Understanding the Performanceof the

Equity Value Factor∗

理解股权价值系数 * 的表现
Lauren Stagnol

劳伦 · 斯塔诺尔
Quantitative ResearchStrategic Investment Advisor Amundi Asset
Management, ParisCPR Asset Management, Paris
lauren.stagnol@amundi.comchristian.lopez@cpr-am.com Thierry
RoncalliBruno Taillardat Quantitative ResearchSmart Beta Team Amundi Asset
Management, ParisAmundi Asset Management, Paris
thierry.roncalli@amundi.combruno.taillardat@amundi.com
定量研究战略投资顾问 Amundi 资产管理，ParisCPR 资产管理，Paris
lauren.stagnol@Amundi. com christian.lopez@cpr-am. com Thierry
RoncalliBruno Taillardat Quantitative ResearchSmart Beta Team Amundi 资产
管理，ParisAmundi 资产管理，Paris thiery.roncalli@Amundi. com
bruno.Taillardat@Amundi。网址:
Christian Lopez
克里斯蒂安 · 洛佩兹

February 2021

二零二一年二月

Abstract

摘要

After decades of sound performance, doubts have been raised on the ability of the equity value
factor to continue to deliver a positive performance in the aftermath of the 2008 Global
Financial Crisis. Indeed, in a context dominated by low yields, sluggish growth and subdued
inflation combined with an accelerating digitalization of the econ-omy, the performance of
value strategies struggled over the past decade. In this paper,we investigate potential drivers
behind this performance lag, such as macroeconomic and microeconomic determinants, ESG
characteristics or credit-borrowed components.Based on European and American data, we find
that inflation and tightening credit spread levels are the most supportive factors for value
stocks. As far as interest rates are concerned, their sustained low levels prevented the value
stock universe from clear-ing its most distressed issuers, also known as “deep value”, and thus
dampened value performance. As a matter of fact, we show that value has not been
systematically an investment strategy bearing a heightened default risk. Our ESG analysis
corroborates the “transatlantic divide”, the historical gap between the U.S. and Europe on this
front,and shows that value and growth stocks are not necessarily all brown and green stocks.In
addition, we demonstrate that the small cap segment has not been the magical cure to value
underperformance. Finally, we conclude that value is not dead yet, and might even have bright
days ahead considering the current improvements in market sentiment,especially if inflation
does materialize. Nevertheless, we also emphasize that the current value risk factor is probably
different in nature from the one we observed during the golden age of value investing at the
beginning of the 2000s. Indeed, trading facilities,ease of access to fundamental data for a large
number of investors, ESG investing and the digitalization of the economy may have changed
the rules of the game.

经过几十年良好的业绩之后，人们开始怀疑股本价值因素是否有能力在 2008 年全球金融危机之后继

续取得积极的业绩。实际上，在低收益率、增长乏力、通胀低迷以及经济数字化加速的背景下，价值
战略的表现在过去十年里举步维艰。在本文中，我们调查了这种表现滞后背后的潜在驱动因素，如宏
观经济和微观经济决定因素、 ESG 特征或信贷借贷成分。基于欧洲和美国的数据，我们发现通货膨胀
和紧缩的信贷利差水平是最有利于价值股票的因素。就利率而言，其持续的低水平阻碍了价值型股票
领域清算其最困难的发行者，也就是所谓的”深度价值”，从而抑制了价值业绩。事实上，我们表明，
价值并不是一个系统性的投资策略，承担着更高的违约风险。我们的 ESG 分析证实了“跨大西洋鸿
沟”，即美国和欧洲在这方面的历史差距，并表明价值型和增长型股票并不一定都是棕色和绿色的。
此外，我们还证明，小盘股并不是治疗价值表现不佳的灵丹妙药。最后，我们得出的结论是，价值尚
未消亡，考虑到目前市场情绪的改善，特别是如果通货膨胀成为现实，甚至可能有光明的未来。然而，
我们也强调，目前的价值风险因素在性质上可能不同于我们在 2000 年代初价值投资黄金时代所观察
到的因素。实际上，交易设施、大量投资者获取基本数据的便利、环境、社会和治理投资以及经济数
字化可能已经改变了游戏规则。

Keywords: Value, risk factor, risk premium, factor investing, valuation, deep value,
infla-tion, interest rates, ESG, carbon risk.

关键词: 价值，风险因素，风险溢价，要素投资，估值，深度价值，通货膨胀，利率，
ESG，碳风险。

JEL classification: C50, G11, G12.

JEL 分类: C50，G11，G12。

∗We would like to thank Fr´ed´eric Lepetit, Alessandro Russo and Takaya Sekine from
Amundi for their helpful comments as well as Anastasiya Semenova, Maxime George,
Tony Hong and Steve Kidd from Moody’s Analytics CreditEdgeTM for their support on
credit data.

* 我们要感谢 Amundi 的 eric Lepetit、 Alessandro Russo 和 Takaya Sekine 神父的有益

评论，以及穆迪分析的 Anastasiya Semenova、 Maxime George、 Tony Hong 和 Steve
Kidd 对信贷数据的支持。

1Introduction
1 引言

First advocated by Graham and Dodd(1934), formalized later by Basu(1983) with his
work on the empirical relationship between price-to-earnings ratios and stock
returns, and then by Fama and French(1992) with their research on book-to-price in
the 1990s, the popularity of the value factor has been increasing ever since among
both academics and practitioners. Value investors buy securities that trade below
their own perceived “intrinsic value”. The differential between an investor’s opinion
on the firm’s intrinsic value and the market price usually resides in diverging views on
the company’s ability to generate cash-flows in the future. Metrics commonly used to
make such an assessment can be divided into two approaches:(1) comparing the
fundamentals to market value such as book-to-market, price-to-earnings or price-to-
cash-flows(Fama and French, 1992, 1993), or(2) more standalone measures such as
dividend yield, dividend growth, operating cash-flows, earnings growth, cash flow
yield(Chan et al., 1991). In this framework, a value investor then bets on the future
outperformance of such stocks.
最初由 Graham 和 Dodd (1934)提倡，后来由 Basu (1983)通过他关于市盈率和股票回报之
间的经验关系的工作正式确定，然后由 Fama 和 French (1992)通过他们在 1990 年代对账
面价值的研究，自那以后，价值因素在学者和从业者中的受欢迎程度一直在增加。价值投资
者购买的证券交易价格低于他们认为的“内在价值”。投资者对公司内在价值的看法与市场
价格之间的差异，通常在于对公司未来产生现金流能力的不同看法。通常用来进行这种评估
的衡量标准可分为两种方法: (1)比较基本面与市值之间的关系，例如账面价值、市盈率或市
盈率(Fama 和 French，1992,1993) ; 或(2)比较独立的衡量标准，例如股息收益率、股息增
长、经营现金流量、盈利增长、现金流量收益率(Chan 等，1991)。在这个框架下，价值投
资者然后押注于这些股票未来的优异表现。
Different explanations compete to elucidate the value premium or the value
anomaly.The first explanation is behavioral, and postulates that investors may tend to
believe that past winners will remain ahead of the losers. Investors would tend to
over-react and would make errors when forecasting future growth path, undermining
value stocks’ earnings paths and inflating those of “glamour ” stocks, which in turn
drives the gap between value and growth performance. Therefore, according to this
theory, value investors would simply be contrarian(Lakonishok et al., 1994). The
second explanation is rather risk-based and lies in the idea that value firms would be
fundamentally riskier. Hence an investor would require a higher premium, because
the riskiest firms being priced lower would have a lower price-to-book ratio, and
would therefore appear more appealing in terms of valuation(Fama and French,
1998). In this approach, value is a skewness risk premium, for which an investor
should be rewarded for bearing default risk.

不同的解释竞相解释价值溢价或价值异常。第一种解释是行为的，并假设投资者可能倾向于
相信过去的赢家将继续领先于失败者。投资者在预测未来增长路径时往往会反应过度，犯错
误，破坏价值型股票的盈利路径，夸大“魅力型”股票的盈利路径，进而拉大价值与增长表
现之间的差距。因此，根据这个理论，价值投资者只会是逆向投资者(Lakonishok 等，
1994)。第二种解释是相当基于风险的，在于价值型公司的基本风险更高。因此，投资者需
要较高的溢价，因为定价较低的风险最大的公司的市净率较低，因此在估值方面似乎更具吸
引力(Fama 和 French，1998)。在这种方法中，价值是一种偏态风险溢价，投资者应该因为
承担违约风险而得到回报。

Nevertheless, the value premium has been under close scrutiny in the past decade.
With sustained underperformance since the 2008 Global Financial Crisis(GFC), it has
been thrown under the bus by an increasing number of academics and practitioners.
Even Eugene Fama and Kenneth French, its famous proponents, attempted to verify if
the value premium still exists(Fama and French, 2020). Such questioning is legitimate,
because the value factor (as defined by Fama and French) lost 32% in the U.S. and
25% in Europe between December 2019 and September 2020. And this lousy
performance was the straw that has broken the camel’s back. Indeed, cumulative
performance has exhibited a severe downward trend since the GFC, dropping to
levels reached as far back as in 2001. Then one could wonder if value is dead yet.

然而，价值溢价在过去十年里一直受到密切关注。自 2008 年全球金融危机(Global

Financial Crisis，简称 GFC)以来，它一直表现不佳，已经被越来越多的学者和从业者抛弃。
即使是著名的支持者 Eugene Fama 和 Kenneth French，也试图验证价值溢价是否仍然存
在(Fama 和 French，2020)。这样的质疑是合理的，因为在 2019 年 12 月至 2020 年 9 月期
间，价值因素(由 Fama 和 French 定义)在美国损失了 32% ，在欧洲损失了 25% 。这种糟
糕的表现是压垮骆驼的最后一根稻草。事实上，自全球金融危机以来，累积的业绩已经呈现
出严重的下降趋势，跌至 2001 年的水平。那么人们可能会怀疑价值是否已经死亡。
The paper is organized as follows. In Section 2, we detail what we mean by value and
its statistical characteristics. We precisely define the differences between value
investing, value style, valuation, value risk factor and value risk premium. We also
touch on recent changes in the market micro-structure, and their possible impact on
the value premium. In Section 3, we focus on both its macroeconomic and
microeconomic determinants. We estimate the empirical relationships between the
value performance and the standard economic risk factors such as interest rates,
inflation and the credit environment. Moreover, we focus on the specific behavior of
deep value stocks. Section 4 highlights the market factors that could be held
responsible for the recent poor performance of value investing. Thus, we explore the
impact of ESG investing, carbon risk, small cap effects and intangible assets. Section
5offers some concluding remarks.

这篇论文的组织方式如下。在第二部分，我们详细说明了我们所说的价值及其统计特征。我
们精确地定义了价值投资、价值风格、价值评估、价值风险因素和价值风险溢价之间的区别。
我们也触及了最近市场微观结构的变化，以及它们可能对价值溢价的影响。在第三部分，我
们关注宏观经济和微观经济的决定因素。我们估计价值表现和标准经济风险因素(如利率、
通货膨胀和信贷环境)之间的经验关系。此外，我们关注深价值股票的具体行为。第四部分
强调了可能导致近期价值投资表现不佳的市场因素。因此，我们探讨了 ESG 投资、碳风险、
小盘股效应和无形资产的影响。第五部分提供了一些结束语。

2What does value mean?

价值是什么意思？

2.1Interconnectedness between value investing, valuation and the value risk

factor

2.1 价值投资、估值和价值风险因素之间的相互关系

In this paper, we want to grasp the drivers behind the recent underperformance of
the value risk factor per-se. Improving our understanding may teach a lesson going
forward. To achieve this goal, we must nevertheless define precisely what we would
like to measure. This is because value covers many concepts that are related and
different. For instance, value investors encompass asset managers and asset owners
that believe in value investing. This means that they apply a stock picking style, where
they compare the fundamental value of the stock with its market value. Therefore,
value investors should buy stocks that are undervalued and sell stocks that are
overvalued. This definition reflects the raison d’ˆetre of active management. Indeed,
active managers think that the market is not efficient. They then buy some stocks
because they hope that their prices will go up. In this case, we notice that value
investing and active management are highly related.
在这篇文章中，我们想要抓住价值风险因素本身最近表现不佳背后的驱动因素。提高我们的
认识可能会给我们上一课。为了实现这个目标，我们必须精确地定义我们想要衡量的东西。
这是因为价值涵盖了许多相关和不同的概念。例如，价值投资者包括相信价值投资的资产管
理者和资产所有者。这意味着他们应用一种选股风格，比较股票的基本价值和市场价值。因
此，价值投资者应该买入被低估的股票，卖出被高估的股票。这个定义反映了积极管理存在
的理由。事实上，主动型管理者认为市场效率不高。然后他们购买一些股票，因为他们希望
他们的价格会上涨。在这种情况下，我们注意到价值投资和积极管理是高度相关的。
However, the academic literature shows that the value style is not unique for building
a stock picking process. Indeed, there are other management styles, such as the low-
volatility,quality or growth styles. In particular, one investment style seems to attract
many investors and dominates the other. This is the momentum style, which has
been exposed by Jegadeesh and Titman(1993). According to Grinblatt at al.(1995), the
momentum style far dominates the value style, since more than 70% of equity
mutual funds are in fact momentum investors.Nevertheless, when we ask investors if
they are value or momentum investors, most of them would pick the first answer.
This is normal since value investing is more rational than momentum investing.
Nobody wants to invest in stocks that are overvalued.
然而，学术文献表明，价值风格并不是建立一个股票选择过程的唯一方式。事实上，还有其
他的管理风格，比如低波动、高质量或增长风格。特别是，一种投资风格似乎吸引了很多投
资者，并主导了另一种。这就是动量风格，已经被 Jegadeesh 和 Titman (1993)揭露出来。
根据 Grinblatt at al。(1995)的研究，动量型基金远远主导了价值型基金，因为超过 70% 的
股票共同基金实际上是动量型投资者。然而，当我们问投资者他们是价值投资者还是动量投
资者时，他们中的大多数会选择第一个答案。这很正常，因为价值投资比动量投资更理性。
没有人愿意投资估值过高的股票。
The concept of valuation is then central when we refer to value. Valuation is also
exten-sively used by commodity investors, currency investors or bond investors to
understand if the asset is overvalued or undervalued by the market. Therefore, we
must think that value is everywhere, whereas in fact, this is not exactly true. When we
refer to value, we implic-itly refer to the stock market, because of the opposition
between value stocks and growth stocks. In other asset classes, we may define
undervalued and overvalued securities, but we do not define value and growth
assets. For instance, the concept of value/growth bonds or value/growth currencies
does not exist. Therefore, value is very specific to the stock market.And even though
value and valuation are related, they cover two different things. Given the
performance of value these last ten years, the question of the death of value
investing is justified. However, the question of the death of valuation is inappropriate
and irrelevant since financial markets and active management exist because of
valuation.
当我们提到价值时，估值的概念是核心。估值也被大宗商品投资者、外汇投资者或债券投资
者广泛用于了解资产是否被市场高估或低估。因此，我们必须认为价值无处不在，然而事实
上，这并不完全正确。当我们提到价值时，我们隐含地指的是股票市场，因为价值型股票和
成长型股票之间存在对立。在其他资产类别中，我们可以定义被低估和被高估的证券，但我
们不定义价值和增长型资产。例如，价值/增长债券或价值/增长货币的概念并不存在。因此，
价值对于股票市场来说是非常特殊的。即使价值和估值是相关的，它们涵盖了两个不同的东
西。考虑到过去十年的价值表现，价值投资的死亡问题是合理的。然而，由于估值导致了金
融市场和积极管理的存在，估值死亡的问题是不恰当和无关紧要的。
We may wonder why equity value investing is so unique or singular. We have partially
answered this question because of the value/growth opposition. Another piece of the
puzzle is the academic consensus of the value definition around the Fama-French
model. As we have already said in the introduction, we need a model or some metrics
to define a firm’s intrinsic or fair value. In the case of equity, the Fama-French model
based on the book-to-market statistic has become widely accepted as the benchmark
model. It has been accepted not only by academics, but also by professionals.
Furthermore, many investment portfolios are based on the Fama-French model or an
extended version of the Fama-French approach with several criteria and metrics. If
we consider corporate bonds for instance, the Merton model may also be seen as a
reference model for defining overvalued and undervalued bonds. Nevertheless,the
calibration of the Merton model is so difficult that it is not a widely accepted
benchmark model by professionals.

