Does socially responsible mutual fund performance vary over the business cycle? New insights on the effect of idiosyncratic SR features
| DOI | http://doi.org/10.1111/beer.12196 |
| Author | Diego Víctor de Mingo‐López,Juan Carlos Matallín‐Sáez,Amparo Soler‐Domínguez,Emili Tortosa‐Ausina |
| Date | 01 January 2019 |
| Published date | 01 January 2019 |
Business Ethic s: A Eur Rev. 2019;28:71– 98. wileyonlinelibrary.com/journal/beer
|
71
© 2018 John Wiley & So ns Ltd
1 | INTRODUCTION
Social respon sibility matte rs. This is especi ally true in finan ce, as
socially resp onsible investing (S RI) is gaining mome ntum. The dy‐
namism of the mutu al fund industr y has led to a prolifer ation of
socially resp onsible (SR) fund s, aiming to satisf y investors’ prefe r‐
ences in terms of ris k and return but also cover ing objectives tha t are
not purely fina ncial (Benson & Hum phrey, 2008; Ferruz , Muñoz, &
Vargas, 2010), such as envi ronmental, social and governa nce (ESG),
religious and ethical attributes.
Most of the stud ies analysing the b ehaviour of SR funds ove r
time compare th eir performan ce with that of conventio nal mutual
funds during th e same period. However, not all SR funds inve st ac‐
cording to the sam e SRI criteria. For in stance, stock s held by envi‐
ronmental fu nds would be expec ted to differ from s tocks religious
funds invest in. A s the funds’ return s mainly depend on the e volution
of the asset classe s held in their por tfolio (Sharpe, 1992), SR funds
belonging to different categories should yield different results, lead‐
ing to erroneou s overall perfor mance conclusions if they are not
treated separately.
Received:5Febr uary2017
|
Revised:29Marc h2018
|
Accepted:30May2 018
DOI: 10 .1111/bee r.12196
ORIGINAL ARTICLE
Does socially responsible mutual fund performance vary over
the business cycle? New insights on the effect of idiosyncratic
SR features
Juan Carlos Matallín‐Sáez1 | Amparo Soler‐Domínguez1 |
Diego Víctor de Mingo‐López1 | Emili Tortosa‐Ausina2
1Departme nt of Finance and
Accounting, U niversitat Jaume I , Castellón
de la Plana, Spain
2Department of Economics and
Ivie, Univer sitat Jaume I, Cas tellón de la
Plana, Spain
Correspondence
Amparo Soler‐Domínguez, Department of
Finance and Acc ounting, Univers itat Jaume
I, Castell ón de la Plana, E‐12071 Castelló n,
Spain.
Email: amparo.soler@uji.es
Funding information
the Universi tat Jaume I, Grant /Award
Number: UJI‐ B2017‐14, E‐2014‐10 and
Predoc‐2014‐04; Spanish Ministerio de
Economía y Competitividad, Grant/Award
Number: ECO2014‐55221‐P, ECO2017‐
85746-P,2014,046,2014and283;
Generalitat Valenciana
Abstract
This study analys es the performa nce and market timing of US sociall y responsible
(SR) mutual funds in r elation to business cycle r egime shifts and dif ferent grouping
criteria: Ethical st rategy focus, SR at tributes scores and Mo rningstar categor y.
Different metho dologies are applied and results high light the importance of consid‐
ering specific be nchmarks related to th e investment style i n evaluating the SR fund
performance. O ur results show that , in aggregate, the abno rmal performan ce of SR
funds is negative and sig nificant in expansion periods, but no s ignificant differ ences
are found in recession p eriods. When specif ic benchmarks are co nsidered, perfor‐
mance improves in recessio n periods, part icularly for environme ntal funds, those
with high SR attrib utes scores, and funds from the nine Mor ningstar style box cate‐
gories. Market timing of S R funds takes positive val ues and is partiall y significant.
Previous evidence of negative timing after a recent financial crisis vanishes when
specific benchm arks are considered. For co mparative purpose s the performance of
conventional US mutual fu nds is also analysed. Th ere are no significant di fferences
between the per formance of SR and conventional mutua l funds when a fair compari‐
son is made within the s ame style categories. When all t he SR funds are considered,
they underper form conventional funds i n expansion sub‐periods , but in recession
sub‐periods they pe rform better, althoug h the differences obs erved are not
significant.
