-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathPerf.html
906 lines (878 loc) · 223 KB
/
Perf.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
<!DOCTYPE html>
<html lang="en">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
<title>Chapter 4 ESG investing and financial performance | Perspectives in sustainable equity investing</title>
<meta name="author" content="Guillaume Coqueret">
<meta name="description" content=".container-fluid main { max-width: 55rem; font-size: 1rem; } One of the central questions in ESG investing pertains to its opportunity cost: is sustainable investing penalizing from a pure...">
<meta name="generator" content="bookdown 0.24 with bs4_book()">
<meta property="og:title" content="Chapter 4 ESG investing and financial performance | Perspectives in sustainable equity investing">
<meta property="og:type" content="book">
<meta property="og:description" content=".container-fluid main { max-width: 55rem; font-size: 1rem; } One of the central questions in ESG investing pertains to its opportunity cost: is sustainable investing penalizing from a pure...">
<meta name="twitter:card" content="summary">
<meta name="twitter:title" content="Chapter 4 ESG investing and financial performance | Perspectives in sustainable equity investing">
<meta name="twitter:description" content=".container-fluid main { max-width: 55rem; font-size: 1rem; } One of the central questions in ESG investing pertains to its opportunity cost: is sustainable investing penalizing from a pure...">
<!-- JS --><script src="https://cdnjs.cloudflare.com/ajax/libs/clipboard.js/2.0.6/clipboard.min.js" integrity="sha256-inc5kl9MA1hkeYUt+EC3BhlIgyp/2jDIyBLS6k3UxPI=" crossorigin="anonymous"></script><script src="https://cdnjs.cloudflare.com/ajax/libs/fuse.js/6.4.6/fuse.js" integrity="sha512-zv6Ywkjyktsohkbp9bb45V6tEMoWhzFzXis+LrMehmJZZSys19Yxf1dopHx7WzIKxr5tK2dVcYmaCk2uqdjF4A==" crossorigin="anonymous"></script><script src="https://kit.fontawesome.com/6ecbd6c532.js" crossorigin="anonymous"></script><script src="libs/header-attrs-2.11/header-attrs.js"></script><script src="libs/jquery-3.6.0/jquery-3.6.0.min.js"></script><meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
<link href="libs/bootstrap-4.6.0/bootstrap.min.css" rel="stylesheet">
<script src="libs/bootstrap-4.6.0/bootstrap.bundle.min.js"></script><script src="libs/bs3compat-0.3.1/transition.js"></script><script src="libs/bs3compat-0.3.1/tabs.js"></script><script src="libs/bs3compat-0.3.1/bs3compat.js"></script><link href="libs/bs4_book-1.0.0/bs4_book.css" rel="stylesheet">
<script src="libs/bs4_book-1.0.0/bs4_book.js"></script><script src="libs/htmlwidgets-1.5.4/htmlwidgets.js"></script><link href="libs/leaflet-1.3.1/leaflet.css" rel="stylesheet">
<script src="libs/leaflet-1.3.1/leaflet.js"></script><link href="libs/leafletfix-1.0.0/leafletfix.css" rel="stylesheet">
<script src="libs/proj4-2.6.2/proj4.min.js"></script><script src="libs/Proj4Leaflet-1.0.1/proj4leaflet.js"></script><link href="libs/rstudio_leaflet-1.3.1/rstudio_leaflet.css" rel="stylesheet">
<script src="libs/leaflet-binding-2.0.4.1/leaflet.js"></script><script src="https://cdnjs.cloudflare.com/ajax/libs/autocomplete.js/0.38.0/autocomplete.jquery.min.js" integrity="sha512-GU9ayf+66Xx2TmpxqJpliWbT5PiGYxpaG8rfnBEk1LL8l1KGkRShhngwdXK1UgqhAzWpZHSiYPc09/NwDQIGyg==" crossorigin="anonymous"></script><script src="https://cdnjs.cloudflare.com/ajax/libs/mark.js/8.11.1/mark.min.js" integrity="sha512-5CYOlHXGh6QpOFA/TeTylKLWfB3ftPsde7AnmhuitiTX4K5SqCLBeKro6sPS8ilsz1Q4NRx3v8Ko2IBiszzdww==" crossorigin="anonymous"></script><!-- CSS -->
</head>
<body data-spy="scroll" data-target="#toc">
<div class="container-fluid">
<div class="row">
<header class="col-sm-12 col-lg-3 sidebar sidebar-book"><a class="sr-only sr-only-focusable" href="#content">Skip to main content</a>
<div class="d-flex align-items-start justify-content-between">
<h1>
<a href="index.html" title="">Perspectives in sustainable equity investing</a>
</h1>
<button class="btn btn-outline-primary d-lg-none ml-2 mt-1" type="button" data-toggle="collapse" data-target="#main-nav" aria-expanded="true" aria-controls="main-nav"><i class="fas fa-bars"></i><span class="sr-only">Show table of contents</span></button>
</div>
<div id="main-nav" class="collapse-lg">
<form role="search">
<input id="search" class="form-control" type="search" placeholder="Search" aria-label="Search">
</form>
<nav aria-label="Table of contents"><h2>Table of contents</h2>
<ul class="book-toc list-unstyled">
<li><a class="" href="index.html"><span class="header-section-number">1</span> Introduction</a></li>
<li><a class="" href="ESGdata.html"><span class="header-section-number">2</span> ESG Data</a></li>
<li><a class="" href="Investors.html"><span class="header-section-number">3</span> Investors and SRI</a></li>
<li><a class="active" href="Perf.html"><span class="header-section-number">4</span> ESG investing and financial performance</a></li>
<li><a class="" href="Quant.html"><span class="header-section-number">5</span> Quantitative portfolio construction with ESG data and criteria</a></li>
<li><a class="" href="Climate.html"><span class="header-section-number">6</span> Climate change risk</a></li>
<li><a class="" href="Equilibrium.html"><span class="header-section-number">7</span> SRI in economic equilibria</a></li>
<li><a class="" href="Conc.html"><span class="header-section-number">8</span> Conclusion</a></li>
<li><a class="" href="Bib.html"><span class="header-section-number">9</span> Bibliography</a></li>
</ul>
<div class="book-extra">
</div>
</nav>
</div>
</header><main class="col-sm-12 col-md-9 col-lg-7" id="content"><div id="Perf" class="section level1" number="4">
<h1>
<span class="header-section-number">4</span> ESG investing and financial performance<a class="anchor" aria-label="anchor" href="#Perf"><i class="fas fa-link"></i></a>
</h1>
<style>
.container-fluid main {
max-width: 55rem;
font-size: 1rem;
}
</style>
<p>One of the central questions in ESG investing pertains to its opportunity cost: is sustainable investing penalizing from a pure pecuniary standpoint? This chapter addresses this complex and highly debated question. We start by a simple model that illustrates some typical mechanisms that can enlighten the situations when green firms can outperform brown ones. Then, we split the contributions of the impact of ethical screens on financial performance into four categories: the ones that document a positive relationship, the ones that find a negative relationship, the ones that report no particular relationship, and finally the ones that conclude that the relationship is contingent on some external factors. We mostly refer to financial performance in the sense of returns, though sometimes it means valuation. Nevertheless, SRI impacts other aspects of corporations’ finances, with risk being a major concern. We list some of these facets in the last two subsections below. </p>
<p>While the purpose of this chapter is to study the link between sustainability choices and financial performance, we do it in one direction (from ESG to finance). <span class="citation"><a href="Bib.html#ref-dasgupta2021financial" role="doc-biblioref">DasGupta</a> (<a href="Bib.html#ref-dasgupta2021financial" role="doc-biblioref">2021</a>)</span> has shown that mediocre financial performance can be an incentive for firms to improve their ESG practices.</p>
<div id="toy-model" class="section level2" number="4.1">
<h2>
<span class="header-section-number">4.1</span> Toy model<a class="anchor" aria-label="anchor" href="#toy-model"><i class="fas fa-link"></i></a>
</h2>
<p>The mathematics-averse reader is advised to skip this subsection.</p>
<div id="theory-assets-agents-equilibrium" class="section level3" number="4.1.1">
<h3>
<span class="header-section-number">4.1.1</span> Theory: assets, agents, equilibrium<a class="anchor" aria-label="anchor" href="#theory-assets-agents-equilibrium"><i class="fas fa-link"></i></a>
</h3>
<p>A full section of the survey is dedicated to theoretical models (Chapter <a href="Equilibrium.html#Equilibrium">7</a>), so the framework we present here is very stylized.<a class="footnote-ref" tabindex="0" data-toggle="popover" data-content='<p>It is inspired from the model of <span class="citation"><a href="Bib.html#ref-kacperczyk2019investor" role="doc-biblioref">Kacperczyk, Nosal, and Stevens</a> (<a href="Bib.html#ref-kacperczyk2019investor" role="doc-biblioref">2019</a>)</span>, though the utility function and the market clearing mechanism are different.</p>'><sup>24</sup></a> We consider two agents and two risky assets (for simplicity). The two assets are indexed by <span class="math inline">\(g\)</span> for <em>green</em> and <span class="math inline">\(b\)</span> for <em>brown</em>, while the two agents are indexed with uppercase letters <span class="math inline">\(G\)</span> (sustainability sensitive agent) and <span class="math inline">\(B\)</span> (less ESG-driven, or even ESG insensitive agent). We will often use <span class="math inline">\(X\)</span> to denote one of the agents (<span class="math inline">\(G\)</span> or <span class="math inline">\(B\)</span>) and <span class="math inline">\(y\)</span> one of the assets (<span class="math inline">\(g\)</span> or <span class="math inline">\(b\)</span>).</p>
<p><strong>Assets</strong>. Prices are denoted with <span class="math inline">\(p_g\)</span> and <span class="math inline">\(p_b\)</span>. Both assets are expected to yield a cash-flow (or payoff): <span class="math inline">\(z_g\)</span> and <span class="math inline">\(z_b\)</span>, respectively, and we assume that they have a correlation equal to <span class="math inline">\(\rho\)</span>. Agents agree on the variances of these payoffs, <span class="math inline">\(\sigma_g^2\)</span> and <span class="math inline">\(\sigma_b^2\)</span>. However, they are allowed to disagree on their means: agent <span class="math inline">\(G\)</span> estimates they are <span class="math inline">\(m_{Gg}\)</span> and <span class="math inline">\(m_{Gb}\)</span> while agent B believes they are <span class="math inline">\(m_{Bg}\)</span> and <span class="math inline">\(m_{Bb}\)</span>. The assets are also characterized by a non-random (observable<a class="footnote-ref" tabindex="0" data-toggle="popover" data-content='<p>Again, this is a strong assumption because in practice, ratings disagree, see Section <a href="ESGdata.html#disagreement">2.2</a>. The article <span class="citation"><a href="Bib.html#ref-avramov2020investment" role="doc-biblioref">Avramov, Cheng, et al.</a> (<a href="Bib.html#ref-avramov2020investment" role="doc-biblioref">2021</a>)</span> proposes a theoretical model with disagreement.</p>'><sup>25</sup></a>) ESG score: <span class="math inline">\(e_g\)</span> and <span class="math inline">\(e_b\)</span>, such that, naturally, <span class="math inline">\(e_g>e_b\)</span>. Both assets have fixed supplies <span class="math inline">\(s_g\)</span> and <span class="math inline">\(s_b\)</span>, which are supplied by so-called <em>noise traders</em> who trade for reasons other than payoffs. One risk-free asset is also available in unlimited supply, with unit price and certain payoff equal to <span class="math inline">\(r>0\)</span>.</p>
<p><strong>Agents</strong>. Agents have initial wealth <span class="math inline">\(W_G\)</span> and <span class="math inline">\(W_B\)</span>, for the <em>Green</em> (sustainability-driven) and <em>Brown</em> agent, respectively. Each agent <span class="math inline">\(X\in \{G,B \}\)</span> buys a quantity <span class="math inline">\(q_{Xy}\)</span> of shares of asset <span class="math inline">\(y\in \{g,b \}\)</span>.
The terminal wealth of agents satisfies
<span class="math display">\[W_X^*=\underbrace{r(W_X-q_{Xg}p_g-q_{Xb}p_b)}_{\text{payoff from riskless asset}}+\underbrace{q_{Xg}z_g+q_{Xb}z_b}_{\text{payoff from risky assets}}, \quad X\in \{G,B \} .\]</span>
We re-write this expression in the form of the gross return:
<span class="math display">\[\frac{W_X^*}{W_X}=r(1-w_{Xg}p_g-w_{Xb}p_b)+w_{Xg}z_g+w_{Xb}z_b, \quad X\in \{G,B \} ,\]</span>
where <span class="math inline">\(w_{Xy}=q_{Xy}/W_X\)</span> is the number of shares of agent <span class="math inline">\(X\)</span> in asset <span class="math inline">\(y\)</span>, divided by the total initial wealth of the agent.<a class="footnote-ref" tabindex="0" data-toggle="popover" data-content='<p>Strictly speaking, the <span class="math inline">\(w_X\)</span> are <em>not</em> portfolio weights.</p>'><sup>26</sup></a>
Also, the ESG score of agent <span class="math inline">\(X\)</span>’s holdings is evaluated as
<span class="math display">\[E_X=w_{Xg}e_g+w_{Xb}e_b.\]</span>
The agents have a quadratic utility on gross return on wealth, plus an ESG component:
<span class="math display">\[U_X=\mathbb{E}\left[\frac{W_X^*}{W_X}\right]-\frac{\gamma}{2}\mathbb{V}\left[\frac{W_X^*}{W_X}\right]+\delta_X E_X,\]</span>
where <span class="math inline">\(\delta_G>\delta_B\)</span> means that agent <span class="math inline">\(G\)</span> cares more about ESG score than agent <span class="math inline">\(B\)</span>. The risk aversion coefficient <span class="math inline">\(\gamma>0\)</span> is common to both agents. If we use bold vector notations for prices <span class="math inline">\(\textbf{p}=[p_g \ p_b]\)</span> and relative portfolio holdings <span class="math inline">\(\textbf{w}_X=[q_{Xg} \ q_{Xb}]\)</span>, then the utility function reads
<span class="math display">\[U_X=r-r\textbf{w}_X'\textbf{p} +\textbf{w}_X'\textbf{m}_X-\frac{\gamma}{2}\textbf{w}_X'\boldsymbol{\Sigma}\textbf{w}_X+\delta_X\textbf{w}_X'\textbf{e},\]</span>
where <span class="math inline">\(\textbf{m}_X=[m_{Xg} \ m_{Xb}]'\)</span>, <span class="math inline">\(\textbf{e}=[e_g \ e_b]'\)</span> and
<span class="math inline">\(\boldsymbol{\Sigma}=\begin{bmatrix} \sigma_g^2 & \rho \sigma_g\sigma_b \\ \rho \sigma_g\sigma_b & \sigma_b^2 \end{bmatrix}.\)</span> The first-order condition (the gradient of <span class="math inline">\(U_X\)</span> must be equal to zero) implies
<span class="math display">\[-r \textbf{p}+ \textbf{m}_X-\gamma \boldsymbol{\Sigma} \textbf{w}_X+ \delta_X \textbf{e}= \textbf{0},\]</span>
so that agents’ relative demands satisfy
<span class="math display" id="eq:demands">\[\begin{equation}
\textbf{w}_X=\gamma^{-1}\boldsymbol{\Sigma}^{-1}(\textbf{m}_X+\delta_X\textbf{e}-r\textbf{p}) .
\tag{4.1}
\end{equation}\]</span>
As <span class="math inline">\(\delta_X\)</span> increases in magnitude, the agent will progressively grant more importance to the ESG score <span class="math inline">\(\textbf{e}\)</span> than to the expected payoff <span class="math inline">\(\textbf{m}_X\)</span>. Also, all other things equal, the magnitude of demand decreases with risk aversion and payoff volatility. Note that the demand can very well be negative. Also, if the correlation between payoffs (<span class="math inline">\(\rho\)</span>) is zero, the expression simplifies to
<span class="math display" id="eq:demands2">\[\begin{equation}
\hspace{18mm}w_{Xy}=\frac{m_{Xy}+\delta_Xe_y-rp_y}{\gamma \sigma_y^2}, \quad X\in\{G,B\}, \quad y \in\{g,b \}.
\tag{4.2}
\end{equation}\]</span>
There are three parts in the above formula. The first can be interpreted as the agent-specific attractiveness of the asset <span class="math inline">\(m_{Xy}+\delta_Xe_y\)</span>, which has two components: payoff and sustainability. The second is the negative impact of the price (demand decreases with price). The third (denominator) is risk. The assumption that <span class="math inline">\(\rho=0\)</span> is quite strong, as it implies that both assets are priced independently.</p>
<p><strong>Equilibrium</strong>. The price-weighted demands must satisfy the market clearing condition (demand equals supply). The aggregate demand for one asset <span class="math inline">\(y\)</span> is simply <span class="math inline">\((q_{Gy}+q_{By})p_y\)</span>, i.e., the price times the total holdings (in shares). In our simple setting, we assume a vector <span class="math inline">\(\textbf{s}=[s_g \ s_b]'\)</span> of supply for assets, thus
<span class="math display" id="eq:clearing">\[\begin{equation}
\underbrace{\text{diag}(W_G\textbf{w}_{G} +W_B \textbf{w}_{B})\textbf{p}}_{\text{total demand for asset }}=\textbf{s},
\tag{4.3}
\end{equation}\]</span>
where diag<span class="math inline">\((\textbf{v})\)</span> takes a vector <span class="math inline">\(\textbf{v}\)</span> as argument and yields a diagonal matrix with <span class="math inline">\(\textbf{v}\)</span> as diagonal values. The above equation highlights the importance of each agent’s relative weight on the market, which is given by the ratio of their wealth to total wealth <span class="math inline">\(W=W_G+W_B\)</span>. To further ease the analysis, we posit that the correlation between payoffs is zero. Consequently, plugging the demands in Equation <a href="Perf.html#eq:demands2">(4.2)</a>, the equation for asset <span class="math inline">\(y\)</span> translates to
<span class="math display">\[-\overbrace{[r\left(W_G+W_B\right)]}^{\text{riskless alternative}}p_y^2 +\overbrace{ \left[W_G(m_{Gy}+\delta_Ge_y)+W_B(m_{By}+\delta_Be_y)\right]}^{A_y=\text{total weighted attractiveness}}p_y-\overbrace{\gamma \sigma_y^2s_y}^{\text{risk/supply}}=0,\]</span>
where we have singled out the total attractiveness of the asset, which we write <span class="math inline">\(A_y\)</span>. The equation is quadratic in the price <span class="math inline">\(p_y\)</span> because of the way the <span class="math inline">\(w_Y\)</span> are defined. In many papers, the market clearing Equation <a href="Perf.html#eq:clearing">(4.3)</a> leads to linear forms.
Under the parametric condition
<span class="math display">\[\begin{equation}
(\textbf{C}) \hspace{6mm} A_y^2-4\gamma \sigma_y^2s_yrW\ge0,
\label{eq:assumpA}
\end{equation}\]</span>
the positive price of asset <span class="math inline">\(y\)</span> is
<span class="math display" id="eq:price">\[\begin{equation}
p_y = \frac{\sqrt{A_y^2-4\gamma \sigma_y^2s_yrW}+A_y}{2rW}.