我们可能想知道为什么股票价值投资是如此独特或独特。我们已经部分回答了这个问题，因
为价值/增长的对立。另一个难题是围绕 Fama-French 模型的价值定义的学术共识。正如我
们在引言中已经说过的，我们需要一个模型或一些指标来定义一个公司的内在或公允价值。
在股权方面，基于账面到市场统计的 Fama-French 模型已经被广泛接受为基准模型。它不
仅被学术界所接受，也被专业人士所接受。此外，许多投资组合是基于 Fama-French 模型
或 Fama-French 方法的扩展版本，有几个标准和指标。如果我们以公司债券为例，默顿模
型也可以作为定义高估和低估债券的参考模型。然而，默顿模型的校准是如此困难，以至于
它不是一个被专业人士广泛接受的基准模型。
Among investment portfolios managed with the Fama-French value factor, factor
invest-

在使用 Fama-French 价值因素管理的投资组合中，因素投资 -

ing portfolios take a special place within value investing. The underlying idea of
factor investing is to assume that there is not only one common risk factor, but
several common risk factors that explain the systematic part of stock returns. Since
the specific part(or the alpha) of investment portfolios decreases with diversification,
it does make sense to directly build a diversified equity portfolio using these common
risk factors. Thus, the professional consensus considers that the common risk factors
are the traditional beta defined by the CAPM and alternative betas such as size, value,
momentum, low-volatility and quality. We notice that the concept of beta only makes
sense with respect to a risk factor. Therefore,value refers to the value risk factor in
factor investing.
投资组合在价值投资中占有特殊的地位。要素投资的基本思想是假设不仅有一个共同的风险
因素，而且有几个共同的风险因素来解释股票收益的系统部分。由于投资组合的特定部分
(或阿尔法)随着多样化而减少，因此利用这些共同的风险因素直接建立一个多样化的股票组
合是有意义的。因此，专业共识认为，共同的风险因素是传统的贝塔定义的资本资产定价模
型和替代贝塔，如规模，价值，动量，低波动性和质量。我们注意到 beta 的概念只对风险
因素有意义。因此，价值是指要素投资中的价值风险因素。

Again, we may wonder whether the issue is the death of the value risk factor,
meaning that the value risk factor is not significant when explaining the cross-section
of stock returns.This is not the case since the explanatory power of the value risk
factor continues to be relatively high. According to Drei et al.(2019), the value risk
factor is behind the quality risk factor in North America, but at the same level as the
momentum risk factor between 2014 and 2019. In the Eurozone, the value risk factor
has the largest explanatory power among the alternative betas.

同样，我们可能想知道问题是否是价值风险因素的死亡，这意味着价值风险因素在解释股票
收益的横截面时并不重要。但事实并非如此，因为价值风险因素的解释力仍然相对较高。根
据 Drei 等人(2019)的研究，在北美，价值风险因素位于质量风险因素之后，但与 2014 年至
2019 年间的动量风险因素处于同一水平。在欧元区，价值风险因素在其他贝塔中具有最大
的解释力。
Nevertheless, there is a clear difference between the traditional beta and the
alternative betas. The traditional beta corresponds to the long-only market portfolio,
whereas alter-native betas are defined with respect to long/short portfolios. The
market risk premium is then defined as the expected return of the market portfolio.
In order to define the value risk premium, the risk factor must be long on the value
stocks and short on the anti-value (or growth) stocks. While switching the long and
short exposures has no impact when con-sidering a risk factor analysis, the choice of
the long and short legs is crucial when defining the risk premium associated with an
alternative beta. When we speak about the value risk premium, we explicitly assume
that value stocks outperform growth(or anti-value) stocks.We face the same issue
with the size, momentum, low-volatility and quality risk premia.These choices are at
the core of the construction of factor investing portfolios. Therefore,the underlying
question does not concern the value style, the valuation approach, or the value risk
factor, but the equity value premium as used in a factor investing framework.
尽管如此，传统的测试版和替代的测试版之间还是有着明显的区别。传统的测试版对应于只
做多的市场投资组合，而另一种原生的测试版是根据多/空投资组合定义的。市场风险溢价
被定义为市场投资组合的预期回报。为了界定价值风险溢价，风险因素必须是价值型股票的
多头，反价值型(或增长型)股票的空头。尽管在考虑风险因素分析时，转换长期和短期风险
敞口没有影响，但在确定与替代 beta 相关的风险溢价时，长期和短期风险敞口的选择至关
重要。当我们谈论价值风险溢价时，我们明确地假设价值型股票的表现优于成长型(或反价
值型)股票，我们面临着规模、动量、低波动性和质量风险溢价的同样问题。这些选择是构
建要素投资组合的核心。因此，根本问题不在于价值风格、估值方法或价值风险因素，而在
于要素投资框架中使用的股票价值溢价。

2.2The value risk premium

2.2 价值风险溢价

In the sequel, we mainly based our analysis on the Fama-French definition of the
value risk factor, namely using the book-to-market and called HML1. Therefore, when
referring to growth stocks, we actually allude to anti-value stocks(the ones with the
lowest book-to-market). We have deliberately decided to take such an agnostic view
because it is the benchmark. Even though the value risk factor may be differently
implemented in factor investing portfolios, we think that the well-known HML factor
is representative of its average performance. Contrary to academic research, we
focus on a shorter period corresponding to the last 20 years, and more specifically in
the last decade corresponding to the value crisis.

在续集中，我们的分析主要基于 Fama-French 对价值风险因素的定义，即使用账面到市场

的方法，称为 HML1。因此，当提到成长型股票时，我们实际上指的是反价值型股票(账面
价值最低的股票)。我们故意决定采取这种不可知论的观点，因为它是基准。尽管价值风险
因素在要素投资组合中的实现方式可能不同，但我们认为著名的 HML 因素代表了其平均表
现。与学术研究相反，我们关注的是与过去 20 年相对应的较短时期，更具体地说，是与价
值危机相对应的过去 10 年。
There is no doubt that value 30 years ago is not the same as today. Indeed, in the last
decade, alongside the developments in the electronic trading market thanks to
advances in the internet, transaction costs substantially shrank, easing investors’
ability to arbitrage in the short-term. For example, Chordia et al.(2001) showed that
the average bid-ask spread was practically cut by half for U.S. equities between 1988
and 1998, and the average
毫无疑问，30 年前的价值与今天不同。事实上，在过去十年，随着互联网的发展，电子交
易市场得以发展，交易成本大幅下降，减少了投资者在短期内套利的能力。例如，Chordia
等人(2001)指出，1988 年至 1998 年间，美国股票的平均买卖价差实际上减少了一半，而美
国股票的平均买卖价差则减少了一半

1The Fama French factor is built using the value-weight portfolios on book-to-market:
1 Fama French 因素是使用“账面对市场”(book-to-market)的价值加权投资组合构建的:

Thus it should be noted that the HML factor is not sector neutral.

因此需要注意的是，HML 因素并不是行业中性的。

Table 1: Empirical moments of the HML monthly returns

表 1: HML 月度回报的经验时刻
2 2 2
0 0 0
0 0 1
0 0 0
- - -
E 2 2 2
u 0 0 0 5
r 2 0 2 t
o 0 9 0 h
p 2 2 2 第
e 0 0 0 五
欧 0 0 1 名
洲 0 0 0
- - -
2 2 2
0 0 0
2 0 2
0 9 0
M 0 1 − −
e . . 0 2
 . .
2 0 3 1
9 4 9 9
% % % %
a
n 0 1 ー -
刻 . . 0 2
薄 2 0 . .
9 4 3 1
% % 9 9
% %

− −
0 0 0 2
. . . .
M
2 7 5 2
e
7 3 4 4
d
% % % %
i
a
0 0 - -
n
. . 0 2
中
2 7 . .
位
7 3 5 2
数
% % 4 4
% %

1
7 7 7 3
V . . . .
a 8 7 0 0
r 7 3 2 4
i % % % %
a
n 7 7 7 1
c . . . 3
e 8 7 0 .
方 7 3 2 0
差 % % % 4
%

S 0 0 0 −
k . . . 0
e 3 2 3 .
w 0 9 0 9
n 9
e -
s 0
s .
扭
9
曲
9
度
K
u
r
t
2 3 3 2
o
. . . .
s
7 0 3 4
i
2 6 0 1
s
峭
壁
病
20 20 20
00 00 10 5
U- - - t
S 20 20 20 h
A 20 09 20 *
5
20 20 20
美 00 00 10 第
国- - - 五
20 20 20 名
20 09 20
0 0
. . − − −
0 6 0. 0. 1.
M
8 5 4 1 2
e
% % 4 2 7
a
% % %
n
0 0 - - -
. . 0. 0. 1.
刻
0 6 4 1 2
薄
8 5 4 2 7
% % % % %

M − 0 − − −
e 0 . 0 2 2
d . 3 . . .
i 1 2 4 1 3
a 0 % 7 3 7
n % % % %
中 0
位 - . - - ー
数 0 3 0 2 2
. 2 . . .
1 % 4 1 3
0 7 3 7
 % % % %

1 1 6 5 4
0 3 . 3 2
. . 9 . .
V
5 8 8 9 9
a
5 6 % 1 5
ri
% % % %
a
6
n
1 1 . 5 4
c
0 3 9 3 2
e
. . 8 . .
方
5 8 % 9 9
差
5 6 1 5
% % % %

− −
0 1
Ske . .
wn 0 0 5 0 2
ess . . 3 . 7
扭 1 1 - 2 -
曲 2 5 0 9 1
度 . .
5 2
3 7
Kur
tosi 3 1 4 0 1
s . . . . .
峭 1 8 9 1 3
壁 1 8 8 8 3
病

Source: Authors’ calculations based on Kenneth French data from 31/12/2000 to 31/12/2020.

资料来源: 作者根据 Kenneth French 2000 年 12 月 31 日至 2020 年 12 月 31 日的数据计算。

Note: 5th corresponds to the lowest 5th percentile of the distribution of the respective market factor returns, that is Nov.
2000, Feb. 2001, Sep. 2001, Jul. 2002, Sep. 2002, Jun. 2008, Sep. 2008, Oct. 2008, Feb. 2009, Dec. 2018,Feb. 2020 and Mar.
2020 for the U.S. In Europe, it corresponds to Sep. 2001, Jul. 2020, Sep. 2002, Jan. 2008,Sep. 2008, Oct. 2008, Jan. 2009, May
2010, Aug. 2011, Sep. 2011, May 2012 and Mar. 2020. For USA 5th*, we discard the monthly return of Nov. 2000.

注: 第 5 个百分位数对应于各自市场要素收益分布的最低第 5 个百分位数，即美国 2000 年 11 月、2001 年 2 月、2001 年 9 月、

2002 年 7 月、2002 年 9 月、2008 年 9 月、2008 年 10 月、2009 年 2 月、2018 年 12 月、2020 年 2 月和 2020 年 3 月。在欧洲，
它对应于 2001 年 9 月、2020 年 7 月、2002 年 9 月、2008 年 9 月、2008 年 10 月、2009 年 1 月、2010 年 5 月、2011 年 8 月、
2011 年 9 月、2012 年 5 月和 2020 年 3 月。对于美国第五 * ，我们丢弃了 2000 年 11 月的每月申报表。

trading daily volume per stock doubled. This trend has continued in the 21st century.
This is key for a mean-reverting strategy such as value. Therefore, one could wonder
whether the investment horizon of value investors has shortened over the years and
if the long-term horizon of value investors is compatible with this short-term
arbitrage. On top of that,internet development has also democratized access to
knowledge and data, and the financial world is no exception. Thanks to data
providers, a rising share of fund managers has gained access to corporate financial
statements with a click of a mouse. A phenomenon that also might have increased
short-term arbitrage of value in the past decades.

