Value premium and macroeconomic variables

• Published date: 01 September 2023
• Author: Elena Beccalli,Nicola Doninelli,Cesare Orsini
• Date: 01 September 2023
• DOI: http://doi.org/10.1111/eufm.12397

DOI: 10.1111/eufm.12397

ORIGINAL ARTICLE

Value premium and macroeconomic variables

Elena Beccalli

1

|Nicola Doninelli

1

|Cesare Orsini

2

1

School of Banking, Finance, and

Insurance, Catholic University of the

Sacred Heart of Milan, Milan, Italy

2

Epsilon Associati Sgr S.p.A., Milan, Italy

Correspondence

Elena Beccalli, Catholic University of the

Sacred Heart of Milan, Largo Gemelli 1,

Milan 20123, Italy.

Email: elena.beccalli@unicatt.it

Abstract

This paper investigates the effect of macroeconomic

expectations on the value premium. We introduce a

two‐pass estimation procedure to extrapolate the

impact of investors' macroexpectations on the firm

fundamental value of Rhodes‐Kropf, Robinson, and

Viswanathan. We find that the level and slope of the

term structure affect valuation, revealing a heavily

industry‐dependent effect. The portfolios sorted on

metrics orthogonal to macroeconomic variables

show a clear association between the misvaluation

component of value premium and size risk. By

removing the influence of the macroeconomic

conditions and size, we separate the portion of the

value premium that rewards macroeconomic

expectations.

KEYWORDS

macroeconomic risk, market‐to‐book decomposition, value

premium

JEL CLASSIFICATION

G12, G14

Eur Financ Manag. 2023;29:1336–1374.1336

|

wileyonlinelibrary.com/journal/eufm

EUROPEAN

FINANCIAL MANAGEMENT

This is an open access article under the terms of the Creative Commons Attribution‐NonCommercial‐NoDerivs License, which permits

use and distribution in any medium, provided the original work is properly cited, the use is non‐commercial and no modifications or

adaptations are made.

© 2022 The Authors. European Financial Management published by John Wiley & Sons Ltd.

We thank anonymous referees, Ettore Croci, Marzia De Donno, Giorgio Di Giorgio, Massimo Guidolin, Roberto

Savona, Andrea Tarelli, and seminar participants at the European Financial Management Association 2021 Conference

for helpful comments and suggestions.

1|INTRODUCTION

Value investors typically buy low price‐to‐book stocks and sell high price‐to‐book securities to

harvest abnormal returns known as the value premium. Early attempts to understand the effect

of macroeconomic conditions on the value premium have been quite unsuccessful (Lakonishok

et al., 1994). More recently, Bergbrant and Kelly (2016) suggest little role for aggregate

macroeconomic risks in explaining the returns to size, value, and momentum factors. Other

studies use macroeconomic factors in asset pricing models but do not focus specifically on the

value premium. According to Flannery and Protopapadakis (2002), macroeconomic indexes are

excellent candidates to represent extra‐market risk factors since they simultaneously affect cash

flows and influence the risk‐adjusted discount rate. For example, Chen et al. (2004) use

industrial production and unexpected inflation in a celebrated five‐factor model and Cochrane

(1996) uses aggregate investment growth in his factor pricing model for stock returns. Petkova

(2006) studies the connection between the Fama–French factors and innovations in state

variables (such as the default spread, the market dividend‐price ratio, the yield spread, and the

1‐month Treasury bill rate) and finds that the average stock returns line up nicely against factor

betas. Bianchi et al. (2017) recently show that macroeconomic shocks have considerable effects

on the cross‐section of US stock returns when risk exposures and idiosyncratic risk are time‐

varying.

In this study, we aim to investigate the relation between the value premium and a set of

macroeconomic variables that the literature on asset pricing acknowledges to have an

information content that plays a crucial role in influencing expectations on macroeconomic

risk. To this end, we use Rhodes‐Kropf, Robinson, and Viswanathan's (2005; RKRV) market‐to‐

book decomposition that they introduced to study the timing of merger waves. This

decomposition breaks the market‐to‐book ratio into market‐to‐value and value‐to‐book

components in which value is defined as a multiple‐based estimate of the fundamental value

of equity. As far as we know, Chang et al. (2013) and Golubov and Konstantinidi (2019; GK) are

the first to apply the RKRV decomposition to an asset pricing context. In particular, GK show

that most of the premium is attributable to the market‐to‐value component while the cross‐

sectional loadings on value‐to‐book are not significantly different from zero. Moreover, they

find that different exposures to cash‐flow risk, long‐run consumption risk, investment‐specific

technology shocks, operating leverage, and duration are not related to the cross‐sectional

distribution of market‐to‐values and provide evidence that supports common behavioral

explanations of the price‐to‐book anomaly. Using a similar approach, Jaffe et al. (2020) report

that after controlling for mispricing error, the RKRV mispricing component predicts abnormal

short‐term and medium‐term returns and that value investing no longer beats the growth

approach. Chang et al. (2013) examine the association of a misvalutaion factor (

M

SV

F

) with

common stock characteristics and empirically test the relation between

M

SV

F

and future

returns.

