench-
mark portfolio is a priori optimal, meaning that the
2 benchmark is optimal under the expected returns and
covariance matrix described by Equations 20 and 21.
Mathematically, referring to Equation 22, this
1 amounts to solving for x in the equation

λΣ thB = X tH−1 V −1 ξ (29)

0

2008 2010 2012 2014 2016 2018 2020 Solving for x in this way might not be possible exactly,
in which case one may solve for it in the least-squares
NOTES: Cumulative wealth from simulations of the MVO (dashes)
and BLB (solid) multifactor portfolios with prescient factor views
sense.
and data-driven priors. The portfolios are scaled such that they
realize the same volatility over the sample period, 2007–2020.
CONCLUSIONS

In this article, we proposed a general framework

referred to as BLB for constructing optimal portfolios for factor-based investing.
In the spirit of the classical Black–Litterman model, this framework allows for the
incorporation of investor views and different priors on factor risk premiums, including
data-driven and benchmark priors. We provided computationally efficient closed-form
formulas for the (posterior) expected returns and return covariance matrix that result
from integrating factor views into an APT multifactor model.
Finally, in a step-by-step procedure, we showed how to build the prior and incorpo-
rate the factor views, demonstrating in a realistic empirical example using a number
of well-known cross-sectional US equity factors that the BLB approach can add value
to mean–variance-optimal multifactor risk premium portfolios.
In closing, we make a few observations regarding application of these results in
transaction cost–aware MVO. The original BL model and BLB extension with factor
views that we have presented here share in common that the portfolio construction
step is a standard mean–variance optimization. The BL and BLB methods are based
on adjusting the posterior security return density (where only the first two posterior
moments are needed, in the case of elliptical distributions). In particular, we denote
by h0 the starting portfolio and c(h0, h) the anticipated cost (predicted before trading
begins) to trade from h0 → h. For institutional asset managers, c(h0, h) is typically
dominated by slippage costs due to temporary and permanent impact. Nevertheless,
the optimization problem for BLB with costs is

λ
max h′ µ BL-fac − h′ Σ BL-fach − c(h0 , h)
h 2

where notation refers to Equations 24 and 25.

 Quantitative Special Issue 2021 The Journal of Portfolio Management | 125

REFERENCES
Aliaga-Diaz, R., G. Renzi-Ricci, A. Daga, and H. Ahluwalia. 2020. “Portfolio Optimization with
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

Active, Passive, and Factors: Removing the Ad Hoc Step.” The Journal of Portfolio Management
46 (4): 39–51.

Ang, A. Asset Management: A Systematic Approach to Factor Investing. Oxford, UK: Oxford University
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

Press, 2014.

Asness, C. S., T. J. Moskowitz, and L. H. Pedersen. 2013. “Value and Momentum Everywhere.”
The Journal of Finance 68 (3): 929–985.

Avramov, D., and G. Zhou. 2010. “Bayesian Portfolio Analysis.” Annual Review of Financial Eco-
nomics 2 (1): 25–47.

Baltussen, G., L. Swinkels, and P. Van Vliet. 2019. “Global Factor Premiums.” Technical report.

Bass, R., S. Gladstone, and A. Ang. 2017. “Total Portfolio Factor, Not Just Asset, Allocation.”
The Journal of Portfolio Management 43 (5): 38–53.

Bender, J., J. Le Sun, and R. Thomas. 2018. “Asset Allocation vs. Factor Allocation—Can We Build
a Unified Method?” The Journal of Portfolio Management 45 (2): 9–22.

Bergeron, A., M. Kritzman, and G. Sivitsky. 2018. “Asset Allocation and Factor Investing: An Inte-
grated Approach.” The Journal of Portfolio Management 44 (4): 32–38.

Black, F., and R. B. Litterman. 1990. “Asset Allocation: Combining Investor Views with Market
Equilibrium.” Technical report.

——. 1991a. “Asset Allocation: Combining Investor Views with Market Equilibrium.” The Journal
of Fixed Income 1 (2): 7–18.

——. 1991b. “Global Asset Allocation with Equities, Bonds and Currencies.” Technical report.

——. 1992. “Global Portfolio Optimization.” Financial Analysts Journal 48 (5): 28–43.

Carhart, M. M. 1997. “On Persistence in Mutual Fund Performance.” The Journal of Finance
52 (1): 57–82.

Chamberlain, G. 1983. “A Characterization of the Distributions That Imply Mean–Variance Utility

Functions.” Journal of Economic Theory 29 (1): 185–201.

Dopfel, F. E., and A. Lester. 2018. “Optimal Blending of Smart Beta and Multifactor Portfolios.”
The Journal of Portfolio Management 44 (4): 93–105.