每股成交量翻了一番。这种趋势在 21 世纪仍在继续。这是价值等均值回归策略的关键。因
此，人们可能想知道价值投资者的投资期限是否已经缩短，价值投资者的长期期限是否符合
这种短期套利。除此之外，互联网的发展也使知识和数据的获取民主化，金融世界也不例外。
多亏了数据提供商，越来越多的基金经理可以通过点击鼠标访问公司财务报表。在过去的几
十年里，这种现象也可能增加了短期套利的价值。
Within the factor investing literature, value belongs to the “pure risk premia”
category.This is not a market anomaly. More specifically, it is considered as a
skewness risk premium,that is supposed to pay off in good market conditions but
would suffer a performance drag in case of market downturns. Indeed, as an
alternative risk premium, it implies that a value investor bears the risk in bad times.
This type of strategy can therefore be compared to the payoff of a short put+ long
call option(Roncalli, 2017). In Figure 20 in Appendix A.1, we have plotted the
distribution of the monthly returns from HML Europe and HML USA. We worked on
the full sample and split it into two sub-periods(up to December 2009 and since
January 2010). In addition, keen to analyze the payoff in turbulent markets, we
isolated the returns in periods of market downturns. Table 1 highlights the main
characteristics of these distributions. For both the U.S. and Europe, we observe a
slight positive skew for the full sample. Average returns were higher for the first
decade of our analysis, both in Europe and the U.S. These results highlight the
deterioration of value since 2010, corroborating stylized facts. Finally, the analysis of
market downturns highlights the negative skewness of the distribution of HML
returns in Europe, a property expected of a skewness risk premium such as the value
premium. In the U.S., oddly we do not observe such an effect. However,we should
bear in mind that the periods of market downturn in the U.S. include the bursting of
the tech bubble, where the value risk factor has strongly rallied. When discarding the
data from November 2000, we do find a negative skewness. These results confirm the
pure risk premium feature of the value risk factor.

在要素投资文献中，价值属于“纯风险溢价”范畴。这不是市场异常现象。更具体地说，它
被认为是一种偏态风险溢价，在市场状况良好的情况下应该获得回报，但在市场低迷的情况
下会受到业绩拖累。事实上，作为另一种风险溢价，它意味着价值投资者在经济不景气时承
担风险。因此，这种类型的策略可以与短期看跌 + 长期看涨期权的收益进行比较
(Roncalli，2017)。在附录 a. 1 中的图 20 中，我们绘制了来自 HML Europe 和 HML USA
的月度回报的分布。我们对整个样本进行了研究，并将其分为两个子时期(截至 2009 年 12
月和 2010 年 1 月)。此外，我们热衷于分析动荡市场的回报，将市场低迷时期的回报分离出
来。表 1 突出了这些分布的主要特征。对于美国和欧洲，我们观察到整个样本略有正向倾斜。
在我们分析的第一个十年里，欧洲和美国的平均回报率都更高。这些结果突出了自 2010 年
以来价值的恶化，证实了程式化的事实。最后，对市场低迷的分析突出了欧洲 HML 收益分
布的负偏态，这是一种预期的偏态风险溢价，如价值溢价。奇怪的是，在美国，我们没有观
察到这种效应。不过，我们应该记住，美国市场低迷时期包括科技泡沫破裂，其中价值风险
因素强劲反弹。当抛开 2000 年 11 月的数据时，我们确实发现了一个负偏差。这些结果证
实了价值风险因素的纯风险溢价特征。
3Economic analysis of the value risk factor
3 价值风险因素的经济学分析

3.1Macroeconomic determinants of the value strategy

3.1 价值战略的宏观经济决定因素

In this section, we are keen to assess the macroeconomic determinants of the value
factor2.We recall that Fama and French(1993) used book-to-market sorting portfolios
to build the High Minus Low or HML risk factor. More specifically, we will study its
relationship with interest rates(distinguishing between the level and slope of the yield
curve), inflation dynamics, volatility regimes, periods of recession and credit market
conditions.

在这一部分，我们热衷于评估宏观经济决定因素的价值因素 2。我们回顾，Fama 和 French

(1993)使用账面到市场的分类投资组合，以建立高负低或 HML 风险因素。更具体地说，我
们将研究它与利率(区分收益率曲线的水平和斜率)、通胀动态、波动机制、衰退周期和信贷
市场状况的关系。

Figure 1: Cumulative performance of the HML factor(long value, short growth)

图 1: HML 因素的累积表现(长期价值，短期增长)

HML EuropeHML USA

欧洲美国

Figure 1 represents the cumulative performance of the HML factor in the U.S. and
Europe since the beginning of the 1990s. Since then, and up to the late 2000s, the
value risk factor has enjoyed strong performance, although it certainly suffered when
it became clear that the dot-com bubble was forming. However, when the bubble
bursts in March 2000, value experienced one of its most powerful rallies. This is
driven by the nature of this crisis marked by the polarization between Technology
and Telecom sector stocks that were crushed by market players, and the other
sectors of the economy that withstood. Being long value and short on growth stocks
was without any doubt detrimental between 1998and 2000, but then a sound
turnaround played out for HML. Merely at that time, being a value or growth investor
actually meant betting on a sector’s performance, rather than on true “valueness” of
underlying stocks. In that matter, the dot-com bubble remains an atypical event for
value, which is not expected to display such outperforming returns during a crisis. In
a sense, the dot-com bubble is a pure financial crisis of valuation and its impact on
the economy has been limited compared to the 1929 Great Depression or the 2008
Global Financial Crisis. However, after peaking in the wake of the dot-com bubble
burst until the beginning of 2007(mid-2008 in Europe, but to a lesser extent), value
has plummeted sharply. Until last September, 2020 was about to be coined the worst
year for value, but a small turnaround was triggered. These findings apply both in the
U.S. and Europe. As a matter of fact, performances on both sides of the Atlantic seem
to co-move, although the

图 1 表示了自 20 世纪 90 年代初以来美国和欧洲的 HML 因素的累积性能。自那时以来，直

到本世纪头十年末，价值风险因素一直表现强劲，尽管当互联网泡沫形成的迹象变得明显时，
它肯定受到了影响。然而，当泡沫在 2000 年 3 月破裂时，价值经历了最强劲的反弹。这是
由于这场危机的性质造成的，其特点是被市场参与者压垮的技术和电信部门股票与经济中经
受住冲击的其他部门股票之间的两极分化。在 1998 年到 2000 年间，做多和做空成长型股
票毫无疑问是有害的，但是之后，一个良好的转变出现在了 HML 身上。仅仅在当时，作为
一个价值投资者或增长投资者实际上意味着押注于一个行业的表现，而不是基础股票的真正
“价值”。在这个问题上，互联网泡沫仍然是一个非典型的价值事件，在危机期间不会显示
出如此优异的回报。从某种意义上说，互联网泡沫是一场纯粹的估值金融危机，与 1929 年
的大萧条(Great Depression)或 2008 年的全球金融危机(Global Financial Crisis)相比，其
对经济的影响一直很有限。然而，在互联网泡沫破灭后达到顶峰直到 2007 年初(2008 年中
期在欧洲，但程度较轻) ，价值急剧下跌。直到去年 9 月，2020 年才被认为是最糟糕的一年，
但是一个小小的转机被触发了。这些发现适用于美国和欧洲。事实上，大西洋两岸的表现似
乎是同步的

2The analysis has been conducted with the Kenneth French library data available at the
following website:http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data library.html.
We use the monthly Fama-French three factors(MKT, SMB and HML) for the U.S. and Europe,
both in USD with historical data starting from July 1926 and July 1990, respectively.

2 该分析是根据 Kenneth French 图书馆的数据进行的，该数据可在以下网站获得:

http://mba.tuck.dartmouth.edu/pages/faculty/ken.French/data library. html。我们使用美国和欧
洲的 Fama-French 三个因素(MKT，SMB 和 HML) ，均以美元计算，历史数据分别从 1926 年 7 月和
1990 年 7 月开始。

U.S. has not managed to catch up with the outperformance Europe accumulated at
the beginning of the 2000s, where they were clearly de-correlated.
美国没有设法赶上欧洲在 21 世纪初积累的优异表现，在那里它们显然是去相关的。
In order to understand the performance drag on value returns in recent decades, we
investigated its ties to the interest rate environment and monetary policy stance.
First,based on data for 10Y interest rates and the 10Y-2Y yield curve slope, we cannot
establish a strong conviction on the co-movements of the value risk factor with
interest rates3 as shown in Figure 2. Indeed, interest rates have experienced a
continuous downward trend, but the value premium was much more volatile.

为了了解近几十年来价值回报的表现拖累，我们研究了其与利率环境和货币政策立场的关系。
首先，根据 10y 利率和 10y-2y 收益率曲线斜率的数据，我们无法确定价值风险因素与利率
3 的共同变动，如图 2 所示。事实上，利率已经经历了一个持续下降的趋势，但价值溢价更
加波动。

Figure 2: Historical interest rates movements in bps(level and slope)

图 2: 英镑(水平和斜率)的历史利率变动

(b) Europe

(b)欧洲

Figure 3: Average monthly return of the value risk factor(in%) across interest rates
and inflation regimes

图 3: 不同利率和通货膨胀制度的价值风险因素(以% 为单位)的平均月回报率
.5.8
0.5.8
(a) Interest rates regimes(b) Inflation regimes
(a)利率制度(b)通货膨胀制度

In Figure 3, we extend the previous analysis by investigating the average

performance of the value factor across different yield curve movements. We find
similar patterns in the U.S. and Europe. Value tended to outperform when the level of
interest rates increased and when the yield curve steepened. This is consistent with
the economic theory. When the economy is growing at a solid pace, this generally
translates into an upward movement in interest rates, combined with a steepening
yield curve. Therefore, value stocks usually perform well in such an environment,
where earnings of value stocks grow faster than growth stocks. Actually, because of
their cash flow structure, the equity duration of value stocks

在图 3 中，我们通过研究价值因素在不同收益率曲线变动中的平均表现，扩展了前面的分析。
我们在美国和欧洲发现了类似的模式。当利率水平上升和收益率曲线变陡时，价值往往表现
更好。这与经济学理论是一致的。当经济以稳定的速度增长时，这通常会转化为利率的上升，
并伴随着收益率曲线的上升。因此，价值型股票通常在这种环境下表现良好，价值型股票的
收益增长快于成长型股票。实际上，由于它们的现金流结构，价值股票的持股期限

3For the U.S., we use the monthly 10Y and 2Y Treasury constant maturity rate from the Federal
Reserve Bank of St. Louis database. The sample starts in July 1976. For Europe, we chose the
monthly German 10Y and 2Y benchmark bond. Data was retrieved from Factset, and the
sample starts in August 1997.

对于美国，我们使用来自圣路易斯联邦储备银行数据库的每月 10 年期和 2 年期美国国债的恒定到期率。

样本开始于 1976 年 7 月。对于欧洲，我们选择了每月一次的德国 10y 和 2y 基准债券。数据从
Factset 中检索，样本从 1997 年 8 月开始。

is often considered as shorter than the equity duration of growth stocks, because the
latter have cash flows that pay out in the distant future(Schr¨oder and Esterer, 2016).
Thus, when interest rates are high, value stocks are priced higher than growth stocks,
since a lower share of their earnings is deeply discounted in the future. Conversely,
growth stocks are generally favored in a low-interest rate regime, where such a
discount is very low. However, value stocks also have a higher cash flow
beta(Campbell and Vuolteenaho, 2004). This makes them fluctuate with the
economy’s health and therefore they tend to be concentrated in cyclical sectors such
as Financials or Industrials. It should be noted that such interaction with a rising
interest rate regime illustrates the antagonistic payoff profile of the value risk factor
compared to the low-volatility factor. Indeed, the latter usually shows the strongest
performance when rates are on a downward trend(Stagnol and Taillardat, 2017).
通常被认为比成长型股票的持股时间短，因为后者的现金流在遥远的未来才会支付(Schr
order and Esterer，2016)。因此，当利率较高时，价值型股票的定价高于成长型股票，因
为其收益中较低的份额在未来会被大幅折现。相反，成长型股票通常在低利率制度下受到青
睐，在这种制度下折扣非常低。然而，价值型股票也有更高的现金流 β 值(Campbell and
Vuolteenaho，2004)。这使得它们随着经济的健康而波动，因此它们倾向于集中在周期性
行业，如金融或工业。应当指出，这种与上升利率制度的相互作用表明，与低波动率因素相
比，价值风险因素的收益状况是相互对立的。事实上，当利率处于下降趋势时，后者通常表
现出最强的表现(Stagnol 和 Taillardat，2017)。
In a similar manner, we examine whether an inflationary environment(usually fueled
by a growing economy) is conducive to the value premium4. Specifically, we analyze
how value returns historically behaved when inflation was rising or high(equal or
above to 2%).In Figure 3, we notice that value outperformed when inflation was in
high regime, both in the U.S. and Europe. Still, judging by the close performance for
increasing or decreasing inflation, HML performance seems more sensitive to the
level of inflation rather than its direction. This is in line with the essence of value
stocks, which have a higher expected profit in the short term than growth stocks, for
which profits are much further down the road.

以类似的方式，我们检验通货膨胀的环境(通常由经济增长推动)是否有利于价值溢价。具体
来说，我们分析了当通货膨胀上升或高(等于或高于 2%)时，价值回报在历史上是如何表现
的。在图 3 中，我们注意到当通货膨胀处于高位时，无论是在美国还是在欧洲，价值都表现
出色。尽管如此，从通胀上升或下降的近似表现来看，HML 的表现似乎对通胀水平更为敏
感，而不是其方向。这符合价值型股票的本质，价值型股票的短期预期利润高于成长型股票，
而成长型股票的利润要远远高于成长型股票。

In Figure 1 we have seen that market turmoil seems to have a significant impact on
the value premium, as suggested by the sharp rise after the dot-com bubble burst,
but also the drops from the GFC, the European debt crisis or the COVID-19 breakout.
In Figure 4, we note that value tended to track the VIX in the U.S. before 2004.
However, the relationship seems to have reversed since this period. Periods of low
volatility appear to have coincided with the outperformance of the value risk factor.