\tag{4.4}
\end{equation}\]</span>
Intuitively, the price of an asset is increasing in attractiveness, and decreasing with risk and supply.</p>
</div>
<div id="numerical-example" class="section level3" number="4.1.2">
<h3>
<span class="header-section-number">4.1.2</span> Numerical example<a class="anchor" aria-label="anchor" href="#numerical-example"><i class="fas fa-link"></i></a>
</h3>
<p>Now, let us test some parametric configurations of the model so as to reveal several key theoretical predictions. We make the following simplifications:</p>
<ul>
<li>We normalize wealths. <span class="math inline">\(W_B=1\)</span>, so that the Brown investor is the benchmark. We can take several values for <span class="math inline">\(W_G\)</span>. For <span class="math inline">\(W_G=0.5\)</span>, the Green investor represents one third of the market and for <span class="math inline">\(W_G=2\)</span>, two-thirds of the market.</li>
<li>For ease of interpretation, we fix the ESG scores to <span class="math inline">\(e_g=1\)</span> (green) and <span class="math inline">\(e_b=0\)</span> (brown). Also, for both assets, <span class="math inline">\(\sigma_y=0.2\)</span> (<span class="math inline">\(\sigma_y^2=0.04\)</span>) and <span class="math inline">\(s_y=1\)</span> (unit supply).</li>
<li>The taste for sustainability of the Green investor is set to <span class="math inline">\(\delta_G=0.15\)</span>. On the other hand, the Brown investor does not care about ESG and <span class="math inline">\(\delta_B=0\)</span>.</li>
<li>Finally, we assume that <span class="math inline">\(r=0.02\)</span>. <span class="math inline">\(r\)</span> has two important effects on prices. First, it plays a scaling role in condition (<span class="math inline">\(\textbf{C}\)</span>): large values of <span class="math inline">\(r\)</span> may lead to a violation of the condition. Second, it normalizes price values in Equation , so that, at a first-order approximation, prices are inversely proportional to <span class="math inline">\(r\)</span>.</li>
</ul>
<p>We consider two alternative versions of agent beliefs:</p>
<ol style="list-style-type: decimal">
<li>
<strong>Extreme polarization</strong>. The Green investor is purely driven by ESG concerns and is agnostic with respect to returns, so that <span class="math inline">\(\textbf{m}_G=\textbf{0}\)</span>. The Brown investor, in contrast, and as is commonly accepted (see <span class="citation"><a href="Bib.html#ref-bolton2020investors" role="doc-biblioref">Bolton and Kacperczyk</a> (<a href="Bib.html#ref-bolton2020investors" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-bolton2020carbon" role="doc-biblioref">Bolton and Kacperczyk</a> (<a href="Bib.html#ref-bolton2020carbon" role="doc-biblioref">2020</a>)</span>), expects incremental returns for the brown asset (akin to a <em>carbon premium</em>). We model that by assuming that <span class="math inline">\(\textbf{m}_B\)</span> is negatively linked to <span class="math inline">\(\textbf{e}\)</span>: <span class="math inline">\(m_{Bg}=0\)</span> (zero expected payoff from the green asset), <span class="math inline">\(m_{Bb}=0.1\)</span> (i.e., a 10% expected payoff from the brown asset).</li>
<li>
<strong>Moderate diversity in tastes</strong>. In this case, agents have less marked preferences. The Green investor is now interested in payoffs (in addition to ESG). Thus, <span class="math inline">\(m_{Gg}=m_{Gb}=0.1\)</span>, which means that both assets are expected to have the same average payoff. The Brown agent agrees with the Green agent on these parameters, so that <span class="math inline">\(m_{Bg}=m_{Bb}=0.1\)</span>. The only difference between the two agents is therefore in the ESG preferences, which remain the same as under extreme polarization.</li>
</ol>
<p>In Figure <a href="Perf.html#fig:thmod1">4.1</a>, we plot the corresponding prices as functions of the wealth of the Green agent. Naturally, because the Green agent favors the ESG criterion, the price of the green asset increases when sustainable demand increases. For the brown asset, however, it is the opposite in the left panel, because the raw demand of the Green agent is negative (see Equation <a href="Perf.html#eq:demands2">(4.2)</a>. In the right panel, the Green agent has a mildly positive demand and the price of the brown asset converges (as it should) to <span class="math inline">\(m_{Gb}/r=5\)</span>. The limiting value for the green asset is <span class="math inline">\((m_{Gg}+\delta_Ge_g)/r=12.5\)</span>.</p>
<p>Below, we code the theoretical price function and the plot.</p>
<div class="sourceCode" id="cb8"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span class="kw"><a href="https://rdrr.io/r/base/library.html">library</a></span><span class="op">(</span><span class="va"><a href="https://github.com/stefano-meschiari/latex2exp">latex2exp</a></span><span class="op">)</span> <span class="co"># Package for LaTex inclusion in plots</span>
<span class="kw"><a href="https://rdrr.io/r/base/library.html">library</a></span><span class="op">(</span><span class="va"><a href="https://patchwork.data-imaginist.com">patchwork</a></span><span class="op">)</span> <span class="co"># Package for graph combinations/layout</span>
<span class="va">prc</span> <span class="op"><-</span> <span class="kw">function</span><span class="op">(</span><span class="va">W_G</span>, <span class="va">W_B</span>, <span class="va">e</span>, <span class="va">m_G</span>, <span class="va">m_B</span>, <span class="va">d_G</span>, <span class="va">d_B</span>, <span class="va">sigma</span>, <span class="va">s</span>, <span class="va">gamma</span>, <span class="va">r</span><span class="op">)</span><span class="op">{</span>
<span class="va">A</span> <span class="op"><-</span> <span class="va">W_G</span><span class="op">*</span><span class="op">(</span><span class="va">m_G</span><span class="op">+</span><span class="va">d_G</span><span class="op">*</span><span class="va">e</span><span class="op">)</span> <span class="op">+</span> <span class="va">W_B</span><span class="op">*</span><span class="op">(</span><span class="va">m_B</span><span class="op">+</span><span class="va">d_B</span><span class="op">*</span><span class="va">e</span><span class="op">)</span>
<span class="va">W</span> <span class="op"><-</span> <span class="va">W_G</span> <span class="op">+</span> <span class="va">W_B</span>
<span class="kw"><a href="https://rdrr.io/r/base/function.html">return</a></span><span class="op">(</span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/r/base/MathFun.html">sqrt</a></span><span class="op">(</span><span class="va">A</span><span class="op">^</span><span class="fl">2</span><span class="op">-</span><span class="fl">4</span><span class="op">*</span><span class="va">gamma</span><span class="op">*</span><span class="va">sigma</span><span class="op">^</span><span class="fl">2</span><span class="op">*</span><span class="va">s</span><span class="op">*</span><span class="va">r</span><span class="op">*</span><span class="va">W</span><span class="op">)</span><span class="op">+</span><span class="va">A</span><span class="op">)</span><span class="op">/</span><span class="fl">2</span><span class="op">/</span><span class="va">r</span><span class="op">/</span><span class="va">W</span><span class="op">)</span>
<span class="op">}</span>
<span class="va">sigma</span> <span class="op"><-</span> <span class="fl">0.2</span>
<span class="va">gamma</span> <span class="op"><-</span> <span class="fl">1</span> <span class="co"># risk aversion</span>
<span class="va">s</span> <span class="op"><-</span> <span class="fl">1</span>
<span class="va">r</span> <span class="op"><-</span> <span class="fl">0.02</span> <span class="co"># risk-free rate</span>
<span class="va">W_B</span> <span class="op"><-</span> <span class="fl">1</span> <span class="co"># wealth of brown agent (fixed)</span>
<span class="va">W_G</span> <span class="op"><-</span> <span class="fu"><a href="https://rdrr.io/r/base/seq.html">seq</a></span><span class="op">(</span><span class="fl">0.1</span>, <span class="fl">2.7</span>, <span class="fl">0.1</span><span class="op">)</span> <span class="co"># vector of wealths for green agent</span>
<span class="va">d_B</span> <span class="op"><-</span> <span class="fl">0</span>
<span class="va">d_G</span> <span class="op"><-</span> <span class="fl">0.15</span> <span class="co"># ESG Importance</span>
<span class="va">m_G</span> <span class="op"><-</span> <span class="fl">0</span>
<span class="va">e</span> <span class="op"><-</span> <span class="fl">0</span>
<span class="va">m_B</span> <span class="op"><-</span> <span class="fl">0.1</span><span class="op">*</span> <span class="op">(</span><span class="fl">1</span><span class="op">-</span><span class="va">e</span><span class="op">)</span>
<span class="va">p_e_0</span> <span class="op"><-</span> <span class="fu">prc</span><span class="op">(</span><span class="va">W_G</span>, <span class="va">W_B</span>, <span class="va">e</span>, <span class="va">m_G</span>, <span class="va">m_B</span>, <span class="va">d_G</span>, <span class="va">d_B</span>, <span class="va">sigma</span>, <span class="va">s</span>, <span class="va">gamma</span>, <span class="va">r</span><span class="op">)</span>
<span class="va">e</span> <span class="op"><-</span> <span class="fl">1</span>
<span class="va">m_B</span> <span class="op"><-</span> <span class="fl">0.1</span><span class="op">*</span> <span class="op">(</span><span class="fl">1</span><span class="op">-</span><span class="va">e</span><span class="op">)</span>
<span class="va">p_e_1</span> <span class="op"><-</span> <span class="fu">prc</span><span class="op">(</span><span class="va">W_G</span>, <span class="va">W_B</span>, <span class="va">e</span>, <span class="va">m_G</span>, <span class="va">m_B</span>, <span class="va">d_G</span>, <span class="va">d_B</span>, <span class="va">sigma</span>, <span class="va">s</span>, <span class="va">gamma</span>, <span class="va">r</span><span class="op">)</span>
<span class="va">g1</span> <span class="op"><-</span> <span class="fu"><a href="https://r-spatial.github.io/sf/reference/tibble.html">tibble</a></span><span class="op">(</span>W_G <span class="op">=</span> <span class="va">W_G</span>, p_e_0 <span class="op">=</span> <span class="va">p_e_0</span>, p_e_1 <span class="op">=</span> <span class="va">p_e_1</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/leaflet/man/leaflet-imports.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/tidyr/man/pivot_longer.html">pivot_longer</a></span><span class="op">(</span><span class="op">-</span><span class="va">W_G</span>, names_to <span class="op">=</span> <span class="st">"Type"</span>, values_to <span class="op">=</span> <span class="st">"Price"</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/leaflet/man/leaflet-imports.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggplot.html">ggplot</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/aes.html">aes</a></span><span class="op">(</span>x <span class="op">=</span> <span class="va">W_G</span>, y <span class="op">=</span> <span class="va">Price</span>, color <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/factor.html">as.factor</a></span><span class="op">(</span><span class="va">Type</span><span class="op">)</span>, shape <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/factor.html">as.factor</a></span><span class="op">(</span><span class="va">Type</span><span class="op">)</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">ggtitle</a></span><span class="op">(</span><span class="st">"Extreme polarization of tastes"</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggtheme.html">theme_light</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/geom_segment.html">geom_segment</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/aes.html">aes</a></span><span class="op">(</span>x <span class="op">=</span> <span class="fl">1</span>, y <span class="op">=</span> <span class="fl">0</span>, xend <span class="op">=</span> <span class="fl">1</span>, yend <span class="op">=</span> <span class="fl">4.3</span><span class="op">)</span>, linetype <span class="op">=</span> <span class="fl">2</span>, color <span class="op">=</span> <span class="st">"black"</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/annotate.html">annotate</a></span><span class="op">(</span>geom <span class="op">=</span> <span class="st">"text"</span>, x <span class="op">=</span> <span class="fl">1</span>, y <span class="op">=</span> <span class="fl">4.9</span>, label<span class="op">=</span> <span class="st">"Weath of \n Brown \n agent"</span>, color<span class="op">=</span><span class="st">"black"</span>, size <span class="op">=</span> <span class="fl">4</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/annotate.html">annotate</a></span><span class="op">(</span>geom <span class="op">=</span> <span class="st">"text"</span>, x <span class="op">=</span> <span class="fl">0.3</span>, y <span class="op">=</span> <span class="fl">0.4</span>, label<span class="op">=</span><span class="fu"><a href="https://rdrr.io/pkg/latex2exp/man/TeX.html">TeX</a></span><span class="op">(</span><span class="st">" No \n green \n price"</span><span class="op">)</span>, color<span class="op">=</span><span class="st">"#0DCD64"</span>, size <span class="op">=</span> <span class="fl">5</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/annotate.html">annotate</a></span><span class="op">(</span>geom <span class="op">=</span> <span class="st">"text"</span>, x <span class="op">=</span> <span class="fl">2.5</span>, y <span class="op">=</span> <span class="fl">0.4</span>, label<span class="op">=</span><span class="fu"><a href="https://rdrr.io/pkg/latex2exp/man/TeX.html">TeX</a></span><span class="op">(</span><span class="st">" No \n brown \n price"</span><span class="op">)</span>, color<span class="op">=</span><span class="st">"#735E50"</span>, size <span class="op">=</span> <span class="fl">5</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/geom_point.html">geom_point</a></span><span class="op">(</span>size <span class="op">=</span> <span class="fl">2.5</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">xlab</a></span><span class="op">(</span><span class="st">"Wealth of Green agent"</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/theme.html">theme</a></span><span class="op">(</span>text <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_text</a></span><span class="op">(</span>size <span class="op">=</span> <span class="fl">14</span><span class="op">)</span>,
legend.position <span class="op">=</span> <span class="st">"None"</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/scale_manual.html">scale_color_manual</a></span><span class="op">(</span>values <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"#735E50"</span>, <span class="st">"#0DCD64"</span><span class="op">)</span>, labels <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"Brown, e = 0"</span>, <span class="st">"Green, e = 1"</span><span class="op">)</span>, name <span class="op">=</span> <span class="st">"Asset"</span><span class="op">)</span>
<span class="va">e</span> <span class="op"><-</span> <span class="fl">0</span>
<span class="va">m_B</span> <span class="op"><-</span> <span class="fl">0.1</span>
<span class="va">m_G</span> <span class="op"><-</span> <span class="fl">0.1</span>
<span class="va">p_e_0</span> <span class="op"><-</span> <span class="fu">prc</span><span class="op">(</span><span class="va">W_G</span>, <span class="va">W_B</span>, <span class="va">e</span>, <span class="va">m_G</span>, <span class="va">m_B</span>, <span class="va">d_G</span>, <span class="va">d_B</span>, <span class="va">sigma</span>, <span class="va">s</span>, <span class="va">gamma</span>, <span class="va">r</span><span class="op">)</span>
<span class="va">e</span> <span class="op"><-</span> <span class="fl">1</span>
<span class="va">m_B</span> <span class="op"><-</span> <span class="fl">0.1</span>
<span class="va">p_e_1</span> <span class="op"><-</span> <span class="fu">prc</span><span class="op">(</span><span class="va">W_G</span>, <span class="va">W_B</span>, <span class="va">e</span>, <span class="va">m_G</span>, <span class="va">m_B</span>, <span class="va">d_G</span>, <span class="va">d_B</span>, <span class="va">sigma</span>, <span class="va">s</span>, <span class="va">gamma</span>, <span class="va">r</span><span class="op">)</span>
<span class="va">g2</span> <span class="op"><-</span> <span class="fu"><a href="https://r-spatial.github.io/sf/reference/tibble.html">tibble</a></span><span class="op">(</span>W_G <span class="op">=</span> <span class="va">W_G</span>, p_e_0 <span class="op">=</span> <span class="va">p_e_0</span>, p_e_1 <span class="op">=</span> <span class="va">p_e_1</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/leaflet/man/leaflet-imports.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/tidyr/man/pivot_longer.html">pivot_longer</a></span><span class="op">(</span><span class="op">-</span><span class="va">W_G</span>, names_to <span class="op">=</span> <span class="st">"Type"</span>, values_to <span class="op">=</span> <span class="st">"Price"</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/leaflet/man/leaflet-imports.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggplot.html">ggplot</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/aes.html">aes</a></span><span class="op">(</span>x <span class="op">=</span> <span class="va">W_G</span>, y <span class="op">=</span> <span class="va">Price</span>, color <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/factor.html">as.factor</a></span><span class="op">(</span><span class="va">Type</span><span class="op">)</span>, shape <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/factor.html">as.factor</a></span><span class="op">(</span><span class="va">Type</span><span class="op">)</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">ggtitle</a></span><span class="op">(</span><span class="st">"Moderate diversity in tastes"</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/geom_segment.html">geom_segment</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/aes.html">aes</a></span><span class="op">(</span>x <span class="op">=</span> <span class="fl">1</span>, y <span class="op">=</span> <span class="fl">4</span>, xend <span class="op">=</span> <span class="fl">1</span>, yend <span class="op">=</span> <span class="fl">9.4</span><span class="op">)</span>, linetype <span class="op">=</span> <span class="fl">2</span>, color <span class="op">=</span> <span class="st">"black"</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/annotate.html">annotate</a></span><span class="op">(</span>geom <span class="op">=</span> <span class="st">"text"</span>, x <span class="op">=</span> <span class="fl">1</span>, y <span class="op">=</span> <span class="fl">10</span>, label<span class="op">=</span> <span class="st">"Weath of \n Brown \n agent"</span>, color<span class="op">=</span><span class="st">"black"</span>, size <span class="op">=</span> <span class="fl">4</span><span class="op">)</span> <span class="op">+</span>
<span class="co">#annotate(geom = "text", x = 0.2, y = 0.5, label=TeX(" No \n green \n price"), color="#0DCD64", size = 5) +</span>
<span class="co">#annotate(geom = "text", x = 2.4, y = 0.5, label=TeX(" No \n brown \n price"), color="#735E50", size = 5) +</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/geom_point.html">geom_point</a></span><span class="op">(</span>size <span class="op">=</span> <span class="fl">2.5</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">xlab</a></span><span class="op">(</span><span class="st">"Wealth of Green agent"</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggtheme.html">theme_light</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/theme.html">theme</a></span><span class="op">(</span>text <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_text</a></span><span class="op">(</span>size <span class="op">=</span> <span class="fl">14</span><span class="op">)</span>,
legend.position <span class="op">=</span> <span class="st">"None"</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/scale_manual.html">scale_color_manual</a></span><span class="op">(</span>values <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"#735E50"</span>, <span class="st">"#0DCD64"</span><span class="op">)</span>,
labels <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"Brown, e = 0"</span>, <span class="st">"Green, e = 1"</span><span class="op">)</span>, name <span class="op">=</span> <span class="st">"Asset"</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">ylab</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_blank</a></span><span class="op">(</span><span class="op">)</span><span class="op">)</span>
<span class="va">g1</span><span class="op">+</span><span class="va">g2</span></code></pre></div>
<div class="figure">
<span style="display:block;" id="fig:thmod1"></span>
<img src="ESG_p_files/figure-html/thmod1-1.png" alt="Theoretical predictions. We plot the price of two assets as a function of the wealth of the Green investor (the wealth of the Brown investor is kept constant at unit value). Prices for the green asset (e=1) are shown with triangles, while those for the brown asset (e=0) are shown with circles. For some values of the x-axis, prices may not be defined in the left plot." width="672"><p class="caption">
FIGURE 4.1: Theoretical predictions. We plot the price of two assets as a function of the wealth of the Green investor (the wealth of the Brown investor is kept constant at unit value). Prices for the green asset (e=1) are shown with triangles, while those for the brown asset (e=0) are shown with circles. For some values of the x-axis, prices may not be defined in the left plot.