1

Hahn and Lee (2006) show that changes in the Default Spread and changes in the

term spread capture most of the systematic risks proxied by Fama and French's (1993) size

(SMB) and book‐to‐market (HML) factors.

We consider our investigation in many ways as complementary to the above studies. First,

our paper aims to assess the impact of macrovariables on each individual sector by

1

Chang et al. (2013) directly focus on the RKRV firm‐level misvaluation and test the power of

M

SVFto predict future

macroeconomic conditions.

BECCALLI ET AL.EUROPEAN

FINANCIAL MANAGEMENT

|

1337

extrapolating in a bottom‐up way the impact of the expectations contained in the

macroeconomic variables on the value premium. Prior literature aims to provide a

macroeconomic explanation of the value premium by focusing on a particular characteristic

of value stocks (e.g., leverage, beta, etc.). The resulting empirical evidence often highlights

consistency problems since macroeconomic expectations exhibit a complex and variable way of

affecting returns of value stocks (Maloney & Moskowitz, 2021). A representative case of this

lack of consistency is, in contradiction with the leverage interpretation of value premium, the

poor performance of value stocks during the period 2009–2016 despite easing financial

conditions. The lack of evidence on the connection between macroeconomic expectations and

value premium suggests that such investigation cannot be grounded on a single characteristic.

On the contrary, the relationship that links an economic sector to the performance of a

macroeconomic variable tends to be more stable and subject to more sporadic shocks related to

structural breaks in the economic activity (e.g., the advent of information technology). In short,

while focusing on a single fundamental characteristic leads to problems of inconsistency, our

analysis grasps more stable relationships by taking into account sectoral heterogeneity.

Second, provided that macroeconomic conditions hit stock prices differently depending on

their economic sector, dissecting this heterogeneous effect helps explaining the mispricing

component of the value premium. Moreover, market analysts normally estimate fundamental

multiples on a sector basis; thus making a sector‐based investigation very relevant also from a

market practice point of view. By adopting a sector‐based approach, we depart from the extant

market‐wide misvaluation analysis cited above by using a distinct approach that provides

further evidence on the effect of macroeconomic conditions on each sector's fundamental

multiple and their contribution to the market value.

Third, our paper advances Hahn and Lee (2006) by extending the range of factors explaining

the sensitivity of value stocks to fluctuations in the business cycle. While Hahn and Lee (2006)

just focus on the high level of debt as the primary explanation for such a sensitivity, our

approach enables us to examine several firms' fundamental multiples and provides evidence on

the significant influence of macroeconomic variables on net income and book value multiples.

The need to move beyond the leveraged capital structure derives from the steady fall in the cost

of debt experimented over the last decade: persistent low interest rates weaken the negative

leverage's effect on the firm's market value.

Our results present relevant asset pricing implications. Using monthly data from 1975 to

2016, we investigate size, risk, and the returns of portfolios sorted in the market‐to‐book

components. Consistent with GK, we find that the return difference between the low and high

market‐to‐book portfolios almost entirely results from the firm‐specific error component.

Furthermore, to investigate the effect of macroeconomic conditions on intrinsic values

within each sector, we use the fundamental multiples suggested by RKRV to directly dissect the

sensitivity of these measures to the macroeconomic environment. Given that these multiples

incorporate predictions on both growth and discount rates, their time‐varying estimates should

reflect the expectations on the macroeconomic scenario. To explore this connection in a

regression setting, we first pick a set of leading macroeconomic variables that affect the

expectational, nonfundamental component of stock prices: the term spread, the 10‐year US

Treasury yield, the ISM Manufacturing Purchasing Managers Index, and the Conference Board

Leading Economic Index. These four variables are widely used in time series macroeconomic

factor models, and very popular among practitioners as broad economic cycle indicators. We

find that the slope of the term structure and the 10‐year Treasury yield significantly influence

the fundamental multiples with a consequential effect on the assessment of intrinsic value. Our

1338

|

EUROPEAN

FINANCIAL MANAGEMENT

BECCALLI ET AL.