在图 1 中，我们看到市场动荡似乎对价值溢价产生了重大影响，正如互联网泡沫破裂后价值
急剧上升所表明的那样，但也包括全球金融危机、欧洲债务危机或 2019 冠状病毒疾病的突
破所表明的那样。在图 4 中，我们注意到在 2004 年之前，价值趋向于跟踪美国的波动率指
数。然而，这种关系似乎从这段时间开始就逆转了。低波动期似乎与价值风险因素的优异表
现相吻合。

Figure 4: Relationship between the historical volatility(in%) and the HML performance

图 4: 历史波动率(%)与 HML 性能之间的关系

European data(VSTOXX for the Eurostoxx volatility), graphically conveys the same
idea5. However, from Figure 5, we cannot draw any clear conclusions on the impact of

欧洲的数据(欧洲斯托克指数波动率的 VSTOXX)图形化地表达了同样的观点 5。然而，从图 5

中，我们不能得出任何明确的结论

4For the U.S., we use the monthly Consumer Price Index(CPI) for All Urban Consumers(All
Items in U.S. City Average since January 1948 from the Federal Reserve Bank of St. Louis
database) while we use the Euro Area HICP from the OECD in Europe(series starting in January
1997). Both time-series are percentage change from the same period of the previous year.
4 对于美国，我们使用所有城市消费者的每月消费物价指数(自 1948 年 1 月以来的美国城市平均消费
物价指数，来自圣路易斯联邦储备银行的数据库) ，而我们使用欧洲经合组织的欧元区 HICP (1997 年
1 月开始的系列)。这两个时间序列都是与前一年同期相比的百分比变化。
5VIX data for the U.S is from the Chicago Board Options Exchange, corresponds to the implied
volatility on the S&P 500, and starts in January 1990. For Europe, we used the VSTOXX index
based on the Euro Stoxx 50, whose inception date goes back to January 1999. We split the
sample between high and low regimes based on the average volatility index value over the
whole sample.

美国的 5vix 数据来自芝加哥期权交易所，对应于标准普尔 500 指数的隐含波动性，始于 1990 年 1 月。

对于欧洲，我们使用的是基于欧元 Stoxx 50 的 VSTOXX 指数，该指数的创始日期可以追溯到 1999 年
1 月。我们根据整个样本的平均波动率指数值将样本分为高和低两个区间。

Figure 5: Average monthly return of the value risk factor(in%) across volatility regimes
图 5: 不同波动率制度的价值风险因素(以% 为单位)的平均月回报率
.5
0.5
2.4

2.4
(a) Full sample 0.6
(a)完整样本 0.6
2.00.4
2.00.4
1.6
1.6
0.2
0.2
0.0
0.0
1.2
1.2
-0.2
-0.2
0.8
0.8
-0.4
-0.4
0.4
0.4
-0.6
-0.6
0.0-0.8
0.0-0.8
-0.4-1.0
-0.4-1.0
VIX upVIX downVIX highVIX lowVIX upVIX downVIX highVIX low
波动率指数，波动率指数，波动率指数，波动率指数，波动率指数
HML EuropeHML USAHML EuropeHML USA
美国欧洲

(b) Before 2004(c) After 2004

(b)2004 年前(c)2004 年后

volatility index’s direction for the whole sample, witnessing antagonistic return
patterns between Europe and the U.S. Therefore, we split the sample between the
period up to the end of December 2003 and after. We note that before 2004, value
returns were higher in periods of rising volatility. However, since then, the value risk
factor underperformed during the same periods of uncertainty. This result is in line
with the changing correlation hinted in Figure 4.

因此，我们将样本分为截至 2003 年 12 月底和 2003 年 12 月底两个时期。我们注意到，在

2004 年之前，在波动性上升的时期，价值回报率更高。然而，从那时起，价值风险因素在
同样的不确定时期表现不佳。这个结果与图 4 中暗示的变化的相关性是一致的。
Although value has tended to underperform in turbulent market environments since
2004,we may wonder if the value risk factor also suffers a performance drag during
subsequent periods of recession. In Figure 6, when isolating recession periods from
our sample, we observe that value indeed tends to underperform during those
difficult periods, both in the U.S. and in Europe6. This result is in line with our
previous finding. Value outperforms when the slope of the yield curve increases,
which implies an acceleration of economic growth.

尽管自 2004 年以来，价值在动荡的市场环境中往往表现不佳，但我们可能想知道，在随后

的衰退期间，价值风险因素是否也会受到业绩拖累。在图 6 中，当从我们的样本中分离出经
济衰退期时，我们观察到，无论是在美国还是在欧洲，在那些困难时期，价值确实往往表现
不佳。这个结果与我们之前的发现一致。当收益率曲线的斜率增加时，价值表现更好，这意
味着经济增长的加速。
Finally, considering that recession periods tend to be synchronized with widening
credit spreads, we would like to analyze the relationship between value and credit
market condi-tions. In Figure 7, we report the relationship between the credit
spread7 and the HML risk factor for both Europe and the U.S. This figure suggests
that narrowing credit spreads seem
最后，考虑到经济衰退期往往与不断扩大的信贷息差同步，我们愿意分析价值和信贷市场条
件之间的关系。在图 7 中，我们报告了欧洲和美国的信贷利差 7 和 HML 风险因素之间的关
系。这个数字表明，缩小的信贷利差似乎

6We use monthly NBER based recession indicators for the United States and OECD based
recession indicators for Euro area from the Federal Reserve Bank of St. Louis database. Data is
available for the whole sample for both HML factors.
我们使用美国国家经济研究局(NBER)的月度衰退指标和圣路易斯联邦储备银行(Federal Reserve Bank
of st. Louis)数据库中欧元区基于经合组织(OECD)的衰退指标。数据可用于两个 HML 因素的整个样本。
7For the U.S., we use the monthly Option Adjusted Spread(OAS) from ICE BofA US Corporate,
which is available starting in January 1997. For Europe, we use the monthly Option Adjusted
Spread from ICE BOFA Euro Corporate available since December 1996. We split each sample
into a high or a low OAS regime based on their respective average over the full sample period.

对于美国，我们使用美国洲际交易所美国公司提供的每月期权调整利差(OAS) ，该利差从 1997 年 1 月

开始提供。对于欧洲，我们使用自 1996 年 12 月以来美银欧洲公司每月的期权调整利差。我们根据每
个样本在整个样本期间的平均值，将每个样本分为高美洲国家组织和低美洲国家组织。

Figure 6: Average monthly return of the value risk factor(in%) across economic
activity
图 6: 整个经济活动的价值风险因素(以% 为单位)的平均月回报率

to be the most favorable environment for the value risk factor. However, results are
mixed when breaking down high versus low spread regimes as shown in Figure 8.
Although value outperforms in a low credit spread regime in Europe, no clear pattern
appears in the U.S.From this standpoint, the directionality of the credit spreads
appears to be more meaningful that its level when it comes to explaining the
performance of the value risk factor.

成为价值风险因素最有利的环境。然而，如图 8 所示，当分解高低传播机制时，结果是混合
的。尽管在欧洲低信贷利差体系中，价值表现优于其他国家，但在美国，从这个角度来看，
信贷利差的方向性在解释价值风险因素的表现时似乎比它的水平更有意义。

Figure 7: Historical credit spread in bps

图 7: 以英镑为单位的历史信贷利差

We saw that the set of macroeconomic factors seems to impact the performance of
value in our sample. With the aim of appraising their combined effect and identifying
the strongest determinants, we ran least squares regressions on the HML monthly
returns in Table 2 for Europe and the U.S. Results generally confirm our previous
conclusions. In line with Figures 5 and 6, we find mixed evidence of recession that is
mildly significant, and of volatility that is only meaningful in the U.S. Although taken
individually, the level of interest rates(and the slope in the U.S.) has a positive
significant impact on the value risk factor. When combined with other variables,
interest rate dynamics do not come out as compelling as in Figure 3. As a matter of
fact, our econometric analysis shows that, when combining the aforementioned
different macroeconomic indicators, credit spread tightening and inflationary
environment (and to a lesser extent the absence of recession) are the most conducive
determinants to the outperformance of the value risk factor.

我们看到，在我们的样本中，一系列宏观经济因素似乎影响价值的表现。为了评估它们的综
合效应和确定最强的决定因素，我们对欧洲和美国的 HML 月收益进行了表 2 中的最小二乘
回归。结果基本上证实了我们以前的结论。与图 5 和图 6 一致，我们发现复杂的证据表明经
济衰退有轻微的影响，而波动性只对美国有意义。虽然单独来看，利率水平(以及美国的斜
率)对价值风险因素有积极的重大影响。当与其他变量相结合时，利率动态并不像图 3 那样
引人注目。事实上，我们的计量经济分析显示，当综合上述不同的宏观经济指标时，信贷利
差收紧和通胀环境(以及在较小程度上没有出现衰退)是价值风险因素表现优异的最有利因素。

Figure 8: Average monthly return of the value risk factor(in%) across credit spread
regimes
图 8: 不同信用利差制度的价值风险因素的平均月回报率(以% 为单位)

Table 2: Time series regressions

表 2: 时间序列回归

Europe
欧洲
Constant
不变的
0.26
−0.32
-0.32
0.39*
0.39 *
0.29*
0.29 *
0.24
−0.37
-0.37
10Y level
10 层
2.28**
2.28 * *
0.63
1− 0
I 0− 0
R −
C − −
V −0
R0 0 0 0 0 0
U
−
−
0
0
.
C .
6
o 7
8
n 7
*
s *
t
-
a 0 00 − ー
n 0
0
t .
.
不 7
6
变 7
8
的
*
*

10 −
11 1
I 0 0
R − −
 C − −
V 02
R0 0 0 0 0 0

Source: Authors’ calculations based on data extracted FRED Database& Factset.

来源: 作者的计算基于数据提取 FRED 数据库和 Factset。

Note: Sample starts from February 1999 to December 2020 for Europe, and February 1997 to December 2020 for the U.S.
Interest rate variables are taken in first difference. Inflation is already in difference and recession is a dummy variable.
Credit spread, VSTOXX and VIX are percentage change from the previous period.

注: 样本从 1999 年 2 月至 2020 年 12 月为欧洲，1997 年 2 月至 2020 年 12 月为美国，利率变量采用第一差值。通货膨胀率已经存

在差异，而经济衰退只是一个虚拟变量。信贷利差，VSTOXX 和 VIX 是与前一时期相比的百分比变化。

3.2Microeconomic factors of the value strategy

3.2 价值战略的微观经济因素

We now turn to microeconomic determinants of the value premium and investigate

whether recent underperformance can be attributed to the highest book-to-market
firms or the so-called “deep value” stocks. These companies are distressed for various
possible reasons such as having the capacity to repay interest, but not the principal,
inadequate income generation to cover debt expenses, etc. For real deep value
stocks, such critical situation can turn into a recovery, the company being acquired or
going bankrupt. Economic downturn or rising interest rates are often the catalysts for
the crash of deep value stocks. Still, such stocks may find buyers. In essence, both
deep value investors and more traditional value investors believe in price mean-
reversion. However, with a shorter time horizon than for traditional value, investing in
this deepest segment is often assimilated to a speculative strategy. Indeed,deep
value investors are believers in short-term rotation where sudden price reversion
could occur, for instance fueled by supportive flows. Managing to capture the latter is
not an easy task, as mean-reversion is generally not as sustained as for generic value
stocks. Nonetheless,such contrarian bets can have very positive outcomes, but it is all
about stock selection.

我们现在转向价值溢价的微观经济决定因素，并调查近期表现不佳是否可归因于账面价值最
高的公司或所谓的“深度价值”股票。这些公司因各种可能的原因而陷入困境，例如有能力
偿还利息而无法偿还本金、创收不足以支付债务开支等。对于真正的高价值股票来说，这种
危急的情况可以转变为复苏，公司被收购或破产。经济衰退或利率上升往往是深价值股票崩
溃的催化剂。尽管如此，这些股票可能会找到买家。本质上，深度价值投资者和传统价值投
资者都相信价格均值回归。然而，由于时间范围比传统价值更短，投资于这个最深层次的部
分通常被等同于投机策略。事实上，深度价值投资者相信短期轮换，在这种情况下，价格可
能会突然回落，例如由支持性资金流推动。设法捕捉后者并不是一件容易的事情，因为均值
回归通常不像普通价值股那样持续。尽管如此，这种反向押注可能会产生非常积极的结果，
但这完全取决于股票选择。

Figure 9: 12 months rolling average monthly returns(in%) of deep value(long 95th

per-centile, short 75th percentile on the book-to-price ratio)

图 9:12 个月滚动深度价值的平均月回报率(百分比)(长期 95% ，短期 75%)

一 EMU Deep-ValueUSA Deep-Value

一个动车组深度价值

In Figure 9, we plotted the difference between the performance of the 25% highest
book-to-market companies versus the highest 5%, the latter being our proxy for deep
value. We work on two distinct universes, the MSCI EMU Value and the MSCI USA
Value, from December 2000 to December 20208. We note that deep value rotation
can occur over a very short period as illustrated during the GFC for the EMU stock
universe. On a more general note, deep value performance was inconsistent across
the past two decades but was broadly aligned across the two sides of the Atlantic
since 2013. The dot-com bubble crash hindered deep value stocks(especially in the
U.S.) that recovered up to the GFC, where they dropped sharply in the EMU but
resisted fairly well in the U.S. Performance oscillated during the European sovereign
debt crisis. The COVID-19 outbreak was also a severe hit (especially in Europe), but a
V-shaped recovery has already occurred. Nonetheless, on average over this sample, a
deep value investor would have been significantly worse-off in the EMU. It is
interesting to note the last two decades were marked by decreasing interest

在图 9 中，我们绘制了账面价值最高的 25% 公司与最高的 5% 公司之间的业绩差异，后者

是深度价值的代表。从 2000 年 12 月到 20208 年 12 月，我们研究了两个不同的领域，
MSCI EMU 值和 MSCI USA 值。我们注意到深度价值旋转可以在非常短的时间内发生，如
在 EMU 股票宇宙的 GFC 期间所说明的那样。更一般地说，深度价值业绩在过去二十年间并
不一致，但自 2013 年以来，大西洋两岸大致一致。互联网泡沫破灭阻碍了深值股票(尤其是
美国股票)的复苏，直至全球金融稳定基金(GFC)。在全球金融稳定基金中，深值股票在经济
与货币联盟(EMU)中大幅下跌，但在美国表现相当不错，在欧洲主权债务危机期间表现起伏
不定。2019 冠状病毒疾病的爆发也是一次严重的打击(特别是在欧洲) ，但是 v 型的恢复已
经出现。尽管如此，在这个样本中，一个深度价值投资者在经济和货币联盟中的平均情况要
糟糕得多。值得注意的是，过去 20 年来，人们的兴趣在不断下降

8As far as data cleaning is concerned, negative book-to-market and those over 100 are left out
of the analysis for this section.