72
|
MATALLíN‐SáEZ ET AL.
In this line, the pr esent study cont ributes towards a be tter un‐
derstand ing of the performance of SR mutu al funds over time. Our
main (but not exclus ive) objective is to analy se the difference s in per‐
formance within SR groups. A proper cla ssification of the fu nds in the
sample is there fore crucial to the analysis. To addre ss this issue, we
consider three different classifications (the funds’ ethical strategy
focus, their SR attributes scores and their Morningstar categories),
in order to accurat ely distinguish amongst SR f unds with heteroge‐
neous investme nt styles in the perfor mance evaluation process. As
Figure 1 shows, th e study therefore contribute s to the literature in
several ways.
Regarding the com parison of SR fund perfo rmance, and in order
to differentia te the value added by manager s’ investment skills from
the behaviour of t he asset classes (or st yles) held in the fund port‐
folios, we apply t wo widely‐used approach es from the conventional
mutual fund lite rature: Asses sing funds’ risk‐adjuste d abnormal re‐
turns (the alpha); a nd evaluating manager s’ market timing skills (ma n‐
agers’ abilitie s to forecast the market). Acc ordingly, the comparative
performan ce analysis is conduc ted for SR funds c ategorised in dif‐
ferent groups. T his analysis is also p erformed for a la rge sample of
conventional fun ds with the purpose of highlighti ng any similarities
with previous ar ticles about SR fund s.
Our second contr ibution, in contr ast to previous st udies on
market conditio ns (e.g., Nofsinger & Varma , 2014), is to consider
several econom ic states (expansio ns and recessions) in t he econo‐
metric analys es, following eviden ce reported in Kosowski ( 2011) and
Kacperczyk, Nieuwerburgh, and Veldkamp (2014), amongst others.
As mutual fun ds seem to perform diffe rently in relation to the busi‐
ness cycle, it is im portant to asse ss their perfo rmance in differ ent
economic peri ods to obtain an accurate over view of fund behaviour
in each categor y. This is especially necessa ry in the case of environ‐
mental funds , which have been aff ected by the poor evo lution of
the green sec tor (as explained by Prior (20 09) and Sabbaghi (2010)).
Our initial resu lts confirm our ex pectations . SR funds belong‐
ing to different c ategories diff er considerably i n terms of perfor‐
mance. For insta nce, Carhar t’s (1997) four‐factor model reveal s
that environme ntal funds behave di fferently fro m other SR funds.
Specifica lly, before the onset of the f inancial crisis env ironmental
funds were the best performers, but their performance worsened
considerably after this point. Also, SR f unds investing in stocks wi th
lower SR charac teristics are b etter able to add va lue for investors
than managers of h igh‐scoring SR funds, who m ay face greater diffi‐
culties in manag ing a more restricte d portfolio. When we e xtend the
four‐factor model to in clude the market timing varia bles defined by
Treynor and Mazuy (1966) and Henrik sson and Merton (1981), which
is related to our thi rd contribution, w e also observe th at SR funds
with a higher comm itment to socially r esponsible inves tments (i.e.,
high‐scoring SR f unds) seem to exhibit wo rse timing abilit ies than
low‐scoring SR fu nds.
Despite the aforementioned evidence, we would expect a bias
in the SR fund per formance resu lts due to the omissi on of spe‐
cific relevant benchmarks on applying Carhart’s (1997) four‐factor
model (Matall ín‐Sáez, 2006; Pás tor & Stambaugh, 2 002). In other
words, part of t he fund’s risk‐adjusted ab normal returns at tributed
FIGURE 1 Contribution to S R literature and manageri al implications
Are there any differences in the performance of
different groups of SR funds?
Bauer et al. (2005), Renneboog et al. (2008b),
Signori (2009), Nofsinger and Varma (2014),
Leite and Cortez (2014a) and Ur Rehman et al.
(2016), among others:
In general there are no significant differences.
Areal et al. (2013) and Nofsinger and Varma
(2014) found that SR funds seem to perform
better than their conventional counterparts in
crisis times.
In aggregate both SR and conventional funds
achieve a negative performance.
In line with the literature, in aggregate there are
no differences when a fair comparison is ma de
within the same style categories of both types of
funds.
When specific SR fundsare considered,they
significantly underperform conventional funds
in times of expansion.