</p>
</div>
<p>Note that in the left panel, when Green demand is too small (on the left), green prices are not defined. Reversely, when Green demand is too large, brown prices are not valued (right part of the graph). When tastes are less diverse (right panel), these issues vanish.</p>
</div>
</div>
<div id="sri-improves-performance" class="section level2" number="4.2">
<h2>
<span class="header-section-number">4.2</span> SRI improves performance<a class="anchor" aria-label="anchor" href="#sri-improves-performance"><i class="fas fa-link"></i></a>
</h2>
<p>
A large number of published work comes to the conclusion that sustainable investments are more profitable than aggregate benchmarks or even unethical portfolios. In fact, according to the survey <span class="citation"><a href="Bib.html#ref-friede2015esg" role="doc-biblioref">Friede, Busch, and Bassen</a> (<a href="Bib.html#ref-friede2015esg" role="doc-biblioref">2015</a>)</span>, 90% of papers report a positive relationship between performance and the propensity to tilt portfolios towards ESG stocks. As early as in <span class="citation"><a href="Bib.html#ref-klassen1996impact" role="doc-biblioref">Klassen and McLaughlin</a> (<a href="Bib.html#ref-klassen1996impact" role="doc-biblioref">1996</a>)</span>, it is found that environmentally friendly corporate management is rewarded by positive returns. In another seminal article, <span class="citation"><a href="Bib.html#ref-gompers2003corporate" role="doc-biblioref">Gompers, Ishii, and Metrick</a> (<a href="Bib.html#ref-gompers2003corporate" role="doc-biblioref">2003</a>)</span> show that corporate governance is a strong (positive) driver of returns in the cross-section. Prior to that, <span class="citation"><a href="Bib.html#ref-core1999corporate" role="doc-biblioref">Core, Holthausen, and Larcker</a> (<a href="Bib.html#ref-core1999corporate" role="doc-biblioref">1999</a>)</span> also concluded that weak governance is detrimental to financial performance. In a related study, <span class="citation"><a href="Bib.html#ref-auer2016socially" role="doc-biblioref">Auer</a> (<a href="Bib.html#ref-auer2016socially" role="doc-biblioref">2016</a>)</span> find that governance-related screens improve performance and that, in fact, firms that have higher ESG ratings earn higher returns.</p>
<p>
In another influential series of papers, <span class="citation"><a href="Bib.html#ref-edmans2011does" role="doc-biblioref">Edmans</a> (<a href="Bib.html#ref-edmans2011does" role="doc-biblioref">2011</a>)</span>, <span class="citation"><a href="Bib.html#ref-edmans2012link" role="doc-biblioref">Edmans</a> (<a href="Bib.html#ref-edmans2012link" role="doc-biblioref">2012</a>)</span> reports that portfolios of firms with highly satisfied employees generate significant alpha, though this could stem from prior underperformance (<span class="citation"><a href="Bib.html#ref-celiker2021undervaluation" role="doc-biblioref">Celiker, Sonaer, et al.</a> (<a href="Bib.html#ref-celiker2021undervaluation" role="doc-biblioref">2021</a>)</span>).<a class="footnote-ref" tabindex="0" data-toggle="popover" data-content='<p>For a critical essay on stakeholder governance, we refer to <span class="citation"><a href="Bib.html#ref-bebchuk2020illusory" role="doc-biblioref">L. A. Bebchuk and Tallarita</a> (<a href="Bib.html#ref-bebchuk2020illusory" role="doc-biblioref">2020</a>)</span>. Also, at the aggregate level, <span class="citation"><a href="Bib.html#ref-chen2020employee" role="doc-biblioref">J. Chen et al.</a> (<a href="Bib.html#ref-chen2020employee" role="doc-biblioref">2020</a>)</span> and symitsi2021employee find that employee sentiment is a predictor of stock returns (possibly a negative predictor).</p>'><sup>27</sup></a> More fundamentally, <span class="citation"><a href="Bib.html#ref-mervelskemper2017enhancing" role="doc-biblioref">Mervelskemper and Streit</a> (<a href="Bib.html#ref-mervelskemper2017enhancing" role="doc-biblioref">2017</a>)</span> find that it is always beneficial for firms to communicate on their ESG policies and disclose related reports and indicators. Notably, this motivates employees, who, in turn, are more productive (<span class="citation"><a href="Bib.html#ref-burbano2019getting" role="doc-biblioref">Burbano</a> (<a href="Bib.html#ref-burbano2019getting" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-hedblom2019toward" role="doc-biblioref">Hedblom, Hickman, and List</a> (<a href="Bib.html#ref-hedblom2019toward" role="doc-biblioref">2021</a>)</span>). </p>
<p>Many additional contributions underline the benefits that can be extracted when resorting to ESG data in the allocation process, often by selecting those firms with the highest scores. We list a few references in chronological order below in Table <a href="Perf.html#tab:positive">4.1</a>.</p>
<div class="inline-table"><table class="table table-sm">
<caption>
<span id="tab:positive">TABLE 4.1: </span> Contributions that conclude to a positive relationship between ESG and financial performance</caption>
<colgroup>
<col width="50%">
<col width="50%">
</colgroup>
<thead><tr class="header">
<th>Reference</th>
<th>Notable finding</th>
</tr></thead>
<tbody>
<tr class="odd">
<td><img width="250/"></td>
<td><img width="580/"></td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-cremers2005governance" role="doc-biblioref">K. M. Cremers and Nair</a> (<a href="Bib.html#ref-cremers2005governance" role="doc-biblioref">2005</a>)</span></td>
<td>A portfolio that is long firms with high level of takeover vulnerability and short firms with low levels of takeover vulnerability generates an annualized abnormal return of 10 to 15% if public pension fund (blockholder) ownership is high.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-derwall2005eco" role="doc-biblioref">Derwall et al.</a> (<a href="Bib.html#ref-derwall2005eco" role="doc-biblioref">2005</a>)</span></td>
<td>Eco-efficient portfolios (consisting of firms that are more environmentally friendly) generate higher returns compared to non-eco-effficient strategies.</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-kempf2007effect" role="doc-biblioref">Kempf and Osthoff</a> (<a href="Bib.html#ref-kempf2007effect" role="doc-biblioref">2007</a>)</span></td>
<td>Buying high ESG stocks and selling their low ESG counterparts yields annual returns of 8.7% on average.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-evans2010relationship" role="doc-biblioref">Evans and Peiris</a> (<a href="Bib.html#ref-evans2010relationship" role="doc-biblioref">2010</a>)</span></td>
<td>ESG is positively linked to stock returns, stock valuation and operating performance.</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-gil2010performance" role="doc-biblioref">Gil-Bazo, Ruiz-Verdú, and Santos</a> (<a href="Bib.html#ref-gil2010performance" role="doc-biblioref">2010</a>)</span></td>
<td>SRI funds perform better than their conventional counterparts, even after management fees, but the outperformance comes from specialized funds (see also <span class="citation"><a href="Bib.html#ref-filbeck2016socially" role="doc-biblioref">G. Filbeck, Krause, and Reis</a> (<a href="Bib.html#ref-filbeck2016socially" role="doc-biblioref">2016</a>)</span>). SRI vehicles from general funds underperform.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-giroud2011corporate" role="doc-biblioref">Giroud and Mueller</a> (<a href="Bib.html#ref-giroud2011corporate" role="doc-biblioref">2011</a>)</span></td>
<td>“<em>Weak governance firms have lower equity returns, worse operating performance, and lower firm value, but only in noncompetitive industries.</em>”</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-deng2013corporate" role="doc-biblioref">Deng, Kang, and Low</a> (<a href="Bib.html#ref-deng2013corporate" role="doc-biblioref">2013</a>)</span></td>
<td>In M&A deals, high ESG acquirers generate higher returns and have higher success rates.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-eccles2014impact" role="doc-biblioref">Eccles, Ioannou, and Serafeim</a> (<a href="Bib.html#ref-eccles2014impact" role="doc-biblioref">2014</a>)</span></td>
<td>Highly sustainable firms outperform poorly sustainable firms both in terms of stock market and accounting performance.</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-cai2014corporate" role="doc-biblioref">L. Cai and He</a> (<a href="Bib.html#ref-cai2014corporate" role="doc-biblioref">2014</a>)</span></td>
<td>It takes time for ESG to pay out. The paper reports that profitability comes after 3 years but not before.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-matsumura2014firm" role="doc-biblioref">Matsumura, Prakash, and Vera-Munoz</a> (<a href="Bib.html#ref-matsumura2014firm" role="doc-biblioref">2014</a>)</span></td>
<td>On average, on a 2006-2008 panel of 256 firms, for every additional thousand metric tons of carbon emissions, market capitalization decreases by $212,000.</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-dimson2015active" role="doc-biblioref">Dimson, Karakaş, and Li</a> (<a href="Bib.html#ref-dimson2015active" role="doc-biblioref">2015</a>)</span></td>
<td>According to the article, firms that commit to successful ESG engagements benefit from positive abnormal returns.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-kruger2015corporate" role="doc-biblioref">Krüger</a> (<a href="Bib.html#ref-kruger2015corporate" role="doc-biblioref">2015</a>)</span></td>
<td>Investors respond strongly negatively to negative CSR events, thereby sanctioning bad firms with negative returns. Also, stocks that face severe ESG controversies significantly underperform their benchmarks (<span class="citation"><a href="Bib.html#ref-de2020esg" role="doc-biblioref">Franco</a> (<a href="Bib.html#ref-de2020esg" role="doc-biblioref">2020</a>)</span>). However, investors may be tempted to overreact to these ESG controversies (<span class="citation"><a href="Bib.html#ref-cui2020stock" role="doc-biblioref">B. Cui and Docherty</a> (<a href="Bib.html#ref-cui2020stock" role="doc-biblioref">2020</a>)</span>).</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-nagy2016can" role="doc-biblioref">Nagy, Kassam, and Lee</a> (<a href="Bib.html#ref-nagy2016can" role="doc-biblioref">2016</a>)</span></td>
<td>ESG portfolios have superior alpha, compared to a global (MSCI World) benchmark.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-verheyden2016esg" role="doc-biblioref">Verheyden, Eccles, and Feiner</a> (<a href="Bib.html#ref-verheyden2016esg" role="doc-biblioref">2016</a>)</span></td>
<td>ESG screens improve risk-adjusted returns.</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-price2017doing" role="doc-biblioref">Price and Sun</a> (<a href="Bib.html#ref-price2017doing" role="doc-biblioref">2017</a>)</span></td>
<td>CSR is positively rewarded and CSI is penalized, but the effects of CSI last longer.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-lending2018corporate" role="doc-biblioref">Lending, Minnick, and Schorno</a> (<a href="Bib.html#ref-lending2018corporate" role="doc-biblioref">2018</a>)</span></td>
<td>Sustainable firms with small boards face lower odds of data breaches, which is beneficial for market returns. index{board}</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-khan2019corporate" role="doc-biblioref">Khan</a> (<a href="Bib.html#ref-khan2019corporate" role="doc-biblioref">2019</a>)</span></td>
<td>The authors report: “<em>In the cross-section, forward stock returns increased monotonically across governance and ESG quartiles</em>.”</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-kumar2019climate" role="doc-biblioref">Kumar, Xin, and Zhang</a> (<a href="Bib.html#ref-kumar2019climate" role="doc-biblioref">2019</a>)</span></td>
<td>Firms with higher climate change exposure experience lower subsequent returns.</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-li2019learning" role="doc-biblioref">Z. Li et al.</a> (<a href="Bib.html#ref-li2019learning" role="doc-biblioref">2019</a>)</span></td>
<td>Best CSR firms earn positive abnormal returns and are more likely to have positive earnings surprises. </td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-awaysheh2020relation" role="doc-biblioref">Awaysheh et al.</a> (<a href="Bib.html#ref-awaysheh2020relation" role="doc-biblioref">2020</a>)</span></td>
<td>Best-in-class firms (in terms of ESG scores) outperform their industry peers.</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-ravina2020impact" role="doc-biblioref">Ravina and Hentati Kaffel</a> (<a href="Bib.html#ref-ravina2020impact" role="doc-biblioref">2020</a>)</span></td>
<td>The Green-minus-Carbon factor is rewarded in Europe. In addition, it helps explain the cross-section of stocks by augmenting the 5-factor model proposed by <span class="citation"><a href="Bib.html#ref-fama2015five" role="doc-biblioref">Fama and French</a> (<a href="Bib.html#ref-fama2015five" role="doc-biblioref">2015</a>)</span>.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-serafeim2020public" role="doc-biblioref">Serafeim</a> (<a href="Bib.html#ref-serafeim2020public" role="doc-biblioref">2020</a>)</span></td>
<td>The article shows that to make ESG data profitable, it is useful to resort to public sentiment about firms’ sustainability performance.</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-madhavan2020toward" role="doc-biblioref">Madhavan, Sobczyk, and Ang</a> (<a href="Bib.html#ref-madhavan2020toward" role="doc-biblioref">2021</a>)</span></td>
<td>The paper analyzes factor loadings of ESG-tilted funds. They show that they are more exposed to the quality and momentum factors and that they also have high alphas.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-khajenouristandard" role="doc-biblioref">Khajenouri and Schmidt</a> (<a href="Bib.html#ref-khajenouristandard" role="doc-biblioref">2020</a>)</span></td>
<td>The authors compare conventional equity indices with their sustainable screened counterparts. The authors find that green indices outperform with respect to Sharpe ratios.</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-guest2020corporate" role="doc-biblioref">Guest and Nerino</a> (<a href="Bib.html#ref-guest2020corporate" role="doc-biblioref">2020</a>)</span></td>
<td>When a firm experiences a downgrade in governance rank from the Institutional Shareholder Services, negative returns follow.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-naffa2020performance" role="doc-biblioref">Naffa and Fain</a> (<a href="Bib.html#ref-naffa2020performance" role="doc-biblioref">2020</a>)</span></td>
<td>Nine ESG-tilted portfolios are built on sustainable themes. All of them yielded positive and significant alpha, even after transaction costs.</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-abate2021level" role="doc-biblioref">Abate, Basile, and Ferrari</a> (<a href="Bib.html#ref-abate2021level" role="doc-biblioref">2021</a>)</span></td>
<td>Mutual funds that invest in high ESG stocks perform better.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-wendt2021climate" role="doc-biblioref">Wendt et al.</a> (<a href="Bib.html#ref-wendt2021climate" role="doc-biblioref">2021</a>)</span></td>
<td>A portfolio that is long low emission firms and short high emission firms earns an annual alpha of 3.6% (equally-weighted portfolio) or 5.9% (value-weighted portfolio).</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-chu2021esg" role="doc-biblioref">Chu et al.</a> (<a href="Bib.html#ref-chu2021esg" role="doc-biblioref">2021</a>)</span></td>
<td>An aggregate ESG index significantly predicts the aggregate market return (with a positive coefficient).</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-xu2021climate" role="doc-biblioref">J. Xu, Sun, and You</a> (<a href="Bib.html#ref-xu2021climate" role="doc-biblioref">2021</a>)</span></td>
<td>Firms with high climate change exposure (measured via the Palmer Drought Severity Index) experience lower future profitability</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-geczy2021esg" role="doc-biblioref">Geczy and Guerard</a> (<a href="Bib.html#ref-geczy2021esg" role="doc-biblioref">2021</a>)</span></td>
<td>“<em>Environmental ratings interact with those forecasted returns and produce excess returns both unconditionally and conditionally</em>” and high ESG stocks earn higher returns than low ESG stocks.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-havlinova2021corporate" role="doc-biblioref">Havlinova and Kukacka</a> (<a href="Bib.html#ref-havlinova2021corporate" role="doc-biblioref">2021</a>)</span></td>
<td>“<em>A 1% increase in the ESG Score is associated with an increase in share price between 0.8% and 0.9%</em>”</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-glossner2021repeat" role="doc-biblioref">Glossner</a> (<a href="Bib.html#ref-glossner2021repeat" role="doc-biblioref">2021</a>)</span></td>
<td>“<em>firms’ past ESG incident rates predict more future incidents, weaker profitability, and lower risk-adjusted stock returns</em>.”</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-chang2021esg" role="doc-biblioref">R. Chang et al.</a> (<a href="Bib.html#ref-chang2021esg" role="doc-biblioref">2021</a>)</span></td>
<td>An aggregate ESG index predicts the stock market index positively between 2009 and 2018.</td>
</tr>
<tr class="even">
<td><span class="citation"><a href="Bib.html#ref-kazdin2021climate" role="doc-biblioref">Kazdin et al.</a> (<a href="Bib.html#ref-kazdin2021climate" role="doc-biblioref">2021</a>)</span></td>
<td>Firms with low carbon intensities have high excess returns and high productivity.</td>
</tr>
<tr class="odd">
<td><span class="citation"><a href="Bib.html#ref-dor2022incorporating" role="doc-biblioref">Dor, Guan, and Sun</a> (<a href="Bib.html#ref-dor2022incorporating" role="doc-biblioref">2022</a>)</span></td>
<td>The ESG premium in US and European equities has been positive in the past decade, after controlling for other factor exposures.</td>
</tr>
</tbody>
</table></div>
<p>
In an attempt to unify several competing theories, <span class="citation"><a href="Bib.html#ref-giese2019foundations" role="doc-biblioref">Giese et al.</a> (<a href="Bib.html#ref-giese2019foundations" role="doc-biblioref">2019</a>)</span> list three channels through which ESG may positively affect performance: cash flows (ESG firms yield higher dividends), risk (ESG firms have lower tail risk) and valuation (ESG firms, via a lower cost of capital, have an increased value). In a related study, <span class="citation"><a href="Bib.html#ref-antoncic2020sustainable" role="doc-biblioref">Antoncic et al.</a> (<a href="Bib.html#ref-antoncic2020sustainable" role="doc-biblioref">2020</a>)</span> find that it is not necessarily the raw ESG ratings that matter, but also their dynamics. They show that firms that experience ESG momentum (when ESG scores increase) generate positive alpha (similarly, <span class="citation"><a href="Bib.html#ref-conen2019hidden" role="doc-biblioref">Conen and Hartmann</a> (<a href="Bib.html#ref-conen2019hidden" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-shanaev2021esg" role="doc-biblioref">Shanaev and Ghimire</a> (<a href="Bib.html#ref-shanaev2021esg" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-tsai2021changes" role="doc-biblioref">Tsai and Wu</a> (<a href="Bib.html#ref-tsai2021changes" role="doc-biblioref">2021</a>)</span> show that ESG score revisions matter). Changes in ESG ratings are found to have a significant impact on subsequent returns in <span class="citation"><a href="Bib.html#ref-gluck2021esg" role="doc-biblioref">Glück, Hübel, and Scholz</a> (<a href="Bib.html#ref-gluck2021esg" role="doc-biblioref">2021</a>)</span>. The authors conclude that an improvement in the E pillar can be seen (and can act) as a risk mitigating trigger. </p>
<p>Likewise, <span class="citation"><a href="Bib.html#ref-kim2020capitalizing" role="doc-biblioref">T. Kim and Kim</a> (<a href="Bib.html#ref-kim2020capitalizing" role="doc-biblioref">2020</a>)</span> report that firms that enhanced their environmental sustainability by adopting cleaner production practices earn significant alpha. On the other end of the spectrum, <span class="citation"><a href="Bib.html#ref-glossner2018price" role="doc-biblioref">Gloßner</a> (<a href="Bib.html#ref-glossner2018price" role="doc-biblioref">2021</a>)</span> reports that a portfolio of firms with ESG <em>incidents</em> (e.g., scandals, accidents, etc.) suffers from negative alpha. This may come from analysts downgrading their earning forecasts for these firms (<span class="citation"><a href="Bib.html#ref-derrien2021esg" role="doc-biblioref">Derrien et al.</a> (<a href="Bib.html#ref-derrien2021esg" role="doc-biblioref">2021</a>)</span>). <span class="citation"><a href="Bib.html#ref-waddock1997corporate" role="doc-biblioref">Waddock and Graves</a> (<a href="Bib.html#ref-waddock1997corporate" role="doc-biblioref">1997</a>)</span> even document a positive feedback loop: positive financial performance fuels ESG behaviors which in turn generate higher profitability (this optimistic finding was however subsequently challenged in the replication study by <span class="citation"><a href="Bib.html#ref-zhao2016revisiting" role="doc-biblioref">X. Zhao and Murrell</a> (<a href="Bib.html#ref-zhao2016revisiting" role="doc-biblioref">2016</a>)</span>). </p>
<p>Finally, <span class="citation"><a href="Bib.html#ref-bennani2018esg" role="doc-biblioref">Bennani et al.</a> (<a href="Bib.html#ref-bennani2018esg" role="doc-biblioref">2018</a>)</span> find that ESG has performed well <em>recently</em>. They document that all three pillars yield positive returns for long-short portfolios, but for different reasons: for E, the performance is driven by higher returns of the long leg, while for S and G, it is driven by the poor performance of the short legs.</p>
</div>
<div id="mixed" class="section level2" number="4.3">
<h2>
<span class="header-section-number">4.3</span> SRI does not impact performance<a class="anchor" aria-label="anchor" href="#mixed"><i class="fas fa-link"></i></a>
</h2>
<p>Theoretically, restricting the investment universe should be detrimental to the profitability of screened portfolios (see next subsection). One reason for this is that voluntarily omitting assets may increase the odds of missing fruitful opportunities. In fact, the reverse argument works as well: screening can also participate to exclude assets that will perform badly. All in all, these two effects compete and the aggregate outcome is unclear. Consequently, it is not surprising that many studies contend that SR investing does not really hurt financial performance (while also not improving it).</p>
<p>