就数据清理而言，本节的分析中没有提到负面的市场账目和超过 100 本的账目。

rates, which usually preserve deep value firms from going bankrupt. We argue that
actually,this supportive monetary stance has been responsible for such
underperformance. Indeed, it prevented the market from clearing its distressed
issuers, which entailed the value premium.Hence, we can argue that these “zombie
stocks” can be partly held responsible for the value performance drag over the last
decade.
利率，这种利率通常可以保护高价值公司免于破产。我们认为，实际上，这种支持性的货币
立场是造成这种表现不佳的原因。事实上，它阻止了市场清算其陷入困境的发行人，这意味
着价值溢价。因此，我们可以认为，这些“僵尸股票”可以部分地为过去十年的价值表现拖
累负责。
As explained above, the sustained low interest rate environment prevented the deep
value market from clearing. As a matter of fact, it may also have created distortions in
the distribution of metrics for the deep value basket. In Figures 22 and 23 in
Appendix A.1,we construct the box-plots of the book-to-price ratio for both the MSCI
EMU and MSCI USA universes at the constituent level. We note that the dispersion
tended to widen in turbulent markets, for example between 2008 and 2012 or in
2020. Very different trends can be observed, such as a continuous rise in dispersion
in the U.S. versus a more varying pattern in EMU. As a matter of fact, dispersion in
the U.S. was almost twice the EMU figure in 2020. This rise in dispersion, even only
temporarily in the EMU, shed light on an increasing number of firms in a distressed
situation, and is likely to imply a performance drag from the deepest value stocks, as
already hinted in Figure 9.

正如上文所解释的，持续的低利率环境阻碍了深层价值市场的清算。事实上，它也可能造成
了深度价值篮子指标分布的扭曲。在附录 a. 1 的图 22 和 23 中，我们构造了 MSCI 经济与货
币单位和 MSCI 美国宇宙在组成水平上的账面价格比的盒图。我们注意到，在动荡的市场中，
分布趋于扩大，例如在 2008 年至 2012 年或 2020 年之间。我们可以观察到非常不同的趋势，
比如美国的分散度持续上升，而欧洲货币联盟的分散度则更加不同。事实上，美国的分散度
几乎是 2020 年 EMU 数字的两倍。这种分散度的上升，即使只是暂时地在欧洲经济货币联
盟中，揭示了越来越多的公司处于困境之中，并且可能意味着来自价值最高的股票的业绩拖
累，如图 9 所示。
In Section Two, we have seen that value is a skewness risk premium, implying
potential significant losses in case of a market downturn. Clearly, this effect should
be magnified for deep value stocks. This can be verified with its performance
distribution plots in Figure 21in Appendix A.1, and the values of the statistical
moments that are reported in Table 3.
在第二部分中，我们已经看到价值是一个偏态风险溢价，意味着在市场低迷的情况下潜在的
重大损失。显然，这种影响应该被放大到深度价值股票上。这可以通过附录 a. 1 中的图 21
中的性能分布图以及表 3 中报告的统计矩的值来验证。

Table 3: Empirical moments of the deep value monthly returns(MSCI universe)

表 3: 深度价值月回报的经验时刻(MSCI 宇宙)

EMU 5th
2000-2020 2000-2009 2010-2020
经济增长 第五
2000-2020 2000-2009 2010-2020
模型 名
Mean −1.74% −1.72% −1.76% −5.18%
刻薄 - 1.74% - 1.72% ー 1.76% - 5.18%
Median −1.39% −1.68% −1.27% −4.13%
中位数 - 1.39% ー 1.68% - 1.27% - 4.13%
Variance 16.23% 17.78% 14.95% 19.46%
方差 16.23% 17.78% 14.95% 19.46%
Skewness −0.48 −1.09 −0.31
0.07
扭曲度 -0.48 ー 1.09 -0.31
Kurtosis −0.81
3.43 3.37 3.59
峭壁病 -0.81
5th
USA 2000-2020 2000-2009 2010-2020
第五
美国 2000-2020 2000-2009 2010-2020
名
Mean 0.08% 0.32% −0.12% −0.88%
刻薄 0.08% 0.32% - 0.12% - 0.88%
Median −0.18% −0.19% −0.17% −0.72%
中位数 - 0.18% - 0.19% - 0.17% - 0.72%
Variance 8.39% 12.68% 4.77% 3.38%
方差 8.39% 12.68% 4.77% 3.38%
Skewness −0.20 −0.20
2.47 2.89
扭曲度 -0.20 -0.20
Kurtosis −0.67
16.00 14.27 1.03
峭壁病 -0.67

Source: Authors’ calculations based on based on MSCI data, from 31/12/2000 to 31/12/2020.

资料来源: 作者根据 2000 年 12 月 31 日至 2020 年 12 月 31 日 MSCI 数据进行的计算。

Note: Deep value is defined as the difference between the performance of the 5% and 25% highest book-to-price stocks.
5th corresponds to the lowest 5th percentile of the distribution of the respective market factor returns (MSCI EMU or MSCI
USA). In EMU, it corresponds to Sep. 2001, Jul. 2002, Sep. 2002, Jan. 2008, Sep. 2008,Oct. 2008, Jan. 2009, May 2010, Aug.
2011, Sep. 2011, May 2012 and Mar. 2020. In the USA, it corresponds to Feb. 2001, Jun. 2002, Sep. 2002, Jun. 2008, Sep.
2008, Oct. 2008, Jan. 2009, Feb. 2009, May 2010, Dec. 2018,Feb. 2020 and Mar. 2020.

注: 深度价值定义为账面价值最高 5% 和 25% 的股票表现之间的差异。第 5 对应于各自市场因素回报率(MSCI EMU 或 MSCI USA)分

布的最低的第 5 个百分点。在 EMU 中，它对应于 2001 年 9 月、2002 年 7 月、2002 年 9 月、2008 年 9 月、2008 年 10 月、2009
年 1 月、2010 年 5 月、2011 年 8 月、2011 年 9 月、2012 年 5 月和 2020 年 3 月。在美国，它对应于 2001 年 2 月、2002 年 6 月、
2002 年 6 月、2008 年 9 月、2008 年 10 月、2009 年 1 月、2009 年 2 月、2010 年 5 月、2018 年 12 月、2020 年 2 月和 2020 年 3
月。

Over the last 10 years, deep value returns have been negative both in the EMU and in
the U.S.(as seen in Figure 9). Indeed, recently in the U.S., deep value performance has
significantly lagged compared to a flat performance for the previous decade. It is
interesting to note that although deep value performance is slightly negatively
skewed for Europe, we observe a significant and positive skew in the U.S. during the
2000-2010 period before it

在过去 10 年中，EMU 和美国的深度价值回报率都是负值(如图 9 所示)。事实上，最近在美

国，深层价值的表现已经明显落后于过去十年持平的表现。值得注意的是，尽管欧洲的深度
价值表现略有负向倾斜，但我们观察到，在此之前的 2000 年至 2010 年期间，美国的深度
价值表现出明显的正向倾斜

turns slightly negative. As expected, in case of market downturns, deep value took a
severe hit, as revealed by the strongly negative skewness, although the U.S. deep
value segment seems to hold up better in case of a market downturn. Our analysis
shows that deep value stocks have caused a performance drag on value, since 2000
for the EMU and 2010 for the U.S. However, this is not the full story. Value rallied until
2007 in the U.S. and until 2009for Europe. Surely deep value stocks clearly cannot
take all the blame.
变得稍微负面。正如预期的那样，在市场低迷的情况下，深层价值受到了严重打击，正如强
烈的负偏差所显示的那样，尽管在市场低迷的情况下，美国深层价值部分似乎表现得更好。
我们的分析显示，自 2000 年以来，深度价值型股票已经对价值造成了拖累，对于 EMU 和
2010 年的美国来说都是如此。价值在美国上涨到 2007 年，在欧洲上涨到 2009 年。当然，
深度价值股显然不能承担所有的责任。

3.3On the interconnectedness between credit and equity value strategies

3.3 论信贷与股权价值策略的相互关联性

Aware that value in equity and credit markets convey the same idea, but translate
very differently in the metrics employed, we decide to test the robustness of our
results. Does changing the value definition yield different performance figures? Thus,
we took a credit-like approach to defining value, more specifically using the Expected
Default Frequency(EDF)computed by Moody’s Analytics CreditEdgeTM. Our quarterly
dataset starts in March 2006and ends in June 20209. In Figures 10 and 11, we observe
that switching the definition in this way does not alter the recent underperformance
of value equity. The stocks with the lowest price-to-book have lagged, so have the
companies with the highest EDFs, which is our value proxy for the fixed-income
world. In both universes, these metrics are actually very efficient for sorting
portfolios. Once again, it points at the recent underperformance of value since the
mid 2000s.

我们意识到股票市场和信贷市场的价值传达了同样的思想，但是在所采用的度量标准中转换
的方式却非常不同，因此我们决定测试我们结果的稳健性。改变价值定义会产生不同的绩效
数据吗？因此，我们采用了一种类似信用的方法来定义价值，更具体地说是使用 Moody’s
Analytics CreditEdgeTM 计算的预期默认频率(EDF)。我们的季度数据集从 2006 年 3 月开
始，到 20209 年 6 月结束。在图 10 和图 11 中，我们观察到以这种方式转换定义并不能改
变最近价值资产表现不佳的情况。市净率最低的股票已经落后，edf 最高的公司也是如此，
edf 是我们在固定收益领域的价值代表。在这两个世界中，这些指标实际上对于分类投资组
合都是非常有效的。它再一次指出，自 2000 年代中期以来，最近的价值表现不佳。

Figure 10: MSCI EMU quarterly returns(in%) of sorted portfolios

图 10: 摩根士丹利资本国际经济与货币联盟季度收益(以% 为单位)的排序投资组合

9For the MSCI EMU universe, EDF data coverage stands at 89% at the end of 2019, and 55% for
price-to-book data from Factset(91% and 93% respectively for the North American universe).

对于摩根士丹利资本国际经济与货币联盟(MSCI EMU)来说，EDF 数据覆盖率在 2019 年底为 89%

，Factset (北美)为 55% (分别为 91% 和 93%)。

Figure 11: MSCI North America quarterly returns(in%) of sorted portfolios

图 11: 摩根士丹利资本国际北美分类投资组合季度回报率(%)
Results for sorted portfolio returns are also available at the GICS sector level for the
EMU10 and North American universes in Figure 25 in Appendix A.1. First, the sorting
by industry does not give the same smooth sorting as in Figures 10 and 11. In fact,
the cleanest ranking similar to the latter is only observed for Materials, Financials,
Utilities and IT, although some inconsistencies exist11. A peculiar effect comes out.
Indeed, for EMU Energy firms, the higher the probability of default, the higher the
returns. In fact,sorted equity portfolios based on price-to-book remain more efficient
than sorted equity portfolios based on EDF. When defining value stocks as the Q1
sorted portfolio on price-to-book, Industrials and Healthcare stocks held up fairly well
over the 2006-2020 period both in North America and the EMU. Similarly, IT value
stocks withstood in North America.On the other hand, value stocks of Materials, EMU
Utilities and Consumer Discretionary suffered. Hence it reveals that value
performance has been uneven across industries.

排序投资组合回报的结果也可以在附录 a. 1 中的图 25 中的 emu10 和北美宇宙的 GICS 部门

级别获得。首先，按行业进行的排序不像图 10 和图 11 那样顺利。事实上，类似于后者的最
干净的排名只在材料、金融、公用事业和 IT 方面被观察到，尽管存在一些不一致的地方。
一个特殊的效应出现了。事实上，对于 EMU 能源公司来说，违约机率越高，回报就越高。
事实上，基于市净率的分类股票投资组合比基于 EDF 的分类股票投资组合更有效率。当将
价值股定义为按市净率计算的第一季度投资组合时，无论是在北美还是在欧洲货币联盟，
2006-2020 年期间，工业和医疗保健类股表现相当不错。同样，IT 价值股在北美也表现不
错。另一方面，材料、 EMU 公用事业和消费者自由选择等价值股则表现不佳。因此，这表
明各行业的价值表现并不均衡。

We now turn to the assessment of the interaction between value credit(EDF) and
value equity(price-to-book) metrics and how this relationship has evolved over the
recent years.We must recognize that the academic literature remains blurry on that
topic. As stated by Fama and French(1992, 1998), if the value premium should reward
heightened default risk, then one should find a negative correlation between
financial distress and price-to-book(Chen and Zhang, 1998). Nonetheless, it is also
argued that bankruptcy risk is not necessarily related to the price-to-book(Dichev,
1998; Griffin and Lemmon, 2002). To test this hypothesis, we plot the rolling one-year
correlation between the two metrics12, and the value quintile percentage overlap13
in Figure 12.

我们现在转向评估价值信贷(EDF)和价值权益(按账面价值计算)指标之间的相互作用，以及这
种关系近年来是如何演变的。正如 Fama 和 French (1992,1998)所指出的，如果价值溢价应
该奖励更高的违约风险，那么人们应该发现财务困境与市净率之间的负相关关系(Chen 和
Zhang，1998)。尽管如此，也有人认为，破产风险不一定与账面价格相关(Dichev，1998;
Griffin 和 Lemmon，2002)。为了验证这一假设，我们绘制了两个指标之间的滚动一年相
关性 12，以及图 12 中的价值五分位数百分比重叠 13。

10For the EMU universe, the number of real estate issuers in the index at each date is too low
to be able to build five quintile portfolios.
对于经济与货币联盟而言，指数中每个日期的房地产发行者数量太低，无法建立五分之一的投资组合。
11Q2 EDF outperforms Q1 EDF sorting in EMU for Utilities, Q4 EDF outperforms Q3 EDF in
North American Financials, Q4 EDF outperforms Q3 EDF that also dominates Q2 EDF for EMU
Healthcare, etc.12Taken at the issuer level for the whole sample.
11Q2 EDF 在电动车组中优于 Q1 EDF 排序，Q4 EDF 在北美金融中优于 Q3 EDF，Q4 EDF 优于 Q3
EDF，Q3 EDF 在电动车组医疗保健中也占优势，等等。
13The latter is defined as follows. Among the quintile of companies with the lowest book-to-
price(the value stocks, it corresponds to the percentage of firms that are also ranked in the
highest EDF quintile(i.e.the highest default risk).