Applying a extended model considering
idiosyncratic SR features, SR fund performance
improves significantlyin recessi on periods,
although it is not enough to infer significant
differences from the co nventional funds. This
result is in line with Areal et al. (2013) and
Nofsinger and Varma (2014), but conditioned to
the state of stock market inst ead of
macroeconomic states.
Question
Findings and
contribution
Managerial
implications
Are there any differences in the performance of
SR funds over the business cycle?
Are there an y market timing abilities in SR
funds over the business cycle?
Are there any differences betweenperformance
of SR and conventional mutual funds?
We propose grouping SR funds according to
three different criteria :
-Morningstar ethical strategy focus
-Morningstar scores in SR attributes
-Morningstar category.
Unlike Mallet and Michelson (2010), but in line
with Muñoz et al. (2014),differences in
performance of SR fund groups are found.
Environmental outperform (und erperform)
other SR funds before (after) financial crisis.
SR funds wi th lower SR scores outperfo rm SR
funds with high scores.
Negative aggregated performa nce of SR funds
is driven by the more specialised fu nds not
included in the 3x3 Morningstar equity style
matrix(3 1.19% of the SR funds).
Jiang et al. (2007) and Ferson and Mo (2016),
among others:
For conventional mutual funds the most
common evidence found is of negative or non-
significant timing.
Ferruz et al.(2010) and Leite and Cortez
(2014b):
For SR funds similar evidence is found.
Apart from the studies by Muño z et al. (2014)
and Nofsinger and Va rma (2014), only studies
on green funds vs other SR funds:
Mallett and Michelson (2010),Climent and
Soriano (2011):find no significant differences
between gr een and other SR funds, butresults
may depend on the sample period.
Chang et al. (2012):green mutual funds
underperform, but this is not uniform.
Muñoz et al. (2014): gr een fund managers
achieve better results during crisis.
In line with Chang et al. (2012) Stewart (2014):
the results of SR fundsare linked to specialised
sectors.
SR investments ar e subj ect to idiosyncratic
risks that are relevant in order to discrimi nate
between performa nce obtained from managers’
ability and performancefrom underlying assets.
Differences found in performance over su b-
periods may not have been driven by
differences in ma nagement skill, but by the
time-varying beha viour of the asset classes in
which the funds invested.
When specific SR benc hmarks are considered,
the evidence of market timing improves, and
this effect is particularly remarkable in the last
two sub-periods after the financial crisis event.
Part of the previous evidence of negative
market timing ability found in the literature
would be driven by the behaviour of the class of
stocks linked to the specific SR benchmarks
considered rather than by the SR fund
managers.
In line with Pástor and Stambaugh (2002) and
Matallín-Sáez (2006), the omission of
appropriate benchmarks –in this case linked to
SR features –is relevant for performance
assessment.
The comparison between SR and conventional
mutual funds should be established with
appropriate peers, mu tual fund style being one
of the most significant factors.
Results depend on the two types of models
proposed.
For basic fi nancial models omitting relevant
specific SR asset classes:
In line with theliterature, in aggregate, market
timing is partially significant, taking positive
(negative) values for sub-periods before (after)
financial crisis.
SR funds with high SR scores do not time as
well as funds with low scores.
For extended models considering idiosyncratic
SR features:
In aggregate, mark et timing takes positive
values, being partially si gnificant for sub-
periods before financial crisis.
Kosowski (2011), De Souza and Lynch (2012)
and Kacperczyk et al. (2014):
Conventional mutual fund performance varies
over the business cycle.
Moskowitz (2000):
The abnormal performance of conventional
mutual funds increase s during crisis times.
Areal et al. (2013):
SR fund performance improves during
recessions.
Results depend on the two types of models
proposed.
For basic fi nancial models omitting relevant
specific SR asset classes:
In aggregateSR funds perform better during
expansions than in recessions, but the
differences are not significant.
For extended models considering idiosyncratic
SR features:
In aggregate and in line with Areal et al. (2013)
performance in recession sub-periods improves
markedly, and so SR funds perform better
during recessions than in expansions, although
thedifferences remain non-significant.
Improvement in recession periods is especially
relevant for Environmentalfu nds, High SR
attribute scores funds andmore specialised
funds that, pr obably because of their higher
socially responsible profile, are the fu rthest
away from the standard Morningstar 3x3 style
box, i.e. the Othergroup.
Evidence
from the
previous
literature
To continue reading
Request your trial