In a rebuke to <span class="citation"><a href="Bib.html#ref-gompers2003corporate" role="doc-biblioref">Gompers, Ishii, and Metrick</a> (<a href="Bib.html#ref-gompers2003corporate" role="doc-biblioref">2003</a>)</span>, <span class="citation"><a href="Bib.html#ref-johnson2009reexamination" role="doc-biblioref">S. A. Johnson, Moorman, and Sorescu</a> (<a href="Bib.html#ref-johnson2009reexamination" role="doc-biblioref">2009</a>)</span> report that screening firms on governance ratings does not yield more profitable portfolios. Similarly, <span class="citation"><a href="Bib.html#ref-core2006does" role="doc-biblioref">Core, Guay, and Rusticus</a> (<a href="Bib.html#ref-core2006does" role="doc-biblioref">2006</a>)</span>, <span class="citation"><a href="Bib.html#ref-post2015women" role="doc-biblioref">Post and Byron</a> (<a href="Bib.html#ref-post2015women" role="doc-biblioref">2015</a>)</span> and <span class="citation"><a href="Bib.html#ref-amihud2017settling" role="doc-biblioref">Amihud, Schmid, and Solomon</a> (<a href="Bib.html#ref-amihud2017settling" role="doc-biblioref">2017</a>)</span> do not find any causal relationship between weak governance, gender balanced boards, or staggered boards, and lower market returns. At the aggregate level, the studies of <span class="citation"><a href="Bib.html#ref-hamilton1993doing" role="doc-biblioref">Hamilton, Jo, and Statman</a> (<a href="Bib.html#ref-hamilton1993doing" role="doc-biblioref">1993</a>)</span>, <span class="citation"><a href="Bib.html#ref-guerard1997there" role="doc-biblioref">Guerard Jr</a> (<a href="Bib.html#ref-guerard1997there" role="doc-biblioref">1997</a>)</span>, <span class="citation"><a href="Bib.html#ref-statman2000socially" role="doc-biblioref">Statman</a> (<a href="Bib.html#ref-statman2000socially" role="doc-biblioref">2000</a>)</span>, <span class="citation"><a href="Bib.html#ref-bauer2005international" role="doc-biblioref">Bauer, Koedijk, and Otten</a> (<a href="Bib.html#ref-bauer2005international" role="doc-biblioref">2005</a>)</span>, <span class="citation"><a href="Bib.html#ref-bello2005socially" role="doc-biblioref">Bello</a> (<a href="Bib.html#ref-bello2005socially" role="doc-biblioref">2005</a>)</span>, <span class="citation"><a href="Bib.html#ref-dolvin2019good" role="doc-biblioref">Dolvin, Fulkerson, and Krukover</a> (<a href="Bib.html#ref-dolvin2019good" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-plagge2020have" role="doc-biblioref">Plagge and Grim</a> (<a href="Bib.html#ref-plagge2020have" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-wee2020effect" role="doc-biblioref">Wee et al.</a> (<a href="Bib.html#ref-wee2020effect" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-curtis2021esg" role="doc-biblioref">Curtis, Fisch, and Robertson</a> (<a href="Bib.html#ref-curtis2021esg" role="doc-biblioref">2021</a>)</span> (on mutual funds), <span class="citation"><a href="Bib.html#ref-sharma2021there" role="doc-biblioref">Sharma et al.</a> (<a href="Bib.html#ref-sharma2021there" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-milonas2022performance" role="doc-biblioref">Milonas, Rompotis, and Moutzouris</a> (<a href="Bib.html#ref-milonas2022performance" role="doc-biblioref">2022</a>)</span>, as well as the survey <span class="citation"><a href="Bib.html#ref-burghof2021investing" role="doc-biblioref">Burghof and Gehrung</a> (<a href="Bib.html#ref-burghof2021investing" role="doc-biblioref">2021</a>)</span> and the meta analysis of <span class="citation"><a href="Bib.html#ref-hornuf2022performance" role="doc-biblioref">Hornuf, Yüksel, et al.</a> (<a href="Bib.html#ref-hornuf2022performance" role="doc-biblioref">2022</a>)</span> compare ethical funds to conventional ones and do not find significant differences in performance metrics (notably, average returns, Sharpe ratio and diversification measures). In the same vein, <span class="citation"><a href="Bib.html#ref-capelle2019every" role="doc-biblioref">Capelle-Blancard and Petit</a> (<a href="Bib.html#ref-capelle2019every" role="doc-biblioref">2019</a>)</span> report that markets react insignificantly both to positive and negative ESG news.</p>
<p>Furthermore, responsible screening processes in portfolio construction do not deteriorate performance (see <span class="citation"><a href="Bib.html#ref-basso2014constant" role="doc-biblioref">Basso and Funari</a> (<a href="Bib.html#ref-basso2014constant" role="doc-biblioref">2014</a>)</span>, <span class="citation"><a href="Bib.html#ref-humphrey2012does" role="doc-biblioref">J. E. Humphrey, Lee, and Shen</a> (<a href="Bib.html#ref-humphrey2012does" role="doc-biblioref">2012</a>)</span>, <span class="citation"><a href="Bib.html#ref-humphrey2014does" role="doc-biblioref">J. E. Humphrey and Tan</a> (<a href="Bib.html#ref-humphrey2014does" role="doc-biblioref">2014</a>)</span>, <span class="citation"><a href="Bib.html#ref-blankenberg2018socially" role="doc-biblioref">Blankenberg and Gottschalk</a> (<a href="Bib.html#ref-blankenberg2018socially" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-cai2021socially" role="doc-biblioref">L. Cai, Cooper, and He</a> (<a href="Bib.html#ref-cai2021socially" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-chava2021doing" role="doc-biblioref">Chava, Kim, and Lee</a> (<a href="Bib.html#ref-chava2021doing" role="doc-biblioref">2021</a>)</span>), but they do not improve it either (<span class="citation"><a href="Bib.html#ref-gibson2019responsible" role="doc-biblioref">Gibson et al.</a> (<a href="Bib.html#ref-gibson2019responsible" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-pyles2020examining" role="doc-biblioref">Pyles</a> (<a href="Bib.html#ref-pyles2020examining" role="doc-biblioref">2020</a>)</span> and <span class="citation"><a href="Bib.html#ref-gorgen2021get" role="doc-biblioref">Görgen, Jacob, and Nerlinger</a> (<a href="Bib.html#ref-gorgen2021get" role="doc-biblioref">2021</a>)</span>). <span class="citation"><a href="Bib.html#ref-sautner2021pricing" role="doc-biblioref">Sautner et al.</a> (<a href="Bib.html#ref-sautner2021pricing" role="doc-biblioref">2021b</a>)</span> report a zero unconditional risk premium associated to climate change risk. In some sectors (e.g., banking), there is no significant link between CSR and CFP (<span class="citation"><a href="Bib.html#ref-soana2011relationship" role="doc-biblioref">Soana</a> (<a href="Bib.html#ref-soana2011relationship" role="doc-biblioref">2011</a>)</span> – though <span class="citation"><a href="Bib.html#ref-buataerelationship" role="doc-biblioref">Bătae, Dragomir, and Feleagă</a> (<a href="Bib.html#ref-buataerelationship" role="doc-biblioref">2021</a>)</span> do find some links depending on the components of ESG).</p>
<p>Likewise, market neutral portfolios built on ESG metrics are neither beneficial nor detrimental to profits (<span class="citation"><a href="Bib.html#ref-breedt2019esg" role="doc-biblioref">Breedt et al.</a> (<a href="Bib.html#ref-breedt2019esg" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-dai2020greenhouse" role="doc-biblioref">W. Dai and Meyer-Brauns</a> (<a href="Bib.html#ref-dai2020greenhouse" role="doc-biblioref">2021</a>)</span>), or, equivalently, have insignificant average returns (<span class="citation"><a href="Bib.html#ref-kaiser2020esg" role="doc-biblioref">Kaiser</a> (<a href="Bib.html#ref-kaiser2020esg" role="doc-biblioref">2020b</a>)</span>). Furthermore, factor models seem to confirm that ESG is not priced (<span class="citation"><a href="Bib.html#ref-xiao2017financial" role="doc-biblioref">Xiao et al.</a> (<a href="Bib.html#ref-xiao2017financial" role="doc-biblioref">2017</a>)</span>), or add no incremental value, compared to traditional factors (<span class="citation"><a href="Bib.html#ref-naffa2021factor" role="doc-biblioref">Naffa and Fain</a> (<a href="Bib.html#ref-naffa2021factor" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-husse2021responsible" role="doc-biblioref">Husse and Pippo</a> (<a href="Bib.html#ref-husse2021responsible" role="doc-biblioref">2021</a>)</span>). In <span class="citation"><a href="Bib.html#ref-bruno2021honey" role="doc-biblioref">Bruno, Esakia, and Goltz</a> (<a href="Bib.html#ref-bruno2021honey" role="doc-biblioref">2021</a>)</span> acknowledge that the performance of ESG leaders is appealing, unless other traditional asset pricing factors are taken into account. Indeed, the alpha of ESG portfolios disappears when controlling for the size, value, momentum, low volatility, profitability and investment factors. Overall, the meta study <span class="citation"><a href="Bib.html#ref-kim2019can" role="doc-biblioref">C.-S. Kim</a> (<a href="Bib.html#ref-kim2019can" role="doc-biblioref">2019</a>)</span> confirms that “<em>SRI performance is not different from conventional investments</em>.” <span class="citation"><a href="Bib.html#ref-knoll2002ethical" role="doc-biblioref">Knoll</a> (<a href="Bib.html#ref-knoll2002ethical" role="doc-biblioref">2002</a>)</span> provides a theoretical argument to explain this lack of impact: the long-run demand curves for the securities of individual firms are not very steep, hence ethical screening does not shift prices much in the long run.</p>
<p>
<span class="citation"><a href="Bib.html#ref-baron2011economics" role="doc-biblioref">Baron, Harjoto, and Jo</a> (<a href="Bib.html#ref-baron2011economics" role="doc-biblioref">2011</a>)</span> estimate a broad economic model in which firms have to deal with three markets: the financial market that prices their values, the customer market that purchases their products and a market of social pressure that incentivizes them to pursue sustainable policies. They find that CFP is uncorrelated with CSP and negatively correlated with social pressure. </p>
<p>Finally, we single out a few additional studies. The early meta-analysis of <span class="citation"><a href="Bib.html#ref-arlow1982social" role="doc-biblioref">Arlow and Gannon</a> (<a href="Bib.html#ref-arlow1982social" role="doc-biblioref">1982</a>)</span> finds little evidence that links social responsiveness to economic performance. <span class="citation"><a href="Bib.html#ref-ng2018let" role="doc-biblioref">A. Ng and Zheng</a> (<a href="Bib.html#ref-ng2018let" role="doc-biblioref">2018</a>)</span> document that green energy firms have similar performance compared to non-green energy firms, while <span class="citation"><a href="Bib.html#ref-halbritter2015wages" role="doc-biblioref">Halbritter and Dorfleitner</a> (<a href="Bib.html#ref-halbritter2015wages" role="doc-biblioref">2015</a>)</span> show that high ESG minus low ESG portfolios do not reveal significantly positive returns. <span class="citation"><a href="Bib.html#ref-cheung2011stock" role="doc-biblioref">Cheung</a> (<a href="Bib.html#ref-cheung2011stock" role="doc-biblioref">2011</a>)</span> finds that stocks that are suddenly included or removed from the Dow Jones Sustainability World Index do not experience any particular shift in average return or risk. Via <span class="citation"><a href="Bib.html#ref-fama1973risk" role="doc-biblioref">Fama and MacBeth</a> (<a href="Bib.html#ref-fama1973risk" role="doc-biblioref">1973</a>)</span> regressions, <span class="citation"><a href="Bib.html#ref-timar2021does" role="doc-biblioref">Timár et al.</a> (<a href="Bib.html#ref-timar2021does" role="doc-biblioref">2021</a>)</span> also does not find any significant pricing ability of ESG scores. This is also true when accounting for errors-in-variables (<span class="citation"><a href="Bib.html#ref-auer2021implementation" role="doc-biblioref">Auer</a> (<a href="Bib.html#ref-auer2021implementation" role="doc-biblioref">2021</a>)</span>).
We sum up this section with one of the conclusions of <span class="citation"><a href="Bib.html#ref-schroder2004performance" role="doc-biblioref">Schröder</a> (<a href="Bib.html#ref-schroder2004performance" role="doc-biblioref">2004</a>)</span> (p. 130): “<em>socially screened assets seem to have no clear disadvantage concerning their performance compared to conventional assets</em>.”</p>
</div>
<div id="sri-is-financially-detrimental" class="section level2" number="4.4">
<h2>
<span class="header-section-number">4.4</span> SRI is financially detrimental<a class="anchor" aria-label="anchor" href="#sri-is-financially-detrimental"><i class="fas fa-link"></i></a>
</h2>
<p>At the other end of the spectrum, several studies argue that unethical firms tend to experience higher profitability. One such controversial article is that of <span class="citation"><a href="Bib.html#ref-fabozzi2008sin" role="doc-biblioref">Fabozzi, Ma, and Oliphant</a> (<a href="Bib.html#ref-fabozzi2008sin" role="doc-biblioref">2008</a>)</span>, which shows that sin stocks (belonging to the adult services, alcohol, biotech, defense, gaming and tobacco industries) produce a combined annual return of 19%, outperforming any reasonable benchmark.<a class="footnote-ref" tabindex="0" data-toggle="popover" data-content='<p>In a follow-up paper, <span class="citation"><a href="Bib.html#ref-blitz2017sin" role="doc-biblioref">Blitz and Fabozzi</a> (<a href="Bib.html#ref-blitz2017sin" role="doc-biblioref">2017</a>)</span>, it is nonetheless found that a five-factor model is able to explain this sin stock anomaly. Recently, <span class="citation"><a href="Bib.html#ref-blitz2021does" role="doc-biblioref">Blitz and Swinkels</a> (<a href="Bib.html#ref-blitz2021does" role="doc-biblioref">2021a</a>)</span> and <span class="citation"><a href="Bib.html#ref-sagbakken2021european" role="doc-biblioref">Sagbakken and Zhang</a> (<a href="Bib.html#ref-sagbakken2021european" role="doc-biblioref">2021</a>)</span> also find little evidence of a sin premium, but warn that it could materialize in the future, when sin firms will be asked higher costs of capital.</p>'><sup>28</sup></a> (<span class="citation"><a href="Bib.html#ref-hong2009price" role="doc-biblioref">H. Hong and Kacperczyk</a> (<a href="Bib.html#ref-hong2009price" role="doc-biblioref">2009</a>)</span> also find outperformance, but to a lesser extent, while <span class="citation"><a href="Bib.html#ref-dimson2020exclusionary" role="doc-biblioref">Dimson, Marsh, and Staunton</a> (<a href="Bib.html#ref-dimson2020exclusionary" role="doc-biblioref">2020b</a>)</span> do confirm superior performance of sin stocks over the long run, since 1900). Likewise, <span class="citation"><a href="Bib.html#ref-bolton2020investors" role="doc-biblioref">Bolton and Kacperczyk</a> (<a href="Bib.html#ref-bolton2020investors" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-bolton2020carbon" role="doc-biblioref">Bolton and Kacperczyk</a> (<a href="Bib.html#ref-bolton2020carbon" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-busch2020corporate" role="doc-biblioref">Busch et al.</a> (<a href="Bib.html#ref-busch2020corporate" role="doc-biblioref">2020</a>)</span> and <span class="citation"><a href="Bib.html#ref-santi2021carbon" role="doc-biblioref">Santi and Moretti</a> (<a href="Bib.html#ref-santi2021carbon" role="doc-biblioref">2021</a>)</span> find that firms with higher total CO<span class="math inline">\(_2\)</span> emissions (and changes in emissions) earn higher returns. The former authors conclude: “<em>investors are already demanding compensation for their exposure to carbon emission risk</em>.” Nevertheless, <span class="citation"><a href="Bib.html#ref-santi2021carbon" role="doc-biblioref">Santi and Moretti</a> (<a href="Bib.html#ref-santi2021carbon" role="doc-biblioref">2021</a>)</span> also mitigate this result when focusing on regions with low level of worries about climate change. In these regions, the carbon premium disappears.</p>
<p>Similarly, <span class="citation"><a href="Bib.html#ref-hsu2020pollution" role="doc-biblioref">P.-H. Hsu, Li, and Tsou</a> (<a href="Bib.html#ref-hsu2020pollution" role="doc-biblioref">2021</a>)</span> document that polluting firms experience higher average returns. Moreover, <span class="citation"><a href="Bib.html#ref-delmas2015dynamics" role="doc-biblioref">Delmas, Nairn-Birch, and Lim</a> (<a href="Bib.html#ref-delmas2015dynamics" role="doc-biblioref">2015</a>)</span> also report that improving corporate environmental performance reduces short-term financial performance. In the same vein, <span class="citation"><a href="Bib.html#ref-trinks2017opportunity" role="doc-biblioref">Trinks and Scholtens</a> (<a href="Bib.html#ref-trinks2017opportunity" role="doc-biblioref">2017</a>)</span> observe that investing in controversial stocks in many cases results in superior risk-adjusted returns. </p>
<p>Conversely, firms that make significant environmental efforts experience lower returns (<span class="citation"><a href="Bib.html#ref-fisher2011voluntary" role="doc-biblioref">Fisher-Vanden and Thorburn</a> (<a href="Bib.html#ref-fisher2011voluntary" role="doc-biblioref">2011</a>)</span>), just as do high ESG firms (<span class="citation"><a href="Bib.html#ref-luo2022esg" role="doc-biblioref">D. Luo</a> (<a href="Bib.html#ref-luo2022esg" role="doc-biblioref">2022</a>)</span>) and impact firms (<span class="citation"><a href="Bib.html#ref-bernal2021impact" role="doc-biblioref">Bernal, Hudon, and Ledru</a> (<a href="Bib.html#ref-bernal2021impact" role="doc-biblioref">2021</a>)</span>). One explanation is that reducing the carbon footprint or any source of pollution is too costly and outweighs the potential benefits (<span class="citation"><a href="Bib.html#ref-hart1996does" role="doc-biblioref">S. L. Hart and Ahuja</a> (<a href="Bib.html#ref-hart1996does" role="doc-biblioref">1996</a>)</span>). A second explanation is under-diversification. On the governance side, <span class="citation"><a href="Bib.html#ref-bohren2016mandatory" role="doc-biblioref">Bøhren and Staubo</a> (<a href="Bib.html#ref-bohren2016mandatory" role="doc-biblioref">2016</a>)</span> report that firms with mandatory gender parity on their boards experience lower returns (though a broader examination of the literature leads to a more nuanced conclusion, see <span class="citation"><a href="Bib.html#ref-post2015women" role="doc-biblioref">Post and Byron</a> (<a href="Bib.html#ref-post2015women" role="doc-biblioref">2015</a>)</span>). </p>
<p>At the aggregate level, in their study on university endowment funds, <span class="citation"><a href="Bib.html#ref-aragon2019socially" role="doc-biblioref">Aragon et al.</a> (<a href="Bib.html#ref-aragon2019socially" role="doc-biblioref">2020</a>)</span>} find that funds which enforce sustainable policies have greater volatility and the authors attribute underperformance to higher divestment costs and inefficient diversification. Similarly, <span class="citation"><a href="Bib.html#ref-anson2020sustainability" role="doc-biblioref">Anson et al.</a> (<a href="Bib.html#ref-anson2020sustainability" role="doc-biblioref">2020</a>)</span> find that sustainable funds have, on average, lower alphas compared to unconstrained funds (see also <span class="citation"><a href="Bib.html#ref-liang2020socially" role="doc-biblioref">Liang, Sun, and Teo</a> (<a href="Bib.html#ref-liang2020socially" role="doc-biblioref">2022</a>)</span>, who show that, nonetheless, ethical hedge funds attract more flows). <span class="citation"><a href="Bib.html#ref-bofinger2021sustainability" role="doc-biblioref">Bofinger et al.</a> (<a href="Bib.html#ref-bofinger2021sustainability" role="doc-biblioref">2021</a>)</span> find that active ESG funds face the risk of lowering their investment skills because of what the authors call the <em>sustainable trap</em> (the push towards ESG increases the risk of mispricing).
Similar findings are outlined in <span class="citation"><a href="Bib.html#ref-renneboog2008price" role="doc-biblioref">Renneboog, Ter Horst, and Zhang</a> (<a href="Bib.html#ref-renneboog2008price" role="doc-biblioref">2008b</a>)</span>, <span class="citation"><a href="Bib.html#ref-el2020green" role="doc-biblioref">El Ghoul et al.</a> (<a href="Bib.html#ref-el2020green" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-jeffers2021risk" role="doc-biblioref">Jeffers, Lyu, and Posenau</a> (<a href="Bib.html#ref-jeffers2021risk" role="doc-biblioref">2021</a>)</span> (for impact funds), though the underperformance in returns sometimes translates to statistically insignificantly lower Sharpe ratios. Finally, focusing on the debate on mandatory reductions of greenhouse gases in the U.S., <span class="citation"><a href="Bib.html#ref-hsu2013does" role="doc-biblioref">A. W. Hsu and Wang</a> (<a href="Bib.html#ref-hsu2013does" role="doc-biblioref">2013</a>)</span> report that markets react positively to negative ESG news at the company level. <span class="citation"><a href="Bib.html#ref-conen2019hidden" role="doc-biblioref">Conen and Hartmann</a> (<a href="Bib.html#ref-conen2019hidden" role="doc-biblioref">2019</a>)</span> document a reverse effect: “<em>markets react to ESG improvements of top ranked firms with negative abnormal returns</em>.” </p>
<p>From an asset pricing perspective, and according to risk factor analysis, investors should be rewarded for the risk they take when investing in stocks that are exposed to ESG externalities. This is one of the theoretical findings of <span class="citation"><a href="Bib.html#ref-pastor2020sustainable" role="doc-biblioref">Pastor, Stambaugh, and Taylor</a> (<a href="Bib.html#ref-pastor2020sustainable" role="doc-biblioref">2021b</a>)</span>: the capital asset pricing model (CAPM) alpha of stocks should be negatively proportional to their ESG scores. And indeed, <span class="citation"><a href="Bib.html#ref-chen2018urge" role="doc-biblioref">X. Chen and Scholtens</a> (<a href="Bib.html#ref-chen2018urge" role="doc-biblioref">2018</a>)</span> argue that both active and passive SRI funds have negative alphas. Empirically, <span class="citation"><a href="Bib.html#ref-brammer2006corporate" role="doc-biblioref">Brammer, Brooks, and Pavelin</a> (<a href="Bib.html#ref-brammer2006corporate" role="doc-biblioref">2006</a>)</span>, <span class="citation"><a href="Bib.html#ref-lee2009corporate" role="doc-biblioref">D. D. Lee and Faff</a> (<a href="Bib.html#ref-lee2009corporate" role="doc-biblioref">2009</a>)</span>, <span class="citation"><a href="Bib.html#ref-becchetti2018fishing" role="doc-biblioref">Becchetti, Ciciretti, and Dalò</a> (<a href="Bib.html#ref-becchetti2018fishing" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-lioui2018corporate" role="doc-biblioref">Lioui, Poncet, and Sisto</a> (<a href="Bib.html#ref-lioui2018corporate" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-lioui2018esg" role="doc-biblioref">Lioui</a> (<a href="Bib.html#ref-lioui2018esg" role="doc-biblioref">2018b</a>)</span>, <span class="citation"><a href="Bib.html#ref-lioui2018esg2" role="doc-biblioref">Lioui</a> (<a href="Bib.html#ref-lioui2018esg2" role="doc-biblioref">2018a</a>)</span>, <span class="citation"><a href="Bib.html#ref-ciciretti2019contributions" role="doc-biblioref">Ciciretti, Dalò, and Dam</a> (<a href="Bib.html#ref-ciciretti2019contributions" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-adriaan2020carbon" role="doc-biblioref">Adriaan Boermans and Galema</a> (<a href="Bib.html#ref-adriaan2020carbon" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-hubel2020integrating" role="doc-biblioref">Hübel and Scholz</a> (<a href="Bib.html#ref-hubel2020integrating" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-lucia2020greenium" role="doc-biblioref">Alessi, Ossola, and Panzica</a> (<a href="Bib.html#ref-lucia2020greenium" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-alessi2021greenium" role="doc-biblioref">Alessi, Ossola, and Panzica</a> (<a href="Bib.html#ref-alessi2021greenium" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-cakici2021responsible" role="doc-biblioref">Cakici and Zaremba</a> (<a href="Bib.html#ref-cakici2021responsible" role="doc-biblioref">2022</a>)</span> all find that the rewarded ESG factors go long irresponsible firms and short responsible ones. The latter documents that a significant part of the premium comes from small firms (which disclose typically less). Similarly, a portfolio, long sin stocks and short non-sin stock earns a monthly return of 1.33% on average (<span class="citation"><a href="Bib.html#ref-luo2017social" role="doc-biblioref">H. A. Luo and Balvers</a> (<a href="Bib.html#ref-luo2017social" role="doc-biblioref">2017</a>)</span>). In <span class="citation"><a href="Bib.html#ref-kanuri2020risk" role="doc-biblioref">Kanuri</a> (<a href="Bib.html#ref-kanuri2020risk" role="doc-biblioref">2020</a>)</span>, it is found that in the long run, conventional funds outperform ESG funds (in terms of average returns and Sharpe ratio), even though the latter sometimes fare better. </p>