13 后者定义如下。在账面价值最低的五分之一的公司(价值股票，它对应于同样排名在最高的 EDF 五
分之一的公司(即最高的违约风险)的百分比。

Figure 12: Price-to-book and EDF interaction(in%)

图 12: Price-to-book 和 EDF 交互(%)

100

100
We note that the correlation between the price-to-book and the EDF is not steady and
has oscillated between negative and positive territories since 2006. Although the
relationship is fairly stable in North America, it does not appear very significant
considering the average correlation close to 0%. In Europe, it decreased up to 2013,
and has stood in negative territory since 2009. As a matter of fact, the correlation
coefficient appears much more meaningful than in North America. Therefore, we
argue that the value premium might have rewarded investors for indeed bearing
default risk in the EMU since 2009. The results are more nuanced for North America
where we may claim that neither growth nor value are per-se synonymous with
financial distress. It appears that EDF is driven by other determinants,not captured by
the price-to-book alone. As far as the overlap is concerned, we note that on average,
a value stock is in 74% of the cases also in the highest EDF quintile(the highest
default frequency) in Europe in 2020. In North America, this figure drops to 51%.
Therefore,between 2006 and 2020, owning a value stock generally meant bearing
heightened default risk, but it was not clearly systematic. An investment holding the
first quintile value stocks (based on the price-to-book definition) that are not in the
highest EDF range may be a true winning bet as it could yield positive performance,
as illustrated by the positive returns reached by Q1 to Q4 EDF quintiles in both North
America and Europe.

我们注意到，股价与账面价值之间的相关性并不稳定，自 2006 年以来一直在负值和正值之

间波动。尽管这种关系在北美相当稳定，但考虑到平均相关性接近于 0% ，这种关系似乎并
不显著。在欧洲，这种关系在 2013 年前有所下降，自 2009 年以来一直处于负值区间。事
实上，相关系数似乎比北美更有意义。因此，我们认为，自 2009 年以来，价值溢价可能回
报了投资者，因为他们确实承担了欧洲货币联盟的违约风险。对于北美来说，结果更加微妙，
我们可以宣称，无论是增长还是价值本身都不是财务困境的同义词。EDF 似乎是由其他决定
因素驱动的，而不仅仅是由市净率决定的。就重叠而言，我们注意到，在 2020 年欧洲最高
的 EDF 五分位数(最高的违约频率)中，平均有 74% 的情况下存在价值股。在北美，这个数
字下降到 51% 。因此，在 2006 年到 2020 年之间，持有价值股通常意味着承担更高的违约
风险，但这并不是一个明确的系统。持有第一批五分之一价值股票(根据市净率定义)但不在
最高 EDF 价格区间的投资可能是一个真正的赢家，因为它可能产生正面的业绩，正如北美
和欧洲第一季度至第四季度 EDF 五分之一股票所达到的正回报所表明的那样。

Figure 13: Value quintile overlap(in%)- historical definition

图 13: 价值五分位重叠(%)-历史定义
North America Value
北美价值

In Figure 13, we test an alternative definition of the overlap. Instead of selecting

stocks based on raw price-to-book data, we use a stock’s price-to-book divided by its
historical average(calculated over the past year). The idea is to cancel out potential
sector bias that can translate into polarized price-to-book data, identify issuers on a
slippery slope and verify if this translates into an increase in EDF. Changing the
overlap definition offers a different picture, where North America and the EMU
appear much more alike than in Figure 12. For both universes, market stress was
usually accompanied by a rise in the overlap, implying that the issuers with price-to-
book significantly lower than their past average were more likely to embody higher
default risk. Finally, we also plot the historical average EDF of the MSCI EMU and
MSCI North America indexes versus a fictitious index built on the lowest price-to book
quintile in Figure 14. For the MSCI indexes, we kept the capitalization weighting
scheme(rebased according to our EDF coverage presented before), while for their
reduced version we aggregate stocks within the Q1 price-to-book quintile using an
equal weighting scheme.
在图 13 中，我们测试了重叠的另一种定义。我们不是根据原始的市净率数据来选择股票，
而是使用股票的市净率除以其历史平均值(计算于过去一年)。这样做的目的是消除可能导致
市净率数据两极分化的潜在行业偏差，确定处于下滑趋势的发行方，并验证这是否会导致
EDF 增加。改变重叠定义提供了一个不同的图景，其中北美和欧洲货币联盟看起来比图 12
更相似。对于两个领域而言，市场压力通常伴随着重叠上升，这意味着市净率远低于过去平
均水平的发行者更有可能体现出更高的违约风险。最后，我们还绘制了 MSCI EMU 和 MSCI
北美指数的历史平均 EDF 与基于最低价格构建的虚拟指数的比较图 14。对于摩根士丹利资
本国际指数，我们保留了资本化加权方案(根据我们之前提供的 EDF 覆盖率重新定位) ，而对
于它们的减少版本，我们使用相等的加权方案在 q1 市净率五分位数内累计股票。

Figure 14: Indexes average EDF(in%)

 图 14: 索引平均 EDF (%)

It is interesting to note that the MSCI EMU index has encompassed more default risk
than the MSCI North America index on average since the 2008 Global Financial
Crisis.However, since 2017, the gap has closed, and they now bear similar levels of
financial distress.As far as value indexes are concerned, for most of the time the EMU
index has been riskier than its North American counterpart. We witnessed a spike in
2009 in North America and in 2011 in the EMU. In Europe, until 2008, the value index
had a similar EDF to its broader index, but since then it has risen. A different story
goes on in North America, where the value index has consistently borne more default
risk than the MSCI North America index.However, since the GFC in North America,
and the sovereign debt crisis in Europe, the gap between the value index and their
broader version has closed. To summarize, investing in value stocks does not
systematically mean bearing greater default risk and thus, stock selection should be
of paramount importance for a value investor.

值得注意的是，自 2008 年全球金融危机以来，摩根士丹利资本国际经济与货币联盟指数

(MSCI EMU index)包含的违约风险平均高于摩根士丹利资本国际北美指数(MSCI North
America index)。然而，自 2017 年以来，这一差距已经缩小，它们现在承受着类似程度的
金融困境。就价值指数而言，在大多数时候，欧洲货币联盟指数的风险一直高于北美的同类
指数。我们目睹了 2009 年北美和 2011 年欧洲货币联盟的峰值。在欧洲，直到 2008 年，价
值指数与其更广泛的指数有着相似的 EDF，但是从那时起它已经上升了。北美的情况则不同，
价值指数一直承担着比摩根士丹利资本国际北美指数更大的违约风险。然而，自从北美的全
球金融危机和欧洲的主权债务危机以来，价值指数与其更广泛的版本之间的差距已经缩小。
总而言之，投资价值型股票并不意味着承担更大的违约风险，因此，股票选择对价值型投资
者来说应该是至关重要的。
4Explaining the recent performance of the value risk factor
解释价值风险因素的近期表现

4.1ESG risk analysis

4.1 ESG 风险分析

Some investors argue that value is simply too old fashioned. In a context marked by
major business model disruptions that were accelerated in 2020 with the COVID-19
crisis, value companies’ earnings outlooks may be overtaken by the glamorous
growth firms. Indeed,value stocks often belong to sectors such as Energy, Banking,
Utilities, Energy, Consumer Staples or Materials, which on top of being considered as
less glamorous, are also often pointed at for poor environmental performance. But is
value really brown and growth green? We would like to disentangle such common
beliefs from an empirical point of view.First, to grasp the broad ESG footprint of the
companies, we retrieved the Amundi monthly ESG scores for both the MSCI USA
Value and the MSCI EMU Value, and for their growthcounterparts since December
2010. Amundi ESG scores stand between −3 and+3 with a mean close to 0 at the
universe level. We computed the average ESG score at the index level, equally
weighting and cap-weighting underlying stocks14. Second, we turned towards their
exposure to carbon risk following the methodology developed by Roncalli et al.(2020).

一些投资者认为，价值观已经过时了。2019 冠状病毒疾病危机加速了重大商业模式的崩溃，
在这种背景下，价值型公司的盈利前景可能被迷人的增长型公司所取代。实际上，价值型股
票通常属于能源、银行、公用事业、能源、日用消费品或材料等行业，这些行业除了被认为
不那么吸引人之外，还经常被指责环保表现不佳。但是，价值真的是棕色的，增长是绿色的
吗？我们希望从经验的角度理清这些共同的信念。首先，为了掌握这些公司广泛的 ESG 足
迹，我们检索了自 2010 年 12 月以来的 Amundi 每月 ESG 分值，包括 MSCI 美国指数和
MSCI EMU 指数，以及它们的增长指数。Amundi ESG 得分介于 -3 和 + 3 之间，平均值接
近于 0。我们计算了指数水平的平均 ESG 分数，同等加权和上限加权的基础股票 14。其次，
我们根据 Roncalli 等人(2020)开发的方法转向他们对碳风险的暴露。

Figure 15: Amundi ESG score of MSCI equally-weighted indexes

图 15: MSCI 等权指数的 Amundi ESG 评分

USA GrowthUSA Value
美国成长价值

In Figure 15, we observe that the average ESG score of European companies is much
higher than in the U.S. We believe this result illustrates the “transatlantic divide”(Drei
et al., 2019). And it is worth noting how the gap consistently widened over the past
decade. In Europe, value stocks are, on average, less well ranked in terms of ESG than
growth stocks,while the opposite applies in the U.S. For the sake of completeness, the
same analysis is run on the cap-weighted indexes15 in Figure 16. As a matter of fact,
such examination is very informative for an investor that holds either the value or the
growth segment of the MSCI indexes. Switching the weighting scheme yields a much
different picture, especially in the

在图 15 中，我们观察到欧洲公司的 ESG 平均得分远高于美国。我们认为这个结果说明了

“跨大西洋分歧”(Drei et al。 ，2019)。值得注意的是，这个差距在过去十年中是如何不
断扩大的。在欧洲，平均而言，价值型股票在 ESG 方面的排名低于成长型股票，而在美国
则相反，为了完整起见，对图 16 中的上限加权指数 15 进行了同样的分析。事实上，对于持
有 MSCI 指数价值或增长部分的投资者来说，这样的检查是非常有用的。改变权重计划会产
生一幅完全不同的画面，尤其是在美国股市

14The average coverage of ESG data is 98% for EMU value, 97% for USA value, 91% for EMU
growth and 97% for USA growth. We ran the same analysis on distinct E, S and G pillars that
yielded the same ranking as in Figure 15 for the four strategies 15Sector breakdown is
available in Figure 24 in Appendix A.1.
14 ESG 数据的平均覆盖率分别为: EMU 值为 98% ，USA 值为 97% ，EMU 增长值为 91% ，USA 增长
值为 97% 。我们对不同的 e、 s 和 g 支柱进行了相同的分析，得出了与图 15 相同的排名。

EMU. Value is not as brown as some of its detractors claim. Actually, holding the MSCI
EMU Value index instead of the MSCI EMU Growth index actually results in a higher
ESG score. And the same applies in the U.S. To sum up, the U.S. growth companies
may not be as glamorous as many thought on the ESG front. If value stocks have on
average a lower ESG score in Europe, the greenness of the value investor’s portfolio
comes down to its stock picking and portfolio construction, as demonstrated by the
MSCI EMU Value index’s superior ESG score compared the MSCI EMU Growth index.
欧洲货币联盟。Value 并不像一些批评者所声称的那样糟糕。事实上，持有 MSCI EMU 价
值指数而不是 MSCI EMU 增长指数实际上会导致 ESG 得分更高。同样的道理也适用于美国。
总而言之，美国的成长型公司在 ESG 方面可能没有许多人想象的那么光鲜亮丽。如果在欧
洲，价值型股票的 ESG 得分平均较低，那么价值型投资者投资组合的绿色性取决于其选股
和投资组合的构建，正如摩根士丹利资本国际 EMU 价值指数的 ESG 得分优于摩根士丹利资
本国际 EMU 增长指数所显示的那样。

Figure 16: Amundi ESG score of MSCI cap-weighted indexes

图 16: MSCI 上限加权指数的 Amundi ESG 评分

EMU GrowthEMU Value USA GrowthUSA Value

EMU 增值美国增值

The impact of ESG investing on asset pricing may be significant because of two
effects:the valuation of extra-financial risks and the impact of investment flows. For
instance, it is obvious that the momentum risk factor may benefit from the behavior
of ESG investors. The value risk factor may also be impacted by ESG. Indeed, by
construction, the extra-financial risk of best-in-class ESG stocks is lower than the
extra-financial risk of worst-in-class ESG stocks. This implies a lower cost-of-capital,
implying a higher valuation for stocks with high ESG scores because the discount
factor decreases. Moreover, ESG investment flows reinforce the higher valuation of
these stocks.

环境、社会和治理投资对资产定价的影响可能很大，因为有两种影响: 额外金融风险的估值
和投资流动的影响。例如，动量风险因素显然可以从 ESG 投资者的行为中获益。价值风险
因素也可能受到 ESG 的影响。实际上，从建筑业来看，最好的 ESG 股票的额外金融风险低
于最差的 ESG 股票的额外金融风险。这意味着较低的资本成本，意味着较高的 ESG 分值的
股票估值较高，因为折现系数降低。此外，ESG 投资流强化了这些股票的高估值。

4.2Carbon risk analysis

4.2 碳风险分析

We now shift the focus towards the environmental footprint of value investing. More
specif-ically, we analyze its sensitivity to carbon risk. Here we define carbon risk as
the exposure to a long-short portfolio of brown minus green companies from the
MSCI World, using their carbon intensity for scopes 1, 2 and 3. We are then able to
calculate a carbon beta for all stocks(Roncalli et al., 2020, 2021). On one hand, a
positive carbon beta implies that a company is positively sensitive to carbon risk. This
is not necessarily because it has a car-bon intensive business model, but rather that
the latter is not equipped for a sudden change in the transition toward a green
economy and therefore the company would be negatively impacted. On the other
hand, a negative beta for a company means that it would benefit from a green
transition to a low-carbon economy.