Fabozzi, F. J., S. M. Focardi, and P. N. Kolm. 2006. “Incorporating Trading Strategies in the Black–
Litterman Framework.” The Journal of Trading 1 (2): 28–37.

——. Quantitative Equity Investing: Techniques and Strategies. Hoboken: John Wiley & Sons, 2010.

Fabozzi, F. J., P. N. Kolm, D, A. Pachamanova, and S. M. Focardi. Robust Portfolio Optimization and
Management. Hoboken: John Wiley & Sons, 2007.

Fama, E. F., and K. R. French. 1993. “Common Risk Factors in the Returns on Stocks and Bonds.”
Journal of Financial Economics 33 (1): 3–56.

Figelman, I. 2017. “Black–Litterman with a Factor Structure Applied to Multi-Asset Portfolios.”

The Journal of Portfolio Management 44 (2): 136–155.

Frazzini, A., and L. H. Pedersen. 2014. “Betting against Beta.” Journal of Financial Economics
111 (1): 1–25.
 126 | Factor Investing with Black–Litterman–Bayes: Incorporating Factor Views and Priors Quantitative Special Issue 2021

Gelman, A., J. B. Carlin, H. S. Stern, and D. B. Rubin. Bayesian Data Analysis. 2nd ed. Taylor &
Francis, 2003.
It is illegal to make unauthorized copies, forward to an unauthorized user, post electronically, or store on shared cloud or hard drive without Publisher permission.

Johnson, B. 2018. “Strategic Beta ETF Market Still Expanding.” https://www.morningstar.co.uk/

uk/news/182188/strategic-beta-etf-market-still-expanding.aspx.

Jones, R. C., T. Lim, and P. J. Zangari. 2007. “The Black–Litterman Model for Structured Equity
Portfolios.” The Journal of Portfolio Management 33 (2): 24–33.
Downloaded from https://jpm.pm-research.com/content/47/2, by guest on November 19, 2021. Copyright 2020 With Intelligence Ltd.

Keloharju, M., J. T. Linnainmaa, and P. Nyberg. 2016. “Return Seasonalities.” The Journal of Finance
71 (4): 1557–1590.

Koijen, R. S. J., T. J. Moskowitz, L. H. Pedersen, and E. B. Vrugt. 2018. “Carry.” Journal of Financial
Economics 127 (2): 197–225.

Kolm, P. N., and G. Ritter. 2017. “On the Bayesian Interpretation of Black–Litterman.” European
Journal of Operational Research 258 (2): 564–572.

Ledoit, O., and M. Wolf. 2004. “Honey, I Shrunk the Sample Covariance Matrix.” The Journal of
Portfolio Management 30 (4): 110–119.

Litterman, R., and G. He. 1999. “The Intuition behind Black–Litterman Model Portfolios.” Technical
report.

Melas, D., Z. Nagy, N. Kumar, and P. Zangari. 2019. “Integrating Factors in Market Indexes and
Active Portfolios.” The Journal of Portfolio Management 45 (6): 16–29.

Meucci, A. Risk and Asset Allocation. New York: Springer Science & Business Media, 2009.

Moskowitz, T. J., Y. H. Ooi, and L. H. Pedersen. 2012. “Time Series Momentum.” Journal of Finan-
cial Economics 104 (2): 228–250.

Pearl, J. Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference. San Fran-
cisco: Morgan Kaufmann, 2014.

Ritter, G. 2016. “Stable Linear-Time Optimization in Arbitrage Pricing Theory Models.” Risk Maga-
zine, https://www.risk.net/asset-management/2469223/stable-linear-time-optimisation-in-arbi-
trage-pricing-theory-models.

Roll, R., and S. A. Ross. 1980. “An Empirical Investigation of the Arbitrage Pricing Theory.”
The Journal of Finance 35 (5): 1073–1103.

Ross, S. A. 1976. “The Arbitrage Theory of Capital Asset Pricing.” Journal of Economic Theory 13
(3): 341–360.

Rubesam, A., and S. Hwang. 2019. “Do Smart Beta ETFs Capture Factor Premiums? A Bayesian
Perspective.” Working paper series 2018-ACF-04, IÉSEG.

Satchell, S., and A. Scowcroft. 2000. “A Demystification of the Black–Litterman Model: Managing
Quantitative and Traditional Portfolio Construction.” Journal of Asset Management 1 (2): 138–150.

Subrahmanyam, A. 2010. “The Cross-Section of Expected Stock Returns: What Have We Learnt
from the Past Twenty-Five Years of Research?” European Financial Management 16 (1): 27–42.

Zhou, G. 2009. “Beyond Black–Litterman: Letting the Data Speak.” The Journal of Portfolio Man-
agement 36 (1): 36–45.

To order reprints of this article, please contact David Rowe at [email protected]

or 646-891-2157.