<p>
Also, from an optimization standpoint, using screening processes reduces the investment set. By construction, this shifts the efficient frontier towards a smaller enveloppe, which is, by definition, suboptimal – at least in-sample. This implies the opportunity cost of renouncing to potential profitable assets. For a theoretical discussion on this subject, we refer to <span class="citation"><a href="Bib.html#ref-pedersen2020responsible" role="doc-biblioref">Pedersen, Fitzgibbons, and Pomorski</a> (<a href="Bib.html#ref-pedersen2020responsible" role="doc-biblioref">2021</a>)</span>. In particular, <span class="citation"><a href="Bib.html#ref-chang2010performance" role="doc-biblioref">C. E. Chang and Witte</a> (<a href="Bib.html#ref-chang2010performance" role="doc-biblioref">2010</a>)</span> argue that ESG investing shrinks both average returns and Sharpe ratios, compared to unscreened benchmarks.</p>
<p>Finally, in an original contribution, <span class="citation"><a href="Bib.html#ref-jorgensen2021cost" role="doc-biblioref">Jørgensen and Plovst</a> (<a href="Bib.html#ref-jorgensen2021cost" role="doc-biblioref">2021</a>)</span> analyze the hedging cost of sustainability. They measure the price of an insurance derivative that would protect an ESG investor against the underperformance of a green fund versus a conventional fund. The authors find that the incurred cost lies between –0.5% and –3% in terms of annual returns.</p>
</div>
<div id="it-depends" class="section level2" number="4.5">
<h2>
<span class="header-section-number">4.5</span> It depends<a class="anchor" aria-label="anchor" href="#it-depends"><i class="fas fa-link"></i></a>
</h2>
<p>Not surprisingly, SRI does not deliver performance unconditionally, if and when it does. There are in fact many drivers of this performance and the profitability of ESG strategy can depend on several factors, which we list below. The meta-analysis <span class="citation"><a href="Bib.html#ref-hang2018economic" role="doc-biblioref">Hang et al.</a> (<a href="Bib.html#ref-hang2018economic" role="doc-biblioref">2018</a>)</span> is a valuable resource on this topic.</p>
<div id="dimensions" class="section level3" number="4.5.1">
<h3>
<span class="header-section-number">4.5.1</span> Dimensions<a class="anchor" aria-label="anchor" href="#dimensions"><i class="fas fa-link"></i></a>
</h3>
<p>First of all, not all <strong>dimensions</strong> of ESG are equal. Some studies find that governance screens work well (<span class="citation"><a href="Bib.html#ref-gompers2003corporate" role="doc-biblioref">Gompers, Ishii, and Metrick</a> (<a href="Bib.html#ref-gompers2003corporate" role="doc-biblioref">2003</a>)</span>, <span class="citation"><a href="Bib.html#ref-bebchuk2009matters" role="doc-biblioref">L. Bebchuk, Cohen, and Ferrell</a> (<a href="Bib.html#ref-bebchuk2009matters" role="doc-biblioref">2009</a>)</span>, <span class="citation"><a href="Bib.html#ref-auer2016socially" role="doc-biblioref">Auer</a> (<a href="Bib.html#ref-auer2016socially" role="doc-biblioref">2016</a>)</span>, <span class="citation"><a href="Bib.html#ref-bruder2019integration" role="doc-biblioref">Bruder et al.</a> (<a href="Bib.html#ref-bruder2019integration" role="doc-biblioref">2019</a>)</span> and <span class="citation"><a href="Bib.html#ref-lee2020combining" role="doc-biblioref">L.-E. Lee, Giese, and Nagy</a> (<a href="Bib.html#ref-lee2020combining" role="doc-biblioref">2020</a>)</span>, but environmental screens do not (<span class="citation"><a href="Bib.html#ref-auer2016socially" role="doc-biblioref">Auer</a> (<a href="Bib.html#ref-auer2016socially" role="doc-biblioref">2016</a>)</span>, <span class="citation"><a href="Bib.html#ref-alareeni2020esg" role="doc-biblioref">Alareeni and Hamdan</a> (<a href="Bib.html#ref-alareeni2020esg" role="doc-biblioref">2020</a>)</span>). <span class="citation"><a href="Bib.html#ref-lepetit2021recent" role="doc-biblioref">Lepetit et al.</a> (<a href="Bib.html#ref-lepetit2021recent" role="doc-biblioref">2021</a>)</span> document some trends for each pillars between 2014 and 2021 in Europe and the US. <span class="citation"><a href="Bib.html#ref-frambo2022esg" role="doc-biblioref">Frambo and Kok</a> (<a href="Bib.html#ref-frambo2022esg" role="doc-biblioref">2022</a>)</span> study the impact of each pillar on valuation and returns during the COVID 2020 market crash.</p>
<p>In fact, even within ratings, some provisions, or subcategories are more impactful than others (see <span class="citation"><a href="Bib.html#ref-bebchuk2009matters" role="doc-biblioref">L. Bebchuk, Cohen, and Ferrell</a> (<a href="Bib.html#ref-bebchuk2009matters" role="doc-biblioref">2009</a>)</span> and <span class="citation"><a href="Bib.html#ref-becchetti2018fishing" role="doc-biblioref">Becchetti, Ciciretti, and Dalò</a> (<a href="Bib.html#ref-becchetti2018fishing" role="doc-biblioref">2018</a>)</span>). In a similar vein, <span class="citation"><a href="Bib.html#ref-ziegler2007effect" role="doc-biblioref">Ziegler, Schröder, and Rennings</a> (<a href="Bib.html#ref-ziegler2007effect" role="doc-biblioref">2007</a>)</span> finds that E is improving returns, while S is deteriorating them. Likewise, <span class="citation"><a href="Bib.html#ref-galema2008stocks" role="doc-biblioref">Galema, Plantinga, and Scholtens</a> (<a href="Bib.html#ref-galema2008stocks" role="doc-biblioref">2008</a>)</span> and <span class="citation"><a href="Bib.html#ref-jayachandran2013product" role="doc-biblioref">Jayachandran, Kalaignanam, and Eilert</a> (<a href="Bib.html#ref-jayachandran2013product" role="doc-biblioref">2013</a>)</span> show that some ratings can be beneficial (diversity, environment and product), while others cannot. Within each branch of ESG, some component may also prove more relevant than others. <span class="citation"><a href="Bib.html#ref-jacobs2010empirical" role="doc-biblioref">Jacobs, Singhal, and Subramanian</a> (<a href="Bib.html#ref-jacobs2010empirical" role="doc-biblioref">2010</a>)</span> find that philanthropic gifts for environmental causes are associated with significant positive market reaction, while it is the opposite for voluntary emission reduction.</p>
<p>Environmental commitments are not all equal: <span class="citation"><a href="Bib.html#ref-king2002exploring" role="doc-biblioref">A. King and Lenox</a> (<a href="Bib.html#ref-king2002exploring" role="doc-biblioref">2002</a>)</span> find that prevention leads to financial gain, but not pollution reduction. Based on KLD data, <span class="citation"><a href="Bib.html#ref-geczy2019efficient" role="doc-biblioref">Geczy, Guerard, and Samonov</a> (<a href="Bib.html#ref-geczy2019efficient" role="doc-biblioref">2019</a>)</span> show that Human Rights, and Diversity criteria contribute to enhancing portfolio returns. <span class="citation"><a href="Bib.html#ref-filbeck2019performance" role="doc-biblioref">A. Filbeck, Filbeck, and Zhao</a> (<a href="Bib.html#ref-filbeck2019performance" role="doc-biblioref">2019</a>)</span> find that G is beneficial, E is detrimental and S is not impactful. <span class="citation"><a href="Bib.html#ref-giese2021deconstructing" role="doc-biblioref">Giese, Nagy, and Lee</a> (<a href="Bib.html#ref-giese2021deconstructing" role="doc-biblioref">2021</a>)</span> find that pillars can matter at different horizons (short for Governance, long for Social and Environmental). Even inside pillars, variables may have mitigating effects. <span class="citation"><a href="Bib.html#ref-naranjo2020influence" role="doc-biblioref">Naranjo Tuesta, Crespo Soler, and Ripoll Feliu</a> (<a href="Bib.html#ref-naranjo2020influence" role="doc-biblioref">2020</a>)</span> find that different types of policies on different types of emissions have contradicting effects on financial performance. According to <span class="citation"><a href="Bib.html#ref-tsai2021changes" role="doc-biblioref">Tsai and Wu</a> (<a href="Bib.html#ref-tsai2021changes" role="doc-biblioref">2021</a>)</span>, enhancing the environmental score leads to higher relative returns in crisis periods.</p>
<p>Moreover, depending on which fields are considered, results differ. <span class="citation"><a href="Bib.html#ref-aswani2021carbon" role="doc-biblioref">Aswani, Raghunandan, and Rajgopal</a> (<a href="Bib.html#ref-aswani2021carbon" role="doc-biblioref">2021</a>)</span> underline that academics often use the raw values of emissions (in metric tons of CO<span class="math inline">\(_2\)</span>), while it makes more sense to consider emission intensities, whereby emissions are scaled by some proxy of firm size (e.g., sales) – which may considerably alter conclusions. Furthermore, depending on the industry (see also below), some ESG fields may be considered as <em>financially material</em> (i.e., expected to have an impact on finances), while other are not. The study <span class="citation"><a href="Bib.html#ref-grewal2020material" role="doc-biblioref">Grewal, Hauptmann, and Serafeim</a> (<a href="Bib.html#ref-grewal2020material" role="doc-biblioref">2021</a>)</span> shows that materiality is a strong driver of price informativeness. </p>
<p>Beyond pure ESG fields, the propensity of firms to simply disclose ESG related data is also useful in the allocation process (<span class="citation"><a href="Bib.html#ref-d2021sustainable" role="doc-biblioref">D’Apice, Ferri, and Intonti</a> (<a href="Bib.html#ref-d2021sustainable" role="doc-biblioref">2021</a>)</span>). Finally, in <span class="citation"><a href="Bib.html#ref-de2021sustainable" role="doc-biblioref">De Franco, Nicolle, and Tran</a> (<a href="Bib.html#ref-de2021sustainable" role="doc-biblioref">2021</a>)</span>, the authors warn that the ESG dimensions are not the only ones in the field. According to them, the sustainable development goals (SDGs) allow to elaborate alternative metrics that capture other facets of sustainability. For an intelligible introduction on SDGs and their relationships with finance, we recommend the overview of <span class="citation"><a href="Bib.html#ref-zhan2021investing" role="doc-biblioref">Zhan and Santos-Paulino</a> (<a href="Bib.html#ref-zhan2021investing" role="doc-biblioref">2021</a>)</span>.</p>
</div>
<div id="geography" class="section level3" number="4.5.2">
<h3>
<span class="header-section-number">4.5.2</span> Geography<a class="anchor" aria-label="anchor" href="#geography"><i class="fas fa-link"></i></a>
</h3>
<p>A second important factor is <strong>geography</strong>. Many articles document contradicting results when switching from one market to another. Often, ESG strategies are shown to perform for one zone (e.g., US, Europe, or Asia), but not all. We refer for instance to <span class="citation"><a href="Bib.html#ref-cortez2012socially" role="doc-biblioref">Cortez, Silva, and Areal</a> (<a href="Bib.html#ref-cortez2012socially" role="doc-biblioref">2012</a>)</span>, <span class="citation"><a href="Bib.html#ref-von2014effect" role="doc-biblioref">Von Arx and Ziegler</a> (<a href="Bib.html#ref-von2014effect" role="doc-biblioref">2014</a>)</span>, <span class="citation"><a href="Bib.html#ref-post2015women" role="doc-biblioref">Post and Byron</a> (<a href="Bib.html#ref-post2015women" role="doc-biblioref">2015</a>)</span>, <span class="citation"><a href="Bib.html#ref-bruder2019integration" role="doc-biblioref">Bruder et al.</a> (<a href="Bib.html#ref-bruder2019integration" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-cheema2019decarbonization" role="doc-biblioref">Cheema-Fox et al.</a> (<a href="Bib.html#ref-cheema2019decarbonization" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-matallin2019ethical" role="doc-biblioref">Matallı́n-Sáez et al.</a> (<a href="Bib.html#ref-matallin2019ethical" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-de2020esg" role="doc-biblioref">Franco</a> (<a href="Bib.html#ref-de2020esg" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-edmans2020employee" role="doc-biblioref">Edmans, Li, and Zhang</a> (<a href="Bib.html#ref-edmans2020employee" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-griffin2019national" role="doc-biblioref">D. W. Griffin et al.</a> (<a href="Bib.html#ref-griffin2019national" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-amon2019passive" role="doc-biblioref">Amon, Rammerstorfer, and Weinmayer</a> (<a href="Bib.html#ref-amon2019passive" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-murata2021esg" role="doc-biblioref">Murata and Hamori</a> (<a href="Bib.html#ref-murata2021esg" role="doc-biblioref">2021</a>)</span>. and <span class="citation"><a href="Bib.html#ref-giese2021foundations" role="doc-biblioref">Giese, Nagy, and Rauis</a> (<a href="Bib.html#ref-giese2021foundations" role="doc-biblioref">2021</a>)</span> In addition, in <span class="citation"><a href="Bib.html#ref-chakrabarti2020time" role="doc-biblioref">Chakrabarti and Sen</a> (<a href="Bib.html#ref-chakrabarti2020time" role="doc-biblioref">2020</a>)</span>, it is shown that global indices outperform the market, but regional indices do not. In some papers, however, it is shown that there is no link between sustainability and performance, no matter the geographical zone (see <span class="citation"><a href="Bib.html#ref-auer2016socially2" role="doc-biblioref">Auer and Schuhmacher</a> (<a href="Bib.html#ref-auer2016socially2" role="doc-biblioref">2016</a>)</span>). Lastly, there are so many country-specific studies that it is impossible to cite them exhaustively.</p>
<p>Another facet of geography pertains to where ESG incidents may happen and <span class="citation"><a href="Bib.html#ref-groen2021esg" role="doc-biblioref">Groen-Xu and Zeume</a> (<a href="Bib.html#ref-groen2021esg" role="doc-biblioref">2021</a>)</span> show that when they occur abroad. Shocks to incidents are negative overall (-0.6%), but less so when they occur outside the country of headquarters (-0.4%). The authors conclude that their is a home bias in shareholder preferences for ESG externalities.</p>
</div>
<div id="time-varying-link" class="section level3" number="4.5.3">
<h3>
<span class="header-section-number">4.5.3</span> Time-varying link<a class="anchor" aria-label="anchor" href="#time-varying-link"><i class="fas fa-link"></i></a>
</h3>
<p>The third important factor is <strong>chronology</strong>. Just like all factor tilts in the world, the ESG <em>style</em> has its moments. The perception of ESG has shifted through time. <span class="citation"><a href="Bib.html#ref-ioannou2015impact" role="doc-biblioref">Ioannou and Serafeim</a> (<a href="Bib.html#ref-ioannou2015impact" role="doc-biblioref">2015</a>)</span> find that analysts’ recommendations <em>used to be</em> pessimistic toward corporate SR policies. Political changes also impact expectations toward green firms and generate shocks to returns (<span class="citation"><a href="Bib.html#ref-ramelli2021investor" role="doc-biblioref">Ramelli et al.</a> (<a href="Bib.html#ref-ramelli2021investor" role="doc-biblioref">2021</a>)</span>). We refer to <span class="citation"><a href="Bib.html#ref-bauer2005international" role="doc-biblioref">Bauer, Koedijk, and Otten</a> (<a href="Bib.html#ref-bauer2005international" role="doc-biblioref">2005</a>)</span>, <span class="citation"><a href="Bib.html#ref-barnett2006beyond" role="doc-biblioref">M. L. Barnett and Salomon</a> (<a href="Bib.html#ref-barnett2006beyond" role="doc-biblioref">2006</a>)</span>, <span class="citation"><a href="Bib.html#ref-statman2006socially" role="doc-biblioref">Statman</a> (<a href="Bib.html#ref-statman2006socially" role="doc-biblioref">2006</a>)</span>, <span class="citation"><a href="Bib.html#ref-climent2011green" role="doc-biblioref">Climent and Soriano</a> (<a href="Bib.html#ref-climent2011green" role="doc-biblioref">2011</a>)</span>, <span class="citation"><a href="Bib.html#ref-bebchuk2013learning" role="doc-biblioref">L. A. Bebchuk, Cohen, and Wang</a> (<a href="Bib.html#ref-bebchuk2013learning" role="doc-biblioref">2013</a>)</span>, <span class="citation"><a href="Bib.html#ref-de2015benefits" role="doc-biblioref">De and Clayman</a> (<a href="Bib.html#ref-de2015benefits" role="doc-biblioref">2015</a>)</span>, <span class="citation"><a href="Bib.html#ref-belghitar2017importance" role="doc-biblioref">Belghitar, Clark, and Deshmukh</a> (<a href="Bib.html#ref-belghitar2017importance" role="doc-biblioref">2017</a>)</span>, <span class="citation"><a href="Bib.html#ref-ibikunle2017european" role="doc-biblioref">Ibikunle and Steffen</a> (<a href="Bib.html#ref-ibikunle2017european" role="doc-biblioref">2017</a>)</span>, <span class="citation"><a href="Bib.html#ref-in2017being" role="doc-biblioref">In, Park, and Monk</a> (<a href="Bib.html#ref-in2017being" role="doc-biblioref">2017</a>)</span>, <span class="citation"><a href="Bib.html#ref-bennani2018esg" role="doc-biblioref">Bennani et al.</a> (<a href="Bib.html#ref-bennani2018esg" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-mutlu2018corporate" role="doc-biblioref">Mutlu et al.</a> (<a href="Bib.html#ref-mutlu2018corporate" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-ciciretti2019contributions" role="doc-biblioref">Ciciretti, Dalò, and Dam</a> (<a href="Bib.html#ref-ciciretti2019contributions" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-li2019learning" role="doc-biblioref">Z. Li et al.</a> (<a href="Bib.html#ref-li2019learning" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-vojtko2019quant" role="doc-biblioref">Vojtko and Padysak</a> (<a href="Bib.html#ref-vojtko2019quant" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-kanuri2020risk" role="doc-biblioref">Kanuri</a> (<a href="Bib.html#ref-kanuri2020risk" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-bansal2018socially" role="doc-biblioref">Bansal, Wu, and Yaron</a> (<a href="Bib.html#ref-bansal2018socially" role="doc-biblioref">2022</a>)</span>, <span class="citation"><a href="Bib.html#ref-faccini2021climate" role="doc-biblioref">Faccini, Matin, and Skiadopoulos</a> (<a href="Bib.html#ref-faccini2021climate" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-lepetit2021recent" role="doc-biblioref">Lepetit et al.</a> (<a href="Bib.html#ref-lepetit2021recent" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-tsai2021changes" role="doc-biblioref">Tsai and Wu</a> (<a href="Bib.html#ref-tsai2021changes" role="doc-biblioref">2021</a>)</span> for contributions that document the time-varying nature of sustainable investments. <span class="citation"><a href="Bib.html#ref-barnett2020climate" role="doc-biblioref">M. Barnett</a> (<a href="Bib.html#ref-barnett2020climate" role="doc-biblioref">2017</a>)</span> reports that the price of climate risk (evaluated via temperatures) is more significant after 1996, compared to the full sample period of 1948-2019. According to <span class="citation"><a href="Bib.html#ref-alda2021socially" role="doc-biblioref">Alda and Vicente</a> (<a href="Bib.html#ref-alda2021socially" role="doc-biblioref">2021</a>)</span>, returns of SRI funds do not exhibit seasonal patterns.</p>
<p>Several of the aforementioned studies underline that the profitability of SRI was highest in the most recent periods (see <span class="citation"><a href="Bib.html#ref-dumitrescu2021hidden" role="doc-biblioref">Dumitrescu, Jarvinen, and Zakriya</a> (<a href="Bib.html#ref-dumitrescu2021hidden" role="doc-biblioref">2021</a>)</span> among many others). The hype surrounding ESG investing (and the related demand) may have pushed the prices up. For instance, it is shown in <span class="citation"><a href="Bib.html#ref-azar2021big" role="doc-biblioref">Azar et al.</a> (<a href="Bib.html#ref-azar2021big" role="doc-biblioref">2021</a>)</span> that the commitment of global asset managers with respect to sustainability is relatively novel and significant. Relatedly, <span class="citation"><a href="Bib.html#ref-bansal2018socially" role="doc-biblioref">Bansal, Wu, and Yaron</a> (<a href="Bib.html#ref-bansal2018socially" role="doc-biblioref">2022</a>)</span> find that SRI performs well in good times, but badly during economic downturns and crises; <span class="citation"><a href="Bib.html#ref-nofsinger2014socially" role="doc-biblioref">J. Nofsinger and Varma</a> (<a href="Bib.html#ref-nofsinger2014socially" role="doc-biblioref">2014</a>)</span> come to the opposite conclusion.</p>
<p><span class="citation"><a href="Bib.html#ref-lioui2021chasing" role="doc-biblioref">Lioui and Tarelli</a> (<a href="Bib.html#ref-lioui2021chasing" role="doc-biblioref">2021</a>)</span> disentangle time-series and cross-sectional ESG factors. They document a significant time variation in the alpha of both types of factors, conditional on firm characteristics.</p>
<p>The contribution of <span class="citation"><a href="Bib.html#ref-pastor2021dissecting" role="doc-biblioref">Pastor, Stambaugh, and Taylor</a> (<a href="Bib.html#ref-pastor2021dissecting" role="doc-biblioref">2021a</a>)</span> points out that the recent outperformance of sustainable firms is entirely due to aggregate concern over the environment. The authors use the Media Climate Change Concerns (MCCC) index of <span class="citation"><a href="Bib.html#ref-ardia2020climate" role="doc-biblioref">Ardia et al.</a> (<a href="Bib.html#ref-ardia2020climate" role="doc-biblioref">2020</a>)</span> and prove that the long-short green factor sees its performance vanish once it has been controlled for shocks in the MCCC index.</p>