我们现在将重点转向价值投资的环境足迹。更具体地说，我们分析其对碳风险的敏感性。在
这里，我们将碳风险定义为从摩根士丹利资本国际世界(MSCI World)的褐色减去绿色公司的
多空投资组合的风险，使用他们的碳强度范围 1、2 和 3。然后我们就可以计算出所有股票
的碳 β 值(Roncalli 等，2020,2021)。一方面，积极的碳贝塔意味着一个公司对碳风险是积
极的敏感。这不一定是因为它有一个碳密集型的商业模式，而是因为后者没有为向绿色经济
过渡的突然变化做好准备，因此公司将受到负面影响。另一方面，如果一家公司的测试结果
为负值，则意味着它将从向低碳经济的绿色转型中受益。
In Figure 17, we plotted the resulting carbon beta of the MSCI Value and Growth
indexes covering the same geographical scope. We observe that value has
traditionally been

在图 17 中，我们绘制了覆盖相同地理范围的 MSCI 价值和增长指数的碳 β 值。我们观察到

这个价值传统上是

Figure 17: Average carbon beta of MSCI cap-weighted indexes

图 17: 摩根士丹利资本国际上限加权指数的平均碳 β 值
better equipped for a green transition of the economy in the EMU, with a lower
carbon beta than growth. Actually, in this region, average growth betas oscillated
around zero, and were particularly steady in the 2016-2019 period. As of the end of
2020, value and growth average carbon betas are very much equal for the EMU. It is a
different story on the other side of the Atlantic. On average, carbon betas are higher
than in Europe for value, just below zero, highlighting a risk for underlying
companies if the green transition were to accelerate.While in the U.S., value carbon
beta has tended to worsen since the end of 2017, the growth index has enjoyed
constant improvements since 2015. All in all, whereas the gap in carbon sensitivity
between growth and value is closing in the EMU, it is actually widening in the U.S.

经济与货币联盟(EMU)更好地为经济的绿色转型做好了准备，碳 β 值低于增长值。实际上，
在这个地区，平均增长贝塔值在零附近波动，并且在 2016-2019 年期间特别稳定。到 2020
年底，经济与货币联盟的平均碳贝塔值和增长值非常相等。在大西洋的另一边，情况就完全
不同了。平均而言，碳贝塔的价值高于欧洲，略低于零，突显出如果绿色转型加速，潜在公
司面临的风险。在美国，自 2017 年底以来，碳排放价值贝塔趋于恶化，而自 2015 年以来，
增长指数一直在持续改善。总而言之，尽管欧洲货币联盟的增长和价值之间的碳敏感度差距
正在缩小，但实际上美国的差距正在扩大。

4.3The small cap effect

4.3 小帽效应

We now ask whether recent value underperformance has widespread to its smallest
players.Small caps have been broadly praised by investors over the past decades.
Indeed, it is a well-documented fact that firms with small market values – that are
therefore less liquid – tend to outperform the largest ones(Banz, 1981; Fama and
French, 1992). And it is no different within the value world. As Bauman et al.(1998)
demonstrated, the value premium gets bigger as the market capitalization gets
smaller. With the aim of verifying if this still holds between 2000 and 2020, we plot
the difference between the performance of the MSCI Small Cap Value index(EMU or
USA) versus the respective performance of their generic value index in Figure 18.

我们现在要问的是，近期的价值表现不佳是否已经广泛地影响到了最小的参与者。在过去的
几十年里，小型股一直受到投资者的广泛称赞。事实上，一个有据可查的事实是，市值较小
的公司——因此流动性较差——往往表现优于最大的公司(Banz，1981; Fama 和
French，1992)。在价值世界里也没有什么不同。正如鲍曼等人(1998)所证明的那样，随着
市值的缩小，价值溢价会越来越大。为了验证这种情况在 2000 年至 2020 年之间是否仍然
存在，我们绘制了摩根士丹利资本国际小盘股价值指数(EMU 或 USA)的表现与其通用价值
指数各自表现之间的差异，如图 18 所示。
Investing in the small cap value segment instead of the broader one between 2000
and 2010 has clearly been a winning bet. After 2010, results are more nuanced, and
outperfor-mance is much more volatile. While it still paid off in Europe, performance
in the U.S. is more mixed. However, after reaching lows at the beginning of 2020, a
turnaround occurred in the U.S. at the end of 2020 with a sharp rise in the
performance of small cap value firms.
在 2000 年至 2010 年之间，投资于小盘股价值部门，而不是更广泛的部门，显然是一个成
功的赌注。2010 年之后，结果变得更加微妙，表现更加不稳定。虽然在欧洲市场仍有回报，
但在美国市场的表现则更加喜忧参半。然而，在 2020 年初触底之后，美国在 2020 年底出
现了转机，小盘股价值型公司的业绩大幅上升。
If small cap value stocks can add marginal performance gain compared to an
exposure to the broader segment, does it mean that it can also beat the small cap
growth universe? To answer this, we plot the difference between the performance of
the MSCI Small Cap Value index(EMU and USA) versus their growth counterpart(Small
Cap EMU Growth and Small Cap USA Growth) in Figure 19. When comparing the value
and growth performance in the

如果小盘股能够增加边际业绩收益，而不是进入更广泛的领域，这是否意味着它也能击败小
盘股增长领域？为了回答这个问题，我们在图 19 中绘制了 MSCI 小型股价值指数(EMU 和
USA)与其增长对应指数(smallcapemu Growth 和 smallcapusa Growth)之间的表现差异。
当比较价值和增长表现时

Figure 18: Small cap value performance(in%) with respect to the value index
图 18: 相对于价值索引的小市值性能(以% 为单位)

small cap universe, similar sustained underperformance can be witnessed on both

sides of the Atlantic, despite a rebound from lows at the end of 2020. In fact, it
appears that small cap value has suffered as much as the broader value index versus
growth. Hence, choosing the small cap segment of value can bring benefits to an
investor versus the broader index,however it has not been the perfect fix for growth
outperformance.

尽管从 2020 年底的低点反弹，但在大西洋两岸都可以看到类似的持续表现不佳。事实上，

看起来小盘股价值遭受的损失和更广泛的价值指数相对于增长的损失一样多。因此，相对于
更广泛的指数，选择小盘股价值可以给投资者带来好处，但这并不是解决增长优势的完美方
案。

Figure 19: Small cap value performance(in%) with respect to the growth index

图 19: 相对于增长指数的小市值业绩(以% 为单位)

4.4The intangible puzzle

4.4 无形之谜

As explained in Section Two, we are keen to approach value performance with an

agnostic view, and as such we use baseline book-to-market data as proposed by
Fama and French (1992). However, we recognize that it can be enhanced, namely with
the inclusion of in-tangible assets, that have been growing at a solid pace in recent
decades. Ocean Tomo16estimates that intangible assets accounted for 17% of the
S&P 500’s assets in 1975. In 2020,this figure stands at 90%. Academic literature
pointing at the outdated strict definition of 16Ocean Tomo’s Intangible Asset Market
Value Study 2020, https://www.oceantomo.com.

正如在第二部分中所解释的，我们热衷于用一种不可知论的观点来看待价值绩效，因此我们
使用 Fama 和 French (1992)提出的基线账面到市场的数据。然而，我们认识到，可以通过
纳入近几十年来稳步增长的无形资产来加强这一点。Ocean tomo16 估计 1975 年无形资产
占标准普尔 500 指数资产的 17% 。到 2020 年，这一数字为 90% 。学术文献指出，16
ocean Tomo 的无形资产市场价值研究 2020，https://www.oceantomo. com。

traditional value has flourished, alongside the transformation of business models,

with more and more knowledge intensive industries(Lev and Zarowin, 1999). Such a
switch toward an increasingly services-centered economy means that intellectual
property(R&D, patents,etc.) but also design, software development and talent
recruitment have risen for many firms, however, this does not reflect in traditional
accounting metrics such as book value.Brand recognition is also another key aspect
of intangible assets. Although intangible as-sets such as patents, software or
franchise agreements are identifiable, and therefore can be priced and capitalized,
those generated internally such as customer loyalty, brand recogni-tion or customer
lists are trickier to account for. Implications are twofold for an investor.First, omitting
the latter might cause her to underestimate the firm’s book value, which in turn could
distort her view on its valueness once compared to stock price. For instance, a
company that makes large investments in intangible assets could be seen as cheaper
when these expenses are reintegrated into the book value. Second, Chan et al.(2001)
and Edmans (2011) showed that spending on intangibles(such as R&D or employee
satisfaction) may drive stock returns upwards. Amenc et al.(2020) propose then to use
the intangible-adjusted book-to-price in order to more accurately capture the value
premium, a metric that has the advantage of not altering the risk profile of the old
value strategy. By adapting the book value with intangibles, they find a stronger
value premium, even after controlling for other traditional CAPM factors.
传统价值随着商业模式的转变而蓬勃发展，知识密集型产业越来越多(Lev and
Zarowin，1999)。这种向日益以服务为中心的经济转变意味着许多公司的知识产权(研发、
专利等)以及设计、软件开发和人才招聘都有所增加，然而，这并没有反映在账面价值等传
统会计指标上。品牌认知也是无形资产的另一个关键方面。虽然专利、软件或特许经营协议
等无形资产是可以识别的，因此可以进行定价和资本化，但内部产生的客户忠诚度、品牌认
知度或客户名单等则较难解释。对于投资者来说，影响是双重的。首先，忽略后者可能会导
致她低估公司的账面价值，这反过来会扭曲她对公司价值的看法，一旦与股价相比。例如，
一家公司在无形资产上进行了大量投资，当这些费用重新计入账面价值时，可能会被视为更
便宜。其次，Chan 等人(2001)和 Edmans (2011)表明，在无形资产上的支出(如研发或员工
满意度)可能会推高股票回报率。Amenc 等人(2020 年)建议使用无形调整账面价格，以便更
准确地捕捉价值溢价，这一指标的优点是不改变旧的价值战略的风险状况。通过调整账面价
值与无形资产，他们发现一个更强的价值溢价，即使控制了其他传统的 CAPM 因素。

Hence, we believe that accounting for intangibles can provide insightful information
for an investor sorting companies on the value/growth spectrum, although it is
unlikely to lead to a major reshuffle of the traditional price-to-book ranking. Indeed,
our simulations show that the value risk premium increases by about 25 bps per year
for both the European and American stock universes. If we compare stock selection
with and without intangible assets, we notice some neutral stocks are reintegrated
into value stocks, which explains the additional performance. These neutral stocks
are those that present a high value of intangible assets. Therefore, we observe a
switch in terms of ranking between these neutral stocks with high intangible assets
with some value stocks with low intangible assets. Nevertheless, deep value stocks
stay in the universe of value stocks, and growth stocks remain growth stocks.

因此，我们认为，对无形资产进行会计核算，可以为投资者在价值/增长谱上对公司进行分
类提供有见地的信息，尽管这不太可能导致传统的市净率排名发生重大变化。事实上，我们
的模拟显示，欧洲和美国股市的价值风险溢价每年都会增加 25 个基点。如果我们比较有无
形资产和无形资产的股票选择，我们注意到一些中性股票被重新整合到价值股票中，这就解
释了额外的表现。这些中性股票是那些具有高无形资产价值的股票。因此，我们观察到这些
中性股票在无形资产高的股票和价值股票无形资产低的股票之间的排名变化。尽管如此，深
价值股仍然属于价值股的范畴，而成长型股仍然属于成长型股。
5Conclusion

5 结论

Due to its inter-linkages with the economic environment, value clearly has some flaws
that can explain its recent performance lag, e.g. its ties to inflation and market
uncertainty, the deep segment that has not cleared due to low interest rates, a mixed
picture on the ESG front, a faltering business model, etc. Still, even against this
backdrop, everything in value should not be thrown away, and it may enhance
portfolio returns. First, we saw that a growing body of literature has shed the light on
the importance of accounting for intangible assets in book value. Improving how
value is defined based on trends in business models, that tend to be less capital
intensive, asset-light and more services-centered, may enable better value capture.
Second, an investor could depart from the pure essence of value investing and
neutralize the inherent sector bias incumbent in value. Finally, value can add
substantial diversification within a multi-factor portfolio. For instance, it is generally
recognized that it is negatively correlated with other factors such as low-volatility or
momentum factors.Additionally, combined with the quality factor, it may act as a
strong buffer against default risk and value trap issues.