<p>We end this subsection by important articles that show that the good performance of ESG stocks in 2019-2021 can be attributed to increased flows. This is the central hypothesis in <span class="citation"><a href="Bib.html#ref-van2021flow" role="doc-biblioref">Beck</a> (<a href="Bib.html#ref-van2021flow" role="doc-biblioref">2021</a>)</span>. In <span class="citation"><a href="Bib.html#ref-avramov2021dynamic" role="doc-biblioref">Avramov, Lioui, et al.</a> (<a href="Bib.html#ref-avramov2021dynamic" role="doc-biblioref">2021</a>)</span>, the authors argue that the theoretically negative green premium can be offset if the demand for greenness is high. Preference shocks appeared to have pushed realized returns up recently for green assets.</p>
</div>
<div id="industry" class="section level3" number="4.5.4">
<h3>
<span class="header-section-number">4.5.4</span> Industry<a class="anchor" aria-label="anchor" href="#industry"><i class="fas fa-link"></i></a>
</h3>
<p>A fourth dimension is <strong>industry</strong>. For instance, <span class="citation"><a href="Bib.html#ref-herremans1993investigation" role="doc-biblioref">Herremans, Akathaporn, and McInnes</a> (<a href="Bib.html#ref-herremans1993investigation" role="doc-biblioref">1993</a>)</span>, <span class="citation"><a href="Bib.html#ref-russo1997resource" role="doc-biblioref">Russo and Fouts</a> (<a href="Bib.html#ref-russo1997resource" role="doc-biblioref">1997</a>)</span>, <span class="citation"><a href="Bib.html#ref-semenova2008financial" role="doc-biblioref">Semenova and Hassel</a> (<a href="Bib.html#ref-semenova2008financial" role="doc-biblioref">2008</a>)</span>, <span class="citation"><a href="Bib.html#ref-hoepner2010corporate" role="doc-biblioref">Hoepner and Yu</a> (<a href="Bib.html#ref-hoepner2010corporate" role="doc-biblioref">2017a</a>)</span>, <span class="citation"><a href="Bib.html#ref-giroud2011corporate" role="doc-biblioref">Giroud and Mueller</a> (<a href="Bib.html#ref-giroud2011corporate" role="doc-biblioref">2011</a>)</span>, <span class="citation"><a href="Bib.html#ref-lee2011relationship" role="doc-biblioref">Jooh Lee, Pati, and Roh</a> (<a href="Bib.html#ref-lee2011relationship" role="doc-biblioref">2011</a>)</span> (oil and gas), <span class="citation"><a href="Bib.html#ref-jo2012does" role="doc-biblioref">Jo and Na</a> (<a href="Bib.html#ref-jo2012does" role="doc-biblioref">2012</a>)</span>, <span class="citation"><a href="Bib.html#ref-auer2016socially2" role="doc-biblioref">Auer and Schuhmacher</a> (<a href="Bib.html#ref-auer2016socially2" role="doc-biblioref">2016</a>)</span>, <span class="citation"><a href="Bib.html#ref-de2016corporate" role="doc-biblioref">De Haan and Vlahu</a> (<a href="Bib.html#ref-de2016corporate" role="doc-biblioref">2016</a>)</span>, <span class="citation"><a href="Bib.html#ref-feng2017corporate" role="doc-biblioref">M. Feng, Wang, and Kreuze</a> (<a href="Bib.html#ref-feng2017corporate" role="doc-biblioref">2017</a>)</span>, <span class="citation"><a href="Bib.html#ref-bertolotti2019carbon" role="doc-biblioref">Bertolotti and Kent</a> (<a href="Bib.html#ref-bertolotti2019carbon" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-alessandrini2020esg" role="doc-biblioref">Alessandrini and Jondeau</a> (<a href="Bib.html#ref-alessandrini2020esg" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-torre2020does" role="doc-biblioref">Torre et al.</a> (<a href="Bib.html#ref-torre2020does" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-abdi2021exploring" role="doc-biblioref">Abdi, Li, and Càmara-Turull</a> (<a href="Bib.html#ref-abdi2021exploring" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-kuo2021corporate" role="doc-biblioref">Kuo, Chen, and Meng</a> (<a href="Bib.html#ref-kuo2021corporate" role="doc-biblioref">2021</a>)</span> (airlines), <span class="citation"><a href="Bib.html#ref-giese2021foundations" role="doc-biblioref">Giese, Nagy, and Rauis</a> (<a href="Bib.html#ref-giese2021foundations" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-okafor2021corporate" role="doc-biblioref">Okafor, Adusei, and Adeleye</a> (<a href="Bib.html#ref-okafor2021corporate" role="doc-biblioref">2021</a>)</span> find that the impact of ESG depends on the sector of firms. In contrast, <span class="citation"><a href="Bib.html#ref-statman2009wages" role="doc-biblioref">Statman and Glushkov</a> (<a href="Bib.html#ref-statman2009wages" role="doc-biblioref">2009</a>)</span> argues that for SRI to reach its full potential, it must only rely on ESG ratings and not on industry screening. <span class="citation"><a href="Bib.html#ref-giese2021deconstructing" role="doc-biblioref">Giese, Nagy, and Lee</a> (<a href="Bib.html#ref-giese2021deconstructing" role="doc-biblioref">2021</a>)</span> argue that the weighting of ESG pillars should be sector-specific. Typically, industries are not hit uniformly by rising temperature (see <span class="citation"><a href="Bib.html#ref-shaw2021evaluating" role="doc-biblioref">Shaw, Evans, and Turner</a> (<a href="Bib.html#ref-shaw2021evaluating" role="doc-biblioref">2021</a>)</span>). For an analysis on the energy sector, we refer to <span class="citation"><a href="Bib.html#ref-brzeszczynski2019socially" role="doc-biblioref">Brzeszczynski et al.</a> (<a href="Bib.html#ref-brzeszczynski2019socially" role="doc-biblioref">2019</a>)</span>.</p>
</div>
<div id="ownership" class="section level3" number="4.5.5">
<h3>
<span class="header-section-number">4.5.5</span> Ownership<a class="anchor" aria-label="anchor" href="#ownership"><i class="fas fa-link"></i></a>
</h3>
<p>Finally, a fifth mitigating effect is firm <strong>ownership</strong>. <span class="citation"><a href="Bib.html#ref-nekhili2017corporate" role="doc-biblioref">Nekhili et al.</a> (<a href="Bib.html#ref-nekhili2017corporate" role="doc-biblioref">2017</a>)</span> and <span class="citation"><a href="Bib.html#ref-abeysekera2020corporate" role="doc-biblioref">Abeysekera and Fernando</a> (<a href="Bib.html#ref-abeysekera2020corporate" role="doc-biblioref">2020</a>)</span> indicate that family involvement and ownership is also likely to impact the relationship between CSR and financial performance. <span class="citation"><a href="Bib.html#ref-brogger2020skills" role="doc-biblioref">Brøgger and Kronies</a> (<a href="Bib.html#ref-brogger2020skills" role="doc-biblioref">2020</a>)</span> find that the positivity (and significance) of the ESG factor exists within firms that are owned by unconstrained investors (e.g., mutual and hedge funds). Relatedly, <span class="citation"><a href="Bib.html#ref-cheema2019decarbonization" role="doc-biblioref">Cheema-Fox et al.</a> (<a href="Bib.html#ref-cheema2019decarbonization" role="doc-biblioref">2019</a>)</span> find that the performance of portfolios is strongly linked to institutional investor flows.<a class="footnote-ref" tabindex="0" data-toggle="popover" data-content='<p>Institutional ownership affects the propensity of firms to indulge in ESG policies, see, e.g., <span class="citation"><a href="Bib.html#ref-martinez2021nonlinear" role="doc-biblioref">Martı́nez-Ferrero and Lozano</a> (<a href="Bib.html#ref-martinez2021nonlinear" role="doc-biblioref">2021</a>)</span> in the case of developing countries.</p>'><sup>29</sup></a>
\end{itemize} </p>
</div>
<div id="nonlinearities" class="section level3" number="4.5.6">
<h3>
<span class="header-section-number">4.5.6</span> Nonlinearities<a class="anchor" aria-label="anchor" href="#nonlinearities"><i class="fas fa-link"></i></a>
</h3>
<p>Some researchers (e.g., <span class="citation"><a href="Bib.html#ref-barnett2006beyond" role="doc-biblioref">M. L. Barnett and Salomon</a> (<a href="Bib.html#ref-barnett2006beyond" role="doc-biblioref">2006</a>)</span>, <span class="citation"><a href="Bib.html#ref-brammer2008does" role="doc-biblioref">Brammer and Millington</a> (<a href="Bib.html#ref-brammer2008does" role="doc-biblioref">2008</a>)</span>, <span class="citation"><a href="Bib.html#ref-fernando2010does" role="doc-biblioref">C. Fernando, Sharfman, and Uysal</a> (<a href="Bib.html#ref-fernando2010does" role="doc-biblioref">2010</a>)</span>, <span class="citation"><a href="Bib.html#ref-harjoto2017institutional" role="doc-biblioref">Harjoto, Jo, and Kim</a> (<a href="Bib.html#ref-harjoto2017institutional" role="doc-biblioref">2017</a>)</span>, <span class="citation"><a href="Bib.html#ref-gerged2020mandatory" role="doc-biblioref">Gerged, Matthews, and Elheddad</a> (<a href="Bib.html#ref-gerged2020mandatory" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-bruna2022investigating" role="doc-biblioref">Bruna et al.</a> (<a href="Bib.html#ref-bruna2022investigating" role="doc-biblioref">2022</a>)</span>) manage to reconcile seemingly contradicting results by showing that the relationship between CSR and performance (or risk, in <span class="citation"><a href="Bib.html#ref-korinth2022corporate" role="doc-biblioref">Korinth and Lueg</a> (<a href="Bib.html#ref-korinth2022corporate" role="doc-biblioref">2022</a>)</span>) is not linear: very good and very bad corporations experience abnormal returns, while those in the bulk of the ESG distribution perform differently. <span class="citation"><a href="Bib.html#ref-xie2019environmental" role="doc-biblioref">Xie et al.</a> (<a href="Bib.html#ref-xie2019environmental" role="doc-biblioref">2019</a>)</span> show that the optimal level of disclosure is in the middle of the distribution and very low or very high disclosure leads to lower performance. <span class="citation"><a href="Bib.html#ref-de2021value" role="doc-biblioref">Fuente, Ortiz, and Velasco</a> (<a href="Bib.html#ref-de2021value" role="doc-biblioref">2021</a>)</span> also document an inverted U-shape. <span class="citation"><a href="Bib.html#ref-vu2022esg" role="doc-biblioref">Vu, Lehkonen, and Junttila</a> (<a href="Bib.html#ref-vu2022esg" role="doc-biblioref">2022</a>)</span> report a U-shape, but for emerging markets.<br>
More generally, <span class="citation"><a href="Bib.html#ref-wong2021stock" role="doc-biblioref">Wong and Zhang</a> (<a href="Bib.html#ref-wong2021stock" role="doc-biblioref">2021</a>)</span> reveal that the nature of shocks to negative ESG news depend on firm characteristics such as firm size, stock liquidity, S&P500 constituency, and corporate reputation status.</p>
<p>
The screening intensity and the type of screens seems to play a role as well (<span class="citation"><a href="Bib.html#ref-capelle2014performance" role="doc-biblioref">Capelle-Blancard and Monjon</a> (<a href="Bib.html#ref-capelle2014performance" role="doc-biblioref">2014</a>)</span>), if only because it impacts the diversification of the portfolio (<span class="citation"><a href="Bib.html#ref-jin2020esg" role="doc-biblioref">Jin</a> (<a href="Bib.html#ref-jin2020esg" role="doc-biblioref">2020</a>)</span>). <span class="citation"><a href="Bib.html#ref-fairhurst2020too" role="doc-biblioref">Fairhurst and Greene</a> (<a href="Bib.html#ref-fairhurst2020too" role="doc-biblioref">2020</a>)</span> also document a non-monotonic impact of ESG scores: extreme CSR policies appear to be harmful, at least on the takeover market. Non-linear patterns are moreover documented in <span class="citation"><a href="Bib.html#ref-huang2018zombie" role="doc-biblioref">S. Huang and Hilary</a> (<a href="Bib.html#ref-huang2018zombie" role="doc-biblioref">2018</a>)</span> for governance proxies. Asymmetric preferences of investors (who are indifferent to best-in-class, but penalize worst-in-class firms with negative E and G scores) are revealed in <span class="citation"><a href="Bib.html#ref-nofsinger2019institutional" role="doc-biblioref">J. R. Nofsinger, Sulaeman, and Varma</a> (<a href="Bib.html#ref-nofsinger2019institutional" role="doc-biblioref">2019</a>)</span>. One other route to explain diverging results in the field is to argue that models are misspecified, e.g., when important independent variables are omitted (see <span class="citation"><a href="Bib.html#ref-mcwilliams2000corporate" role="doc-biblioref">McWilliams and Siegel</a> (<a href="Bib.html#ref-mcwilliams2000corporate" role="doc-biblioref">2000</a>)</span>). The way ESG criteria are integrated in the portfolio design can also matter. In their study on the Australian market, <span class="citation"><a href="Bib.html#ref-fan2020sustainable" role="doc-biblioref">Fan and Michalski</a> (<a href="Bib.html#ref-fan2020sustainable" role="doc-biblioref">2020</a>)</span> find that simple ESG sorts have disappointing performance, but combined with other factors, like quality or momentum, boosts their returns.</p>
<p>The way and reason why firms disclose CSR actions is also likely to matter. <span class="citation"><a href="Bib.html#ref-bams2021heterogeneous" role="doc-biblioref">Bams, Kroft, and Maas</a> (<a href="Bib.html#ref-bams2021heterogeneous" role="doc-biblioref">2021</a>)</span> separate three dimensions in CSR performance and disclosure. First, strategic CSR refers to genuine sustainability for the sake of sustainability. Second, CSR as insurance is a more passive approach to sustainability through which boards and firms “<em>conform to the institutional pressure for CSR by providing a minimal level of CSR to mitigate risks and maintain their licence to operate</em>.” Finally, there is greenwashing. <span class="citation"><a href="Bib.html#ref-bams2021heterogeneous" role="doc-biblioref">Bams, Kroft, and Maas</a> (<a href="Bib.html#ref-bams2021heterogeneous" role="doc-biblioref">2021</a>)</span> find that firms which choose the former outperform the others in both realised social and financial dimensions. </p>
<p>We end this subsection with five references. <span class="citation"><a href="Bib.html#ref-barnett2006beyond" role="doc-biblioref">M. L. Barnett and Salomon</a> (<a href="Bib.html#ref-barnett2006beyond" role="doc-biblioref">2006</a>)</span> contend that risk-adjusted returns of ESG strategies depend non-linearly in screening intensities. <span class="citation"><a href="Bib.html#ref-lopez2018evaluation" role="doc-biblioref">López-Arceiz, Bellostas-Pérezgrueso, and Moneva</a> (<a href="Bib.html#ref-lopez2018evaluation" role="doc-biblioref">2018</a>)</span> argue that the profitability of SR funds is strongly impacted by the cultural environment in which the fund operates. <span class="citation"><a href="Bib.html#ref-dorfleitner2020esg" role="doc-biblioref">Dorfleitner, Kreuzer, and Sparrer</a> (<a href="Bib.html#ref-dorfleitner2020esg" role="doc-biblioref">2020</a>)</span> find that the weighting scheme of portfolios can matter, as well as the size of companies. <span class="citation"><a href="Bib.html#ref-assael2021esg" role="doc-biblioref">Assael and Challet</a> (<a href="Bib.html#ref-assael2021esg" role="doc-biblioref">2021</a>)</span> show that ESG attribute help explain, via boosted tree algorithms, the sign of firms’ CAPM alpha. In addition, higher ESG scores are generally associated with larger returns for small firms, but smaller ones for large companies. Finally, <span class="citation"><a href="Bib.html#ref-ardia2020climate" role="doc-biblioref">Ardia et al.</a> (<a href="Bib.html#ref-ardia2020climate" role="doc-biblioref">2020</a>)</span> (following <span class="citation"><a href="Bib.html#ref-pastor2020sustainable" role="doc-biblioref">Pastor, Stambaugh, and Taylor</a> (<a href="Bib.html#ref-pastor2020sustainable" role="doc-biblioref">2021b</a>)</span>) find that the profitability of green minus brown portfolios depends on the aggregate concern with respect to climate threats.</p>
</div>
</div>
<div id="csr-and-risk" class="section level2" number="4.6">
<h2>
<span class="header-section-number">4.6</span> CSR and risk<a class="anchor" aria-label="anchor" href="#csr-and-risk"><i class="fas fa-link"></i></a>
</h2>
<p>
The question of whether SRI is an efficient way to hedge risk remains open. This has critical implications for institutions such as pension funds (<span class="citation"><a href="Bib.html#ref-sautner2021esg" role="doc-biblioref">Sautner and Starks</a> (<a href="Bib.html#ref-sautner2021esg" role="doc-biblioref">2021</a>)</span>) in the case of downside risk. <span class="citation"><a href="Bib.html#ref-becchetti2015socially" role="doc-biblioref">Becchetti et al.</a> (<a href="Bib.html#ref-becchetti2015socially" role="doc-biblioref">2015</a>)</span> find that SR funds performed less badly during the 2007-2008 financial crisis, compared to conventional funds. <span class="citation"><a href="Bib.html#ref-de2015benefits" role="doc-biblioref">De and Clayman</a> (<a href="Bib.html#ref-de2015benefits" role="doc-biblioref">2015</a>)</span> and <span class="citation"><a href="Bib.html#ref-hoepner2020esg" role="doc-biblioref">Hoepner et al.</a> (<a href="Bib.html#ref-hoepner2020esg" role="doc-biblioref">2021</a>)</span> document a negative relationship between ESG ratings and risk (measured by stock volatility and downside variance). <span class="citation"><a href="Bib.html#ref-dunbar2021corporate" role="doc-biblioref">C. G. Dunbar, Li, and Shi</a> (<a href="Bib.html#ref-dunbar2021corporate" role="doc-biblioref">2021</a>)</span> find that CSR is only a vector of risk reduction if the governance of firms seeks transparency and corporate social performance. In their study on European funds, <span class="citation"><a href="Bib.html#ref-gonccalves2021risk" role="doc-biblioref">Gonçalves, Pimentel, and Gaio</a> (<a href="Bib.html#ref-gonccalves2021risk" role="doc-biblioref">2021</a>)</span> reveal that green funds outperform conventional funds in times of crises.</p>
<p>Finally, two rather contrarian articles, <span class="citation"><a href="Bib.html#ref-brav2021brown" role="doc-biblioref">Brav and Heaton</a> (<a href="Bib.html#ref-brav2021brown" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-heaton2021expected" role="doc-biblioref">Heaton</a> (<a href="Bib.html#ref-heaton2021expected" role="doc-biblioref">2021</a>)</span> suggest that “prudent” investors might want to invest in brown assets in order to hedge against the likelihood that the transition to a greener economy fails to materialize. They highlight the probabilities of scenarii in which brown assets would outperform their green counterparts.</p>
<p>
The COVID-19 pandemic yielded a thread of event studies that aim at understanding if sustainability mitigated risk. We list a few such contributions below. <span class="citation"><a href="Bib.html#ref-singh2020covid" role="doc-biblioref">Singh</a> (<a href="Bib.html#ref-singh2020covid" role="doc-biblioref">2020</a>)</span> finds that the ESG factor performed well during the COVID-19 market crash (see <span class="citation"><a href="Bib.html#ref-mahmoud2021social" role="doc-biblioref">Mahmoud and Meyer</a> (<a href="Bib.html#ref-mahmoud2021social" role="doc-biblioref">2021</a>)</span> for an in-depth analysis of the drivers of ESG preferences posterior to this drawdown). Similarly, <span class="citation"><a href="Bib.html#ref-pastor2020mutual" role="doc-biblioref">Pástor and Vorsatz</a> (<a href="Bib.html#ref-pastor2020mutual" role="doc-biblioref">2020</a>)</span> and <span class="citation"><a href="Bib.html#ref-omura2020does" role="doc-biblioref">Omura, Roca, and Nakai</a> (<a href="Bib.html#ref-omura2020does" role="doc-biblioref">2020</a>)</span> report that sustainable funds outperformed conventional ones during the market meltdown. <span class="citation"><a href="Bib.html#ref-akhtar2021board" role="doc-biblioref">Akhtar, Veeraraghavan, and Zolotoy</a> (<a href="Bib.html#ref-akhtar2021board" role="doc-biblioref">2021</a>)</span> document a positive effect of gender diversity on the abnormal performance of US stocks during the period between March and April 2020. In <span class="citation"><a href="Bib.html#ref-xiong2021impact" role="doc-biblioref">Xiong</a> (<a href="Bib.html#ref-xiong2021impact" role="doc-biblioref">2021</a>)</span>, firms with low ESG risk are found to be outperforming those with high ESG risk.
Likewise, in their studes on Indian stocks, <span class="citation"><a href="Bib.html#ref-arora2021does" role="doc-biblioref">Arora, Sur, and Chauhan</a> (<a href="Bib.html#ref-arora2021does" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-beloskar2022did" role="doc-biblioref">Beloskar and Rao</a> (<a href="Bib.html#ref-beloskar2022did" role="doc-biblioref">2022</a>)</span> reveal that sustainable firms fared better than brown ones during the pandemic. Similar results are obtained on European data in <span class="citation"><a href="Bib.html#ref-pizzutilo2021esg" role="doc-biblioref">Pizzutilo</a> (<a href="Bib.html#ref-pizzutilo2021esg" role="doc-biblioref">2021</a>)</span>.</p>
<p>However, <span class="citation"><a href="Bib.html#ref-folger2020esg" role="doc-biblioref">Folger-Laronde et al.</a> (<a href="Bib.html#ref-folger2020esg" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-demers2020esg" role="doc-biblioref">Demers et al.</a> (<a href="Bib.html#ref-demers2020esg" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-mahmoud2020anatomy" role="doc-biblioref">Mahmoud and Meyer</a> (<a href="Bib.html#ref-mahmoud2020anatomy" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-capelle2021socially" role="doc-biblioref">Capelle-Blancard, Desroziers, and Zerbib</a> (<a href="Bib.html#ref-capelle2021socially" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-chiappini2021impact" role="doc-biblioref">Chiappini, Vento, and De Palma</a> (<a href="Bib.html#ref-chiappini2021impact" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-gianfrate2021resilience" role="doc-biblioref">Gianfranco Gianfrate, Kievid, and Dijk</a> (<a href="Bib.html#ref-gianfrate2021resilience" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-pavlova2021esg" role="doc-biblioref">Pavlova and Boyrie</a> (<a href="Bib.html#ref-pavlova2021esg" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-yousaf2021green" role="doc-biblioref">Yousaf, Suleman, and Demirer</a> (<a href="Bib.html#ref-yousaf2021green" role="doc-biblioref">2021</a>)</span> see little or no hedging power of ESG-driven funds during the pandemic.<a class="footnote-ref" tabindex="0" data-toggle="popover" data-content='<p><span class="citation"><a href="Bib.html#ref-yousaf2021green" role="doc-biblioref">Yousaf, Suleman, and Demirer</a> (<a href="Bib.html#ref-yousaf2021green" role="doc-biblioref">2021</a>)</span> find that green <em>bonds</em> are however useful.</p>'><sup>30</sup></a> <span class="citation"><a href="Bib.html#ref-glossner2020institutional" role="doc-biblioref">Glossner et al.</a> (<a href="Bib.html#ref-glossner2020institutional" role="doc-biblioref">2021</a>)</span> find no evidence that investors shifted toward ESG firms during the COVID-19 crisis and <span class="citation"><a href="Bib.html#ref-singh2021covid" role="doc-biblioref">Singh</a> (<a href="Bib.html#ref-singh2021covid" role="doc-biblioref">2021</a>)</span> finds evidence that investors shifted from ESG <em>equities</em> to ESG <em>bonds</em>. Finally, ESG returns may also be contingent on investor sentiment (<span class="citation"><a href="Bib.html#ref-azevedo2020investor" role="doc-biblioref">Azevedo, Kaserer, and MS Campos</a> (<a href="Bib.html#ref-azevedo2020investor" role="doc-biblioref">2021</a>)</span>).
</p>
<p>Below, we list further contributions that conclude that ESG is positively, negatively, or weakly linked to risk.</p>
<ul>
<li>
<strong>ESG reduces risk</strong>. CSR engagement reduces risk in controversial industries (<span class="citation"><a href="Bib.html#ref-jo2012does" role="doc-biblioref">Jo and Na</a> (<a href="Bib.html#ref-jo2012does" role="doc-biblioref">2012</a>)</span>). Social irresponsibility is linked to higher risks (<span class="citation"><a href="Bib.html#ref-oikonomou2012impact" role="doc-biblioref">Oikonomou, Brooks, and Pavelin</a> (<a href="Bib.html#ref-oikonomou2012impact" role="doc-biblioref">2012</a>)</span>). Board quality reduces many types of risk (<span class="citation"><a href="Bib.html#ref-kaiser2020board" role="doc-biblioref">Kaiser</a> (<a href="Bib.html#ref-kaiser2020board" role="doc-biblioref">2020a</a>)</span>). Hedge funds can benefit from risk-mitigating properties of ESG investments (<span class="citation"><a href="Bib.html#ref-duanmu2019can" role="doc-biblioref">Duanmu et al.</a> (<a href="Bib.html#ref-duanmu2019can" role="doc-biblioref">2021</a>)</span>). A few articles favorably compare traditional indices to their green counterparts during downturns (<span class="citation"><a href="Bib.html#ref-becchetti2015socially" role="doc-biblioref">Becchetti et al.</a> (<a href="Bib.html#ref-becchetti2015socially" role="doc-biblioref">2015</a>)</span>, <span class="citation"><a href="Bib.html#ref-gonccalves2021risk" role="doc-biblioref">Gonçalves, Pimentel, and Gaio</a> (<a href="Bib.html#ref-gonccalves2021risk" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-ouchen2021esg" role="doc-biblioref">Ouchen</a> (<a href="Bib.html#ref-ouchen2021esg" role="doc-biblioref">2021</a>)</span>). Other contributions include: <span class="citation"><a href="Bib.html#ref-mcguire1988corporate" role="doc-biblioref">McGuire, Sundgren, and Schneeweis</a> (<a href="Bib.html#ref-mcguire1988corporate" role="doc-biblioref">1988</a>)</span>, <span class="citation"><a href="Bib.html#ref-el2011does" role="doc-biblioref">El Ghoul et al.</a> (<a href="Bib.html#ref-el2011does" role="doc-biblioref">2011</a>)</span>, <span class="citation"><a href="Bib.html#ref-humphrey2012does" role="doc-biblioref">J. E. Humphrey, Lee, and Shen</a> (<a href="Bib.html#ref-humphrey2012does" role="doc-biblioref">2012</a>)</span>, <span class="citation"><a href="Bib.html#ref-kong2020can" role="doc-biblioref">Kong et al.</a> (<a href="Bib.html#ref-kong2020can" role="doc-biblioref">2020</a>)</span> and <span class="citation"><a href="Bib.html#ref-he2022csr" role="doc-biblioref">F. He et al.</a> (<a href="Bib.html#ref-he2022csr" role="doc-biblioref">2022</a>)</span> (idiosyncratic risk), <span class="citation"><a href="Bib.html#ref-harjoto2017institutional" role="doc-biblioref">Harjoto, Jo, and Kim</a> (<a href="Bib.html#ref-harjoto2017institutional" role="doc-biblioref">2017</a>)</span>, <span class="citation"><a href="Bib.html#ref-bernile2018board" role="doc-biblioref">Bernile, Bhagwat, and Yonker</a> (<a href="Bib.html#ref-bernile2018board" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-monti2019does" role="doc-biblioref">Monti et al.</a> (<a href="Bib.html#ref-monti2019does" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-gibson2019responsible" role="doc-biblioref">Gibson et al.</a> (<a href="Bib.html#ref-gibson2019responsible" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-lopez2020esg" role="doc-biblioref">Lopez de Silanes, McCahery, and Pudschedl</a> (<a href="Bib.html#ref-lopez2020esg" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-karwowski2021application" role="doc-biblioref">Karwowski and Raulinajtys-Grzybek</a> (<a href="Bib.html#ref-karwowski2021application" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-olofsson2021ethical" role="doc-biblioref">Olofsson et al.</a> (<a href="Bib.html#ref-olofsson2021ethical" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-brodmann2021liquidity" role="doc-biblioref">Brodmann, Wuthisatian, and Malladi</a> (<a href="Bib.html#ref-brodmann2021liquidity" role="doc-biblioref">2021</a>)</span>. ESG also seems negatively related to <em>tail</em> risk (<span class="citation"><a href="Bib.html#ref-bax2021esg" role="doc-biblioref">Bax et al.</a> (<a href="Bib.html#ref-bax2021esg" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-xiong2021impact" role="doc-biblioref">Xiong</a> (<a href="Bib.html#ref-xiong2021impact" role="doc-biblioref">2021</a>)</span>), to stock crash risk (<span class="citation"><a href="Bib.html#ref-wu2021stabilizer" role="doc-biblioref">G. Wu and You</a> (<a href="Bib.html#ref-wu2021stabilizer" role="doc-biblioref">2021</a>)</span> find negative correlation with the quality of green patents, but not their quantity; <span class="citation"><a href="Bib.html#ref-bae2021esg" role="doc-biblioref">JinCheol Bae, Yang, and Kim</a> (<a href="Bib.html#ref-bae2021esg" role="doc-biblioref">2021</a>)</span> report a negative overall relationship with ESG); <span class="citation"><a href="Bib.html#ref-feng2021esg" role="doc-biblioref">J. Feng, Goodell, and Shen</a> (<a href="Bib.html#ref-feng2021esg" role="doc-biblioref">2021</a>)</span> analyze the Chinese market, to default risk (<span class="citation"><a href="Bib.html#ref-aslan2021sustainable" role="doc-biblioref">Aslan, Poppe, and Posch</a> (<a href="Bib.html#ref-aslan2021sustainable" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-li2022esg" role="doc-biblioref">H. Li, Zhang, and Zhao</a> (<a href="Bib.html#ref-li2022esg" role="doc-biblioref">2022</a>)</span> (in China)) and to bank instability (<span class="citation"><a href="Bib.html#ref-chiaramonte2021esg" role="doc-biblioref">Chiaramonte et al.</a> (<a href="Bib.html#ref-chiaramonte2021esg" role="doc-biblioref">2021</a>)</span>).</li>
<li>
<strong>ESG is riskier</strong>. <span class="citation"><a href="Bib.html#ref-bianchi2010should" role="doc-biblioref">R. J. Bianchi et al.</a> (<a href="Bib.html#ref-bianchi2010should" role="doc-biblioref">2010</a>)</span> document a negative effect of ESG: green funds have higher betas than conventional funds during recessions. <span class="citation"><a href="Bib.html#ref-ng2020business" role="doc-biblioref">A. C. Ng and Rezaee</a> (<a href="Bib.html#ref-ng2020business" role="doc-biblioref">2020</a>)</span> find that ESG performance is positively linked to idiosyncratic volatility. <span class="citation"><a href="Bib.html#ref-fiordelisi2021esg" role="doc-biblioref">Fiordelisi et al.</a> (<a href="Bib.html#ref-fiordelisi2021esg" role="doc-biblioref">2021</a>)</span> find that ESG ETFs perform worse than the market during economic downturns. Similar conclusions are obtained by <span class="citation"><a href="Bib.html#ref-bansal2018socially" role="doc-biblioref">Bansal, Wu, and Yaron</a> (<a href="Bib.html#ref-bansal2018socially" role="doc-biblioref">2022</a>)</span>.</li>
<li>
<strong>Link is not clear.</strong> Just like for pure performance, risk is not unequivocally linked to ESG. CSR is weakly linked to risk (<span class="citation"><a href="Bib.html#ref-oikonomou2012impact" role="doc-biblioref">Oikonomou, Brooks, and Pavelin</a> (<a href="Bib.html#ref-oikonomou2012impact" role="doc-biblioref">2012</a>)</span>).</li>
<li>
<strong>Downside risk</strong>. On this topic, we refer to <span class="citation"><a href="Bib.html#ref-monti2019does" role="doc-biblioref">Monti et al.</a> (<a href="Bib.html#ref-monti2019does" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-shafer2018environmental" role="doc-biblioref">Shafer and Szado</a> (<a href="Bib.html#ref-shafer2018environmental" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-nofsinger2019institutional" role="doc-biblioref">J. R. Nofsinger, Sulaeman, and Varma</a> (<a href="Bib.html#ref-nofsinger2019institutional" role="doc-biblioref">2019</a>)</span> and <span class="citation"><a href="Bib.html#ref-richey2020good" role="doc-biblioref">Richey</a> (<a href="Bib.html#ref-richey2020good" role="doc-biblioref">2020</a>)</span>. The latter shows that sin stocks are defensive in bad times. For a focus on US banks, we recommend the contributions of <span class="citation"><a href="Bib.html#ref-tasnia2020impact" role="doc-biblioref">Tasnia, AlHabshi, and Rosman</a> (<a href="Bib.html#ref-tasnia2020impact" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-chaudhry2021carbon" role="doc-biblioref">Chaudhry, Saeed, and Ahmed</a> (<a href="Bib.html#ref-chaudhry2021carbon" role="doc-biblioref">2021</a>)</span>.</li>
<li>
<strong>Implied volatility</strong>. Firms that belong to industries with higher ESG-sales dynamism are associated with lower implied volatilities (<span class="citation"><a href="Bib.html#ref-patel2020not" role="doc-biblioref">Patel, Pearce II, and Oghazi</a> (<a href="Bib.html#ref-patel2020not" role="doc-biblioref">2020</a>)</span>).<br>
</li>
<li>
<strong>Systemic risk</strong>. Better corporate governance in banks helps reducing bank interconnectedness and maintaining financial stability (<span class="citation"><a href="Bib.html#ref-aevoae2022esg" role="doc-biblioref">Aevoae et al.</a> (<a href="Bib.html#ref-aevoae2022esg" role="doc-biblioref">2022</a>)</span>). <span class="citation"><a href="Bib.html#ref-cerqueti2021esg" role="doc-biblioref">Cerqueti et al.</a> (<a href="Bib.html#ref-cerqueti2021esg" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-cerqueti2022mitigating" role="doc-biblioref">Cerqueti et al.</a> (<a href="Bib.html#ref-cerqueti2022mitigating" role="doc-biblioref">2022</a>)</span> find similar mitigating effects of ESG in an exercise of stress scenarios of portfolio litigation.</li>
</ul>
</div>
<div id="other" class="section level2" number="4.7">
<h2>
<span class="header-section-number">4.7</span> ESG and other financial metrics<a class="anchor" aria-label="anchor" href="#other"><i class="fas fa-link"></i></a>
</h2>
<p>Beyond pure stock-market profitability, ESG seems to be favorably related to the metrics listed in Table <a href="Perf.html#tab:other">4.2</a>. With respect to valuation and cost of equity, let us briefly recall that, according to Gordon’s dividend model, the firm value is equal to <span class="math inline">\(V_0/(c-g)\)</span>, where <span class="math inline">\(c\)</span> is the cost of capital (e.g., weighted average cost of capital (WACC)) and <span class="math inline">\(g\)</span> is the growth rate of cash flows (or dividends). ESG issues can impact both channels. Firm value can increase if <span class="math inline">\(g\)</span> increases, or if <span class="math inline">\(c\)</span> decreases, e.g., because investors perceive lower risk. <span class="citation"><a href="Bib.html#ref-derrien2021esg" role="doc-biblioref">Derrien et al.</a> (<a href="Bib.html#ref-derrien2021esg" role="doc-biblioref">2021</a>)</span> show that with respect to ESG incidents, it is the first channel that matters most. </p>
<div class="inline-table"><table class="table table-sm">
<caption>
<span id="tab:other">TABLE 4.2: </span> ESG and other performance metrics</caption>
<colgroup>
<col width="50%">
<col width="50%">
</colgroup>
<thead><tr class="header">
<th>Performance metric</th>
<th>References</th>
</tr></thead>
<tbody>
<tr class="odd">
<td><img width="230/"></td>
<td><img width="600/"></td>
</tr>
<tr class="even">
<td><strong>Raw valuation</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-hillman2001shareholder" role="doc-biblioref">Hillman and Keim</a> (<a href="Bib.html#ref-hillman2001shareholder" role="doc-biblioref">2001</a>)</span>, <span class="citation"><a href="Bib.html#ref-konar2001does" role="doc-biblioref">Konar and Cohen</a> (<a href="Bib.html#ref-konar2001does" role="doc-biblioref">2001</a>)</span>, <span class="citation"><a href="Bib.html#ref-clarkson2004market" role="doc-biblioref">Clarkson, Li, and Richardson</a> (<a href="Bib.html#ref-clarkson2004market" role="doc-biblioref">2004</a>)</span>, <span class="citation"><a href="Bib.html#ref-bebchuk2005costs" role="doc-biblioref">L. A. Bebchuk and Cohen</a> (<a href="Bib.html#ref-bebchuk2005costs" role="doc-biblioref">2005</a>)</span>, <span class="citation"><a href="Bib.html#ref-hill2007corporate" role="doc-biblioref">R. P. Hill et al.</a> (<a href="Bib.html#ref-hill2007corporate" role="doc-biblioref">2007</a>)</span>, <span class="citation"><a href="Bib.html#ref-hong2009price" role="doc-biblioref">H. Hong and Kacperczyk</a> (<a href="Bib.html#ref-hong2009price" role="doc-biblioref">2009</a>)</span>, <span class="citation"><a href="Bib.html#ref-ammann2011corporate" role="doc-biblioref">Ammann, Oesch, and Schmid</a> (<a href="Bib.html#ref-ammann2011corporate" role="doc-biblioref">2011</a>)</span>, <span class="citation"><a href="Bib.html#ref-guenster2011economic" role="doc-biblioref">Guenster et al.</a> (<a href="Bib.html#ref-guenster2011economic" role="doc-biblioref">2011</a>)</span>, <span class="citation"><a href="Bib.html#ref-cai2012doing" role="doc-biblioref">Ye Cai, Jo, and Pan</a> (<a href="Bib.html#ref-cai2012doing" role="doc-biblioref">2012</a>)</span> (in sin industries), <span class="citation"><a href="Bib.html#ref-lourencco2012does" role="doc-biblioref">Lourenço et al.</a> (<a href="Bib.html#ref-lourencco2012does" role="doc-biblioref">2012</a>)</span>, <span class="citation"><a href="Bib.html#ref-servaes2013impact" role="doc-biblioref">Servaes and Tamayo</a> (<a href="Bib.html#ref-servaes2013impact" role="doc-biblioref">2013</a>)</span>, <span class="citation"><a href="Bib.html#ref-lund2016golden" role="doc-biblioref">Lund and Schonlau</a> (<a href="Bib.html#ref-lund2016golden" role="doc-biblioref">2016</a>)</span>, <span class="citation"><a href="Bib.html#ref-amihud2017settling" role="doc-biblioref">Amihud, Schmid, and Solomon</a> (<a href="Bib.html#ref-amihud2017settling" role="doc-biblioref">2017</a>)</span>, <span class="citation"><a href="Bib.html#ref-bajic2018csr" role="doc-biblioref">Bajic and Yurtoglu</a> (<a href="Bib.html#ref-bajic2018csr" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-tsukioka2018does" role="doc-biblioref">Tsukioka</a> (<a href="Bib.html#ref-tsukioka2018does" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-de2020csr" role="doc-biblioref">De Villiers, Ma, and Marques</a> (<a href="Bib.html#ref-de2020csr" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-johnson2020influence" role="doc-biblioref">J. A. Johnson et al.</a> (<a href="Bib.html#ref-johnson2020influence" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-ahsan2021corporate" role="doc-biblioref">Ahsan, Al-Gamrh, and Mirza</a> (<a href="Bib.html#ref-ahsan2021corporate" role="doc-biblioref">2021</a>)</span> (in China), <span class="citation"><a href="Bib.html#ref-bofinger2022corporate" role="doc-biblioref">Bofinger, Heyden, and Rock</a> (<a href="Bib.html#ref-bofinger2022corporate" role="doc-biblioref">2022</a>)</span> (misvaluation). <span class="citation"><a href="Bib.html#ref-cremers2017staggered" role="doc-biblioref">K. M. Cremers, Litov, and Sepe</a> (<a href="Bib.html#ref-cremers2017staggered" role="doc-biblioref">2017</a>)</span> document an absence of impact of staggered boards. This topic is reviewed in <span class="citation"><a href="Bib.html#ref-gerard2018esg" role="doc-biblioref">Gerard</a> (<a href="Bib.html#ref-gerard2018esg" role="doc-biblioref">2018</a>)</span>. </td>
</tr>
<tr class="odd">
<td><strong>Tobin’s q</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-mehran1995executive" role="doc-biblioref">Mehran</a> (<a href="Bib.html#ref-mehran1995executive" role="doc-biblioref">1995</a>)</span>, <span class="citation"><a href="Bib.html#ref-king2001does" role="doc-biblioref">A. A. King and Lenox</a> (<a href="Bib.html#ref-king2001does" role="doc-biblioref">2001</a>)</span>, <span class="citation"><a href="Bib.html#ref-king2002exploring" role="doc-biblioref">A. King and Lenox</a> (<a href="Bib.html#ref-king2002exploring" role="doc-biblioref">2002</a>)</span>, <span class="citation"><a href="Bib.html#ref-jiao2010stakeholder" role="doc-biblioref">Jiao</a> (<a href="Bib.html#ref-jiao2010stakeholder" role="doc-biblioref">2010</a>)</span>, <span class="citation"><a href="Bib.html#ref-jo2011corporate" role="doc-biblioref">Jo and Harjoto</a> (<a href="Bib.html#ref-jo2011corporate" role="doc-biblioref">2011</a>)</span>, <span class="citation"><a href="Bib.html#ref-cai2012doing" role="doc-biblioref">Ye Cai, Jo, and Pan</a> (<a href="Bib.html#ref-cai2012doing" role="doc-biblioref">2012</a>)</span> (in sin industries), <span class="citation"><a href="Bib.html#ref-jo2012causal" role="doc-biblioref">Jo and Harjoto</a> (<a href="Bib.html#ref-jo2012causal" role="doc-biblioref">2012</a>)</span>, <span class="citation"><a href="Bib.html#ref-lioui2012environmental" role="doc-biblioref">Lioui and Sharma</a> (<a href="Bib.html#ref-lioui2012environmental" role="doc-biblioref">2012</a>)</span> (negative relationship), <span class="citation"><a href="Bib.html#ref-cremers2013thirty" role="doc-biblioref">M. Cremers and Ferrell</a> (<a href="Bib.html#ref-cremers2013thirty" role="doc-biblioref">2014</a>)</span>, <span class="citation"><a href="Bib.html#ref-gregory2013exploring" role="doc-biblioref">A. Gregory and Whittaker</a> (<a href="Bib.html#ref-gregory2013exploring" role="doc-biblioref">2013</a>)</span>, <span class="citation"><a href="Bib.html#ref-lee2015impacts" role="doc-biblioref">K.-H. Lee, Min, and Yook</a> (<a href="Bib.html#ref-lee2015impacts" role="doc-biblioref">2015</a>)</span> (negatively impacted by CO<span class="math inline">\(_2\)</span> emissions, but positively by environmental R&D), <span class="citation"><a href="Bib.html#ref-ferrell2016socially" role="doc-biblioref">Ferrell, Liang, and Renneboog</a> (<a href="Bib.html#ref-ferrell2016socially" role="doc-biblioref">2016</a>)</span>, <span class="citation"><a href="Bib.html#ref-kang2016washing" role="doc-biblioref">C. Kang, Germann, and Grewal</a> (<a href="Bib.html#ref-kang2016washing" role="doc-biblioref">2016</a>)</span>, <span class="citation"><a href="Bib.html#ref-velte2017does" role="doc-biblioref">Velte</a> (<a href="Bib.html#ref-velte2017does" role="doc-biblioref">2017</a>)</span> (no impact), <span class="citation"><a href="Bib.html#ref-hasan2018corporate" role="doc-biblioref">Hasan et al.</a> (<a href="Bib.html#ref-hasan2018corporate" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-tsukioka2018does" role="doc-biblioref">Tsukioka</a> (<a href="Bib.html#ref-tsukioka2018does" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-radhouane2018impact" role="doc-biblioref">Radhouane et al.</a> (<a href="Bib.html#ref-radhouane2018impact" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-ioannou2019consequences" role="doc-biblioref">Ioannou and Serafeim</a> (<a href="Bib.html#ref-ioannou2019consequences" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-alareeni2020esg" role="doc-biblioref">Alareeni and Hamdan</a> (<a href="Bib.html#ref-alareeni2020esg" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-gantchev2020costs" role="doc-biblioref">Gantchev and Giannetti</a> (<a href="Bib.html#ref-gantchev2020costs" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-gerged2021corporate" role="doc-biblioref">Gerged, Beddewela, and Cowton</a> (<a href="Bib.html#ref-gerged2021corporate" role="doc-biblioref">2021</a>)</span> (in the Gulf countries). </td>
</tr>
<tr class="even">