由于其与经济环境的相互联系，价值显然存在一些缺陷，可以解释其最近的表现滞后，例如
与通货膨胀和市场不确定性的联系，由于低利率而未能清理的深层次部分，环境、社会和治
理方面的情况好坏参半，商业模式摇摇欲坠等等。尽管如此，即使在这样的背景下，所有的
价值都不应该被抛弃，它可能会提高投资组合的回报。首先，我们看到越来越多的文献阐明
了无形资产账面价值会计的重要性。根据商业模式的趋势来改进价值的定义方式，这些商业
模式往往资本密集程度较低，资产较轻，以服务为中心，可以更好地捕捉价值。其次，投资
者可以脱离价值投资的纯粹本质，中和固有的价值部门偏见。最后，价值可以在一个多因素
投资组合中增加实质性的多样化。例如，人们普遍认为它与其他因素如低波动性或动量因素
呈负相关。此外，与质量因素相结合，它可能成为对违约风险和价值陷阱问题的强大缓冲。
Despite the ongoing technological and structural changes in the economy(social
media dominance, working from home, etc.), we believe that value firms may still
have bright days ahead. As we learned when the dot-com bubble burst, the
emergence of a new economy

尽管经济中正在发生技术和结构性的变化(社交媒体占据主导地位，在家工作等等) ，我们仍
然相信价值型公司前途一片光明。正如我们在互联网泡沫破灭时所了解到的那样，一个新经
济的出现

(increasingly widespread internet use at this time) does not mean that companies
operating in this trendy sector will enjoy exponential earnings in the long term nor
that they will all survive. We also learned that these firms can be genuinely over-
valued. Additionally,this shift towards teleworking technologies and increased social
media presence has been partially forced by the COVID-19 pandemic and therefore
their momentum may fade out once the health crisis is over, and the sector has
streamlined. In October 2020, the American Congress raised doubts about the impact
of the FAANG monopoly on consumers. It was quickly followed in December by the
European Commission, which is keen to prevent anti-competitive practices. Future
legislation imposed on tech leaders could be a catalyst for the much-awaited value
rally.
(目前，互联网的使用越来越普遍)并不意味着在这个时尚领域运营的公司长期将享受指数收
益，也不意味着它们都能生存下去。我们还了解到，这些公司可能真的被高估了。此外，这
种向远程工作技术和增加社交媒体存在的转变在一定程度上是由于 2019 冠状病毒疾病大流
行所迫，因此，一旦卫生危机结束，卫生部门精简后，其势头可能会减弱。2020 年 10 月，
美国国会对 FAANG 垄断对消费者的影响提出了质疑。12 月，欧盟委员会紧随其后，热衷于
防止反竞争行为。未来针对科技行业领导者的立法可能成为期待已久的价值回升的催化剂。
We have seen that a brightening of the economic outlook, as illustrated in November
2020by the hopes placed in the COVID-19 vaccines, can cause strong value rallies.
Although an improved economic outlook is usually accompanied by rising interest
rates, considering the current monetary stance adopted by the major central banks,
we argue that this transition channel may not be as effective as in the past. However,
we also think that today is different than yesterday. We have not experienced
equivalent low rate levels for such a prolonged period. In addition, the fact that
governments are starting a massive fiscal stimulus push may bolster inflation
expectations upward, which should benefit value sectors such as Bank-ing, Energy or
Industrials. Improvements in consumer and business confidence would also push in
that direction. The low-rate environment may have become the new normal, but
governments taking over with an unprecedented spending boost, if it materializes by
fueling inflation expectations, could be real a game changer for value.
我们已经看到，正如 2020 年 11 月对 2019 冠状病毒毒疾病疫苗寄予的希望所表明的那样，
经济前景的光明可以引起强劲的价值回升。虽然经济前景改善通常会伴随利率上升，但考虑
到主要中央银行目前采取的货币政策，我们认为这个过渡渠道可能不如过去有效。然而，我
们也认为今天与昨天不同。在如此长的时间里，我们从未经历过同样的低利率水平。此外，
各国政府正在启动大规模财政刺激计划，这一事实可能会推高通胀预期，从而使银行、能源
或工业等价值型行业受益。消费者和企业信心的改善也将推动这一方向。低利率环境可能已
经成为新的常态，但如果政府通过刺激通货膨胀预期来实现空前的支出增长，那么它将真正
改变游戏规则的价值。
Nevertheless, the negative performance of the value risk factor over the last 10 years
should encourage caution about its future performance. From an ex-post point of
view,we can always find reasons to explain this underperformance. From an ex-ante
point of view, we can also find arguments to support the value risk premium. For
some years, many professionals and academics have predicted a new golden age for
value investing, but it has been slow to arrive. Moreover, the new value risk factor
may be very different from the value risk factor of the 2000s, which was perceived as
a long-term reversal bet or a sector long/short strategy. The empirical structural
relationships between value investing and economic factors must fit in with the new
trends of financial markets because of trading facilities, ESG investing and the
digitalization of the economy. Moreover, computing the value of a stock by using
fundamental data that are easily accessible to everybody may be a challenge, and
perhaps requires more reactive alternative data.
然而，价值风险因素在过去 10 年中的负面表现应该鼓励人们对其未来表现持谨慎态度。从
事后的观点来看，我们总能找到理由来解释这种表现不佳的现象。从事前的观点来看，我们
也可以找到支持价值风险溢价的论据。多年来，许多专业人士和学者都预言价值投资将迎来
一个新的黄金时代，但是这个时代的到来却很缓慢。此外，新的价值风险因素可能与 2000
年代的价值风险因素大不相同，2000 年代的价值风险因素被视为长期逆转押注或行业多/空
策略。价值投资与经济因素之间的经验结构关系必须适应金融市场的新趋势，因为交易设施、
环境、社会治理投资和经济数字化。此外，通过使用每个人都能轻易获得的基本数据来计算
股票价值可能是一个挑战，也许需要更多的反应性替代数据。

References
参考文献
Amenc, N., Goltz, F., and Luyten, B.(2020), Intangible Capital and the Value Factor:Has
Your Value Definition Just Expired?, Journal of Portfolio Management, 46(7), pp.83-99.
无形资本与价值因素: 你的价值定义刚刚过期了吗？投资组合管理杂志，46(7) ，83-99 页。

Basu, S.(1983), The Relationship between Earnings’ Yield, Market Value and Return for
NYSE Common Stocks: Further Evidence, Journal of Financial Economics, 12(1),
pp.129-156.

《纽约证券交易所普通股收益率、市值和收益率之间的关系: 进一步的证据》 ，《金融经济

学杂志》 ，第 12(1)页，第 129-156 页。

Bauman, W.S., Conover, C.M., and Miller, R.E.(1998), Growth versus Value and Large-
cap versus Small-cap Stocks in International Markets, Financial Analysts Journal,54(2),
pp. 75-89.

鲍曼，w.s. ，Conover，c.m. 和 Miller，r.e. (1998) ，《国际市场上的增长与价值及大盘股

与小盘股》 ，《金融分析师杂志》 ，第 54(2)页，第 75-89 页。

Banz, R.W.(1981), The Relationship between Return and Market Value of Common
Stocks,Journal of Financial Economics, 9(1), pp. 3-18.

《普通股收益率与市场价值的关系》 ，《金融经济学杂志》 ，第 9(1)页，第 3-18 页。

Campbell, J.Y., and Vuolteenaho, T.(2004), Bad Beta, Good Beta, American Economic
Review, 94(5), pp. 1249-1275.

《美国经济评论》 ，第 94(5)页，第 1249-1275 页。

Chan, L.K.C., Hamao, Y., and Lakonishok, J.(1991), Fundamentals and Stock Returns in
Japan, Journal of Finance, 46(5), pp. 1739-1764.

(1991) ，《日本的基本面与股票回报》 ，金融期刊，46(5)页，1739-1764。

Chan, L.K.C., Lakonishok, J., and Sougiannis, T.(2001), The Stock Market Valuation of
Research and Development Expenditures, Journal of Finance, 56(6), pp. 2431-2456.

(2001) ，《研究与发展支出的股票市场估值》 ，《金融期刊》 ，第 56(6)页，第 2431-

2456 页。

Chen, N.F., and Zhang, F.(1998), Risk and Return of Value Stocks, Journal of
Business,71(4), pp. 501-535.

《股票的风险与价值回报》 ，《商业杂志》 ，第 71(4)页，第 501-535 页。

Chordia, T., Roll, R., and Subrahmanyam, A.(2001), Market Liquidity and Trading
Activity, Journal of Finance, 56(2), pp. 501-530.

(2001) ，《市场流动性与交易活动》 ，《金融期刊》 ，第 56(2)页，第 501-530 页。

Dichev, I.D.(1998), Is the Risk of Bankruptcy a Systematic Risk?, Journal of

Finance,53(3), pp. 1131-1147.

破产风险是一种系统性风险吗? ，《金融期刊》 ，第 53(3)页，第 1131-1147 页。

Drei, A., Le Guenedal, T., Lepetit, F., Mortier, V., Roncalli, T., and Sekine,T.(2019), ESG
Investing in Recent Years: New Insights from Old Challenges,
SSRN,www.ssrn.com/abstract=3683469.
Lepetit，f. Mortier，v. Roncalli，t. 和 Sekine，t. (2019) ，《近年来的 ESG 投资: 来自旧
挑战的新见解》 ，SSRN，www.SSRN。

Edmans, A.(2011), Does the Stock Market Fully Value Intangibles? Employee
Satisfaction and Equity Prices. Journal of Financial Economics, 101(3),pp. 621-640.

股票市场是否完全重视无形资产? 员工满意度与股票价格。《金融经济学杂志》 ，101(3) ，

第 621-640 页。

Fama, E.F., and French, K.R.(1992), The Cross-section of Expected Stock

Returns,Journal of Finance, 47(2), pp. 427-465.

《预期股票回报的横截面》 ，《金融期刊》 ，第 47(2)页，第 427-465 页。

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

《股票和债券回报中的常见风险因素》 ，《金融经济学杂志》 ，第 33(1)页，第 3-56 页。

Fama, E.F., and French, K.R.(1998), Value versus Growth: The International
Evidence,Journal of Finance, 53(6), pp. 1975-1999.

《价值与增长: 国际证据》 ，《金融学杂志》 ，第 53(6)页，1975-1999。

Fama, E.F., and French, K.R.(2020), The Value Premium, SSRN,

www.ssrn.com/abstract=3525096.

Fama，e.f。 and French，k.r。(2020) ，The Value Premium，SSRN，www.SSRN.

com/abstract = 3525096。

Graham, B., and Dodd, D.L.(1934), Security Analysis, Whittlesey House, New York(sixth
edition, 2008, McGraw-Hill).

格雷厄姆 b 和多德 d.l. (1934) ，《安全分析》 ，惠特尔西出版社，纽约(第六版，2008 年，

麦格劳-希尔)。

Griffin, J.M., and Lemmon, M.L.(2002), Book-to-market Equity, Distress Risk, and Stock
Returns, Journal of Finance, 57(5), pp. 2317-2336.
(2002) ，《账面到市场股权，困境风险和股票回报》 ，《金融期刊》 ，57(5)页，第 2317-
2336 页。

Grinblatt, M., Titman, S., and Wermers, R.(1995), Momentum Investment Strate-gies,
Portfolio Performance, and Herding: A Study of Mutual Fund Behavior, American
Economic Review, 85(5), pp. 1088-1105.

《动量投资策略、投资组合绩效与羊群效应: 共同基金行为研究》 ，《美国经济评论》 ，第

85(5)页，第 1088-1105 页。

Jegadeesh, N., and Titman, S.(1993), Returns to Buying Winners and Selling
Losers:Implications for Stock Market Efficiency, Journal of Finance, 48(1), pp. 65-91.
《回归买入赢家和卖出输家: 对股票市场效率的影响》 ，《金融期刊》 ，第 48(1)页，65-91
页。

Lakonishok, J., Shleifer, A., and Vishny, R.W.(1994), Contrarian Investment, Extrap-
olation, and Risk, Journal of Finance, 49(5), pp. 1541-1578.

《反向投资，额外投资与风险》 ，《金融杂志》 ，第 49(5)页，1541-1578 页。

Lev, B., and Srivastava, A.(2019), Explaining the Recent Failure of Value
Investing,SSRN, www.ssrn.com/abstract=3442539.

Lev，b，and Srivastava，a。(2019) ，《解释价值投资最近的失败》 ，

SSRN，www.SSRN. com/abstract = 3442539。

Lev, B., and Zarowin, P.(1999), The Boundaries of Financial Reporting and How to
Extend Them, Journal of Accounting research, 37(2), pp. 353-385.

列夫，b，和 Zarowin，p。(1999) ，财务报告的边界和如何扩展它们，会计研究杂志，

37(2) ，353-385 页。

Roncalli, T.(2017), Alternative Risk Premia: What Do We Know?, in Jurczenko, E.(Ed.),

Factor Investing and Alternative Risk Premia, ISTE Press – Elsevier, pp. 227-264.

Roncalli，t。(2017) ，另类风险溢价: 我们知道什么？，在 Jurczenko，e。(Ed。)因素投资

和另类风险溢价，ISTE 出版社-Elsevier，第 227-264 页。

Roncalli, T., Le Guenedal, T., Lepetit, F., Roncalli, T., and Sekine, T.(2020),Measuring
and Managing Carbon Risk in Investment Portfolios, SSRN,
www.ssrn.com/abstract=3681266.

Roncalli，t。 Le Guenedal，t。 Lepetit，f。 Roncalli，t。和 Sekine，t。(2020) ，《投

资组合中碳风险的测量与管理》 ，SSRN，www.SSRN. com/abstract = 3681266。

Roncalli, T., Le Guenedal, T., Lepetit, F., Roncalli, T., and Sekine, T.(2021),The Market
Measure of Carbon Risk and its Impact on the Minimum Variance Portfolio,SSRN,
www.ssrn.com/abstract=3772707.

《碳风险的市场测量及其对最小方差组合的影响》 ，SSRN，www.SSRN. com/abstract =

3772707。

Schro¨der, D., and Esterer, F.(2011), A New Measure of Equity and Cash Flow
Duration:The Duration-based Explanation of the Value Premium Revisited, Journal of
Money,Credit and Banking, 48(5), 857-900.

Schro der，d，and Esterer，f. (2011) ，一种衡量股权和现金流持续时间的新方法: 对价值

溢价的基于持续时间的解释，《货币、信贷和银行期刊》 ，48(5) ，857-900。

Stagnol, L., and Taillardat, B.(2017), Analysing the Exposure of Low-Volatility Equity
Strategies to Interest Rates, SSRN, www.ssrn.com/abstract=3074763.
Stagnol，l. 和 Taillardat，b. (2017) ，《分析低波动性股票策略对利率的影响》 ，
SSRN，www.SSRN. com/abstract = 3074763。

A Appendix
一个附录
A.1Additional results

A. 1 附加结果

Figure 20: Distribution of monthly HML returns(in%)

图 20: 每月 HML 报表的分布(%)

Figure 21: Distribution of deep value monthly returns(in%)
图 21: 深度价值月回报的分布(%)
 Figure 22: Boxplots of MSCI EMU constituents’ book-to-market
图 22: MSCI 经济与货币联盟成员的账面到市场的盒子图
3.53.5
3.53.5
 Figure 23: Boxplots of MSCI USA constituents’ book-to-market

图 23: 摩根士丹利资本国际美国分公司成员图书到市场的 boxplot

Figure 24: MSCI indexes sectors composition(in%)- Capitalization weighted

图 24: 摩根士丹利资本国际指数类别构成(%)-资本化加权
(c) Growth EMU(d) Growth USA

(c)增长动车组(d)增长美国

Figure 25: Average quarterly returns(in%) of sorted portfolios by GICS sector

图 25: 按 GICS 部门分类的投资组合的平均季度收益率
665
665
Understanding the Performance of the Equity Value Factor