<td><strong>Return on assets, and return on equity</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-herremans1993investigation" role="doc-biblioref">Herremans, Akathaporn, and McInnes</a> (<a href="Bib.html#ref-herremans1993investigation" role="doc-biblioref">1993</a>)</span>, <span class="citation"><a href="Bib.html#ref-mehran1995executive" role="doc-biblioref">Mehran</a> (<a href="Bib.html#ref-mehran1995executive" role="doc-biblioref">1995</a>)</span>, <span class="citation"><a href="Bib.html#ref-hart1996does" role="doc-biblioref">S. L. Hart and Ahuja</a> (<a href="Bib.html#ref-hart1996does" role="doc-biblioref">1996</a>)</span>, <span class="citation"><a href="Bib.html#ref-russo1997resource" role="doc-biblioref">Russo and Fouts</a> (<a href="Bib.html#ref-russo1997resource" role="doc-biblioref">1997</a>)</span>, <span class="citation"><a href="Bib.html#ref-king2002exploring" role="doc-biblioref">A. King and Lenox</a> (<a href="Bib.html#ref-king2002exploring" role="doc-biblioref">2002</a>)</span>, <span class="citation"><a href="Bib.html#ref-simpson2002link" role="doc-biblioref">Simpson and Kohers</a> (<a href="Bib.html#ref-simpson2002link" role="doc-biblioref">2002</a>)</span>, <span class="citation"><a href="Bib.html#ref-semenova2008financial" role="doc-biblioref">Semenova and Hassel</a> (<a href="Bib.html#ref-semenova2008financial" role="doc-biblioref">2008</a>)</span>, <span class="citation"><a href="Bib.html#ref-peiris2010relationship" role="doc-biblioref">Peiris and Evans</a> (<a href="Bib.html#ref-peiris2010relationship" role="doc-biblioref">2010</a>)</span>, <span class="citation"><a href="Bib.html#ref-guenster2011economic" role="doc-biblioref">Guenster et al.</a> (<a href="Bib.html#ref-guenster2011economic" role="doc-biblioref">2011</a>)</span>, <span class="citation"><a href="Bib.html#ref-jo2012causal" role="doc-biblioref">Jo and Harjoto</a> (<a href="Bib.html#ref-jo2012causal" role="doc-biblioref">2012</a>)</span>, <span class="citation"><a href="Bib.html#ref-lioui2012environmental" role="doc-biblioref">Lioui and Sharma</a> (<a href="Bib.html#ref-lioui2012environmental" role="doc-biblioref">2012</a>)</span> (negative relationship), <span class="citation"><a href="Bib.html#ref-christiansen2016gender" role="doc-biblioref">Christiansen et al.</a> (<a href="Bib.html#ref-christiansen2016gender" role="doc-biblioref">2016</a>)</span>, <span class="citation"><a href="Bib.html#ref-velte2017does" role="doc-biblioref">Velte</a> (<a href="Bib.html#ref-velte2017does" role="doc-biblioref">2017</a>)</span>, <span class="citation"><a href="Bib.html#ref-tsukioka2018does" role="doc-biblioref">Tsukioka</a> (<a href="Bib.html#ref-tsukioka2018does" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-yin2019relationship" role="doc-biblioref">Yin et al.</a> (<a href="Bib.html#ref-yin2019relationship" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-alareeni2020esg" role="doc-biblioref">Alareeni and Hamdan</a> (<a href="Bib.html#ref-alareeni2020esg" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-gantchev2020costs" role="doc-biblioref">Gantchev and Giannetti</a> (<a href="Bib.html#ref-gantchev2020costs" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-gerged2021corporate" role="doc-biblioref">Gerged, Beddewela, and Cowton</a> (<a href="Bib.html#ref-gerged2021corporate" role="doc-biblioref">2021</a>)</span> (in the Gulf countries), <span class="citation"><a href="Bib.html#ref-mohamed2021sustainability" role="doc-biblioref">Mohamed Buallay et al.</a> (<a href="Bib.html#ref-mohamed2021sustainability" role="doc-biblioref">2021</a>)</span> (financial sector), <span class="citation"><a href="Bib.html#ref-rossi2021does" role="doc-biblioref">Matteo Rossi et al.</a> (<a href="Bib.html#ref-rossi2021does" role="doc-biblioref">2021</a>)</span>. </td>
</tr>
<tr class="odd">
<td><strong>Cost of equity</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-sharfman2008environmental" role="doc-biblioref">M. P. Sharfman and Fernando</a> (<a href="Bib.html#ref-sharfman2008environmental" role="doc-biblioref">2008</a>)</span>, <span class="citation"><a href="Bib.html#ref-dhaliwal2011voluntary" role="doc-biblioref">Dhaliwal et al.</a> (<a href="Bib.html#ref-dhaliwal2011voluntary" role="doc-biblioref">2011</a>)</span>, <span class="citation"><a href="Bib.html#ref-el2011does" role="doc-biblioref">El Ghoul et al.</a> (<a href="Bib.html#ref-el2011does" role="doc-biblioref">2011</a>)</span>, <span class="citation"><a href="Bib.html#ref-he2013carbon" role="doc-biblioref">Yu He, Tang, and Wang</a> (<a href="Bib.html#ref-he2013carbon" role="doc-biblioref">2013</a>)</span>, <span class="citation"><a href="Bib.html#ref-chava2014environmental" role="doc-biblioref">Chava</a> (<a href="Bib.html#ref-chava2014environmental" role="doc-biblioref">2014</a>)</span>, <span class="citation"><a href="Bib.html#ref-ng2015business" role="doc-biblioref">A. C. Ng and Rezaee</a> (<a href="Bib.html#ref-ng2015business" role="doc-biblioref">2015</a>)</span>, <span class="citation"><a href="Bib.html#ref-park2018relationship" role="doc-biblioref">Park and Noh</a> (<a href="Bib.html#ref-park2018relationship" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-gao2019standing" role="doc-biblioref">Gao, He, and Wu</a> (<a href="Bib.html#ref-gao2019standing" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-matsumura2020climate" role="doc-biblioref">Matsumura, Prakash, and Vera-Muñoz</a> (<a href="Bib.html#ref-matsumura2020climate" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-piechocka2021impact" role="doc-biblioref">Piechocka-Kaluzna et al.</a> (<a href="Bib.html#ref-piechocka2021impact" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-mariani2021does" role="doc-biblioref">Mariani et al.</a> (<a href="Bib.html#ref-mariani2021does" role="doc-biblioref">2021</a>)</span> (impact on the weighted average cost of capital). </td>
</tr>
<tr class="even">
<td><strong>Operating performance</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-guenster2011economic" role="doc-biblioref">Guenster et al.</a> (<a href="Bib.html#ref-guenster2011economic" role="doc-biblioref">2011</a>)</span>) and <strong>earnings</strong> ( <span class="citation"><a href="Bib.html#ref-borgers2013stakeholder" role="doc-biblioref">Borgers et al.</a> (<a href="Bib.html#ref-borgers2013stakeholder" role="doc-biblioref">2013</a>)</span>, <span class="citation"><a href="Bib.html#ref-velte2019bidirectional" role="doc-biblioref">Velte</a> (<a href="Bib.html#ref-velte2019bidirectional" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-kim2020capitalizing" role="doc-biblioref">T. Kim and Kim</a> (<a href="Bib.html#ref-kim2020capitalizing" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-jia2021corporate" role="doc-biblioref">Jia and Li</a> (<a href="Bib.html#ref-jia2021corporate" role="doc-biblioref">2021</a>)</span>. </td>
</tr>
<tr class="odd">
<td><strong>Cost of debt and credit rating</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-ashbaugh2006effects" role="doc-biblioref">Ashbaugh-Skaife, Collins, and LaFond</a> (<a href="Bib.html#ref-ashbaugh2006effects" role="doc-biblioref">2006</a>)</span>, <span class="citation"><a href="Bib.html#ref-attig2013corporate" role="doc-biblioref">Attig et al.</a> (<a href="Bib.html#ref-attig2013corporate" role="doc-biblioref">2013</a>)</span>, <span class="citation"><a href="Bib.html#ref-chava2014environmental" role="doc-biblioref">Chava</a> (<a href="Bib.html#ref-chava2014environmental" role="doc-biblioref">2014</a>)</span>, <span class="citation"><a href="Bib.html#ref-jiraporn2014does" role="doc-biblioref">Jiraporn et al.</a> (<a href="Bib.html#ref-jiraporn2014does" role="doc-biblioref">2014</a>)</span>, <span class="citation"><a href="Bib.html#ref-jung2018carbon" role="doc-biblioref">Jung, Herbohn, and Clarkson</a> (<a href="Bib.html#ref-jung2018carbon" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-caragnano2020worth" role="doc-biblioref">Caragnano et al.</a> (<a href="Bib.html#ref-caragnano2020worth" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-kling2021impact" role="doc-biblioref">Kling et al.</a> (<a href="Bib.html#ref-kling2021impact" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-michalski2021corporate" role="doc-biblioref">Michalski and Low</a> (<a href="Bib.html#ref-michalski2021corporate" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-raimo2021extending" role="doc-biblioref">Raimo et al.</a> (<a href="Bib.html#ref-raimo2021extending" role="doc-biblioref">2021</a>)</span>; and <strong>default risk</strong> (<span class="citation"><a href="Bib.html#ref-nadaraja2020does" role="doc-biblioref">Nadaraja et al.</a> (<a href="Bib.html#ref-nadaraja2020does" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-li2022esg" role="doc-biblioref">H. Li, Zhang, and Zhao</a> (<a href="Bib.html#ref-li2022esg" role="doc-biblioref">2022</a>)</span> (in China)).</td>
</tr>
<tr class="even">
<td><strong>Equity forecasts</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-gregory2014corporate" role="doc-biblioref">A. Gregory, Tharyan, and Whittaker</a> (<a href="Bib.html#ref-gregory2014corporate" role="doc-biblioref">2014</a>)</span> (with rationale of discounted cash flows), <span class="citation"><a href="Bib.html#ref-derrien2021esg" role="doc-biblioref">Derrien et al.</a> (<a href="Bib.html#ref-derrien2021esg" role="doc-biblioref">2021</a>)</span>.</td>
</tr>
<tr class="odd">
<td><strong>Dividend policy</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-cheung2018corporate" role="doc-biblioref">Cheung, Hu, and Schwiebert</a> (<a href="Bib.html#ref-cheung2018corporate" role="doc-biblioref">2018</a>)</span>, <span class="citation"><a href="Bib.html#ref-benlemlih2019corporate" role="doc-biblioref">Benlemlih</a> (<a href="Bib.html#ref-benlemlih2019corporate" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-de2020csr" role="doc-biblioref">De Villiers, Ma, and Marques</a> (<a href="Bib.html#ref-de2020csr" role="doc-biblioref">2020</a>)</span> and <span class="citation"><a href="Bib.html#ref-matos2020does" role="doc-biblioref">P. V. Matos, Barros, and Sarmento</a> (<a href="Bib.html#ref-matos2020does" role="doc-biblioref">2020</a>)</span>.</td>
</tr>
<tr class="even">
<td><strong>Trade credit</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-cheung2019corporate" role="doc-biblioref">Cheung and Pok</a> (<a href="Bib.html#ref-cheung2019corporate" role="doc-biblioref">2019</a>)</span>, <span class="citation"><a href="Bib.html#ref-xu2019corporate" role="doc-biblioref">H. Xu, Wu, and Dao</a> (<a href="Bib.html#ref-xu2019corporate" role="doc-biblioref">2020</a>)</span>.</td>
</tr>
<tr class="odd">
<td><strong>Exit scenarios and M&A</strong></td>
<td>For public firms (being acquired, going bankrupt, or going private), see <span class="citation"><a href="Bib.html#ref-goktan2018corporate" role="doc-biblioref">Goktan, Kieschnick, and Moussawi</a> (<a href="Bib.html#ref-goktan2018corporate" role="doc-biblioref">2018</a>)</span>. For M&A, see <span class="citation"><a href="Bib.html#ref-gomes2018does" role="doc-biblioref">Gomes and Marsat</a> (<a href="Bib.html#ref-gomes2018does" role="doc-biblioref">2018</a>)</span> and <span class="citation"><a href="Bib.html#ref-caiazza2021role" role="doc-biblioref">Caiazza, Galloppo, and Paimanova</a> (<a href="Bib.html#ref-caiazza2021role" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-arouri2019corporate" role="doc-biblioref">Arouri, Gomes, and Pukthuanthong</a> (<a href="Bib.html#ref-arouri2019corporate" role="doc-biblioref">2019</a>)</span> (on uncertainty), <span class="citation"><a href="Bib.html#ref-alexandridis2021corporate" role="doc-biblioref">Alexandridis et al.</a> (<a href="Bib.html#ref-alexandridis2021corporate" role="doc-biblioref">2021</a>)</span> (on CSR culture), <span class="citation"><a href="Bib.html#ref-jost2021does" role="doc-biblioref">Jost et al.</a> (<a href="Bib.html#ref-jost2021does" role="doc-biblioref">2021</a>)</span> (the effect is limited), <span class="citation"><a href="Bib.html#ref-yen2019market" role="doc-biblioref">Yen and André</a> (<a href="Bib.html#ref-yen2019market" role="doc-biblioref">2019</a>)</span> (on emerging markets) and <span class="citation"><a href="Bib.html#ref-reynolds2021role" role="doc-biblioref">Reynolds and Hassett</a> (<a href="Bib.html#ref-reynolds2021role" role="doc-biblioref">2021</a>)</span> for a topical discussion.</td>
</tr>
<tr class="even">
<td><strong>Equity offerings</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-dutordoir2018corporate" role="doc-biblioref">Dutordoir, Strong, and Sun</a> (<a href="Bib.html#ref-dutordoir2018corporate" role="doc-biblioref">2018</a>)</span>.</td>
</tr>
<tr class="odd">
<td><strong>IPOs</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-reber2021esg" role="doc-biblioref">Reber, Gold, and Gold</a> (<a href="Bib.html#ref-reber2021esg" role="doc-biblioref">2021</a>)</span> (voluntary disclosure reduces risk post-IPO), <span class="citation"><a href="Bib.html#ref-huang2019ipo" role="doc-biblioref">F. Huang et al.</a> (<a href="Bib.html#ref-huang2019ipo" role="doc-biblioref">2019</a>)</span>.</td>
</tr>
<tr class="even">
<td><strong>Employee morale and productivity</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-burbano2019getting" role="doc-biblioref">Burbano</a> (<a href="Bib.html#ref-burbano2019getting" role="doc-biblioref">2021</a>)</span>, <span class="citation"><a href="Bib.html#ref-hedblom2019toward" role="doc-biblioref">Hedblom, Hickman, and List</a> (<a href="Bib.html#ref-hedblom2019toward" role="doc-biblioref">2021</a>)</span>. </td>
</tr>
<tr class="odd">
<td><strong>Customer satisfaction</strong></td>
<td><span class="citation"><a href="Bib.html#ref-fornell2016stock" role="doc-biblioref">Fornell, Morgeson III, and Hult</a> (<a href="Bib.html#ref-fornell2016stock" role="doc-biblioref">2016</a>)</span></td>
</tr>
<tr class="even">
<td><strong>Insider trading</strong></td>
<td>
<span class="citation"><a href="Bib.html#ref-cui2015corporate" role="doc-biblioref">J. Cui, Jo, and Li</a> (<a href="Bib.html#ref-cui2015corporate" role="doc-biblioref">2015</a>)</span> (positive link!)</td>
</tr>
<tr class="odd">
<td>
<strong>Economic perf</strong> (sustainable growth and high ROI)</td>
<td><span class="citation"><a href="Bib.html#ref-ferrero2016effect" role="doc-biblioref">Ferrero-Ferrero, Fernández-Izquierdo, and Muñoz-Torres</a> (<a href="Bib.html#ref-ferrero2016effect" role="doc-biblioref">2016</a>)</span></td>
</tr>
</tbody>
</table></div>
</div>
<div id="empirical-illustration" class="section level2" number="4.8">
<h2>
<span class="header-section-number">4.8</span> Empirical illustration<a class="anchor" aria-label="anchor" href="#empirical-illustration"><i class="fas fa-link"></i></a>
</h2>
<p>We close this chapter with a small empirical exercise. We compare two US equity indices: one conventional, and one ESG-based. The conventional index is the S&P 500, arguably the reference yardstick for US equities, both among practitioners and scholars alike. The ESG portfolio is the iShares MSCI USA ESG Select ETF, which performs sustainability screens that are based on sectors, as well as on ESG scores. In December 2020, the index comprised 202 stocks, which makes it less diversified, compared to the S&P 500.<a class="footnote-ref" tabindex="0" data-toggle="popover" data-content="<p>This is subject for debate because the index is capitalization-weighted, which means that a few dozens of stocks account for a large majority of the weights of the portfolio.</p>"><sup>31</sup></a> The series start on January 28<span class="math inline">\(^{th}\)</span>, 2005, which is the inception date of the ESG index. It is notoriously complicated to find reliable ESG data prior to 2005.
In Figure <a href="Perf.html#fig:perf">4.2</a>, we plot the time-series of the two indices.</p>
<div class="sourceCode" id="cb9"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span class="kw"><a href="https://rdrr.io/r/base/library.html">library</a></span><span class="op">(</span><span class="va"><a href="http://www.quantmod.com">quantmod</a></span><span class="op">)</span> <span class="co"># Package for financial data retrieval</span>
<span class="kw"><a href="https://rdrr.io/r/base/library.html">library</a></span><span class="op">(</span><span class="va"><a href="https://lubridate.tidyverse.org">lubridate</a></span><span class="op">)</span> <span class="co"># Package for date management</span>
<span class="va">tickers</span> <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"SUSA"</span>, <span class="st">"SPY"</span><span class="op">)</span> <span class="co"># Ticker names</span>
<span class="va">prices</span> <span class="op"><-</span> <span class="fu"><a href="https://rdrr.io/pkg/quantmod/man/getSymbols.html">getSymbols</a></span><span class="op">(</span><span class="va">tickers</span>, src <span class="op">=</span> <span class="st">'yahoo'</span>, <span class="co"># Yahoo source </span>
from <span class="op">=</span> <span class="st">"2005-01-28"</span>,
to <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/Sys.time.html">Sys.Date</a></span><span class="op">(</span><span class="op">)</span>,
auto.assign <span class="op">=</span> <span class="cn">TRUE</span>,
warnings <span class="op">=</span> <span class="cn">FALSE</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/leaflet/man/leaflet-imports.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/purrr/man/map.html">map</a></span><span class="op">(</span><span class="op">~</span><span class="fu"><a href="https://rdrr.io/pkg/quantmod/man/OHLC.Transformations.html">Ad</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/r/base/get.html">get</a></span><span class="op">(</span><span class="va">.</span><span class="op">)</span><span class="op">)</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/leaflet/man/leaflet-imports.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/purrr/man/reduce.html">reduce</a></span><span class="op">(</span><span class="va">merge</span><span class="op">)</span>
<span class="va">norm_</span> <span class="op"><-</span> <span class="kw">function</span><span class="op">(</span><span class="va">v</span><span class="op">)</span><span class="op">{</span><span class="kw"><a href="https://rdrr.io/r/base/function.html">return</a></span><span class="op">(</span><span class="va">v</span><span class="op">/</span><span class="va">v</span><span class="op">[</span><span class="fl">1</span><span class="op">]</span><span class="op">)</span><span class="op">}</span>
<span class="va">prices</span> <span class="op"><-</span> <span class="fu"><a href="https://rdrr.io/r/base/apply.html">apply</a></span><span class="op">(</span><span class="va">prices</span>, <span class="fl">2</span>, <span class="va">norm_</span><span class="op">)</span>
<span class="va">prices</span> <span class="op"><-</span> <span class="fu"><a href="https://r-spatial.github.io/sf/reference/tibble.html">tibble</a></span><span class="op">(</span>date <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/zoo/man/yearmon.html">as.Date</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/tibble/man/rownames.html">rownames</a></span><span class="op">(</span><span class="va">prices</span><span class="op">)</span><span class="op">)</span>, <span class="fu"><a href="https://rdrr.io/pkg/tibble/man/as_tibble.html">as_tibble</a></span><span class="op">(</span><span class="va">prices</span><span class="op">)</span><span class="op">)</span>
<span class="fu"><a href="https://rdrr.io/r/base/colnames.html">colnames</a></span><span class="op">(</span><span class="va">prices</span><span class="op">)</span><span class="op">[</span><span class="fl">2</span><span class="op">:</span><span class="fl">3</span><span class="op">]</span><span class="op"><-</span><span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"MSCI_ESG"</span>, <span class="st">"SP500"</span><span class="op">)</span>
<span class="va">prices</span> <span class="op"><a href="https://rdrr.io/pkg/leaflet/man/leaflet-imports.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/tidyr/man/pivot_longer.html">pivot_longer</a></span><span class="op">(</span><span class="op">-</span><span class="va">date</span>, names_to <span class="op">=</span> <span class="st">"Index"</span>, values_to <span class="op">=</span> <span class="st">"Value"</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/leaflet/man/leaflet-imports.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggplot.html">ggplot</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/aes.html">aes</a></span><span class="op">(</span>x <span class="op">=</span> <span class="va">date</span>, y <span class="op">=</span> <span class="va">Value</span>, color <span class="op">=</span> <span class="va">Index</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/geom_path.html">geom_line</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggtheme.html">theme_light</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/scale_manual.html">scale_color_manual</a></span><span class="op">(</span>values <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"#0DCD64"</span>, <span class="st">"#0D70CD"</span><span class="op">)</span>, labels <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"MSCI ESG"</span>, <span class="st">"S&P 500"</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/theme.html">theme</a></span><span class="op">(</span>legend.position <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="fl">0.2</span>, <span class="fl">0.8</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">xlab</a></span><span class="op">(</span><span class="st">"Date"</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">ylab</a></span><span class="op">(</span><span class="st">"Index Value"</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/theme.html">theme</a></span><span class="op">(</span>text <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_text</a></span><span class="op">(</span>size <span class="op">=</span> <span class="fl">16</span><span class="op">)</span><span class="op">)</span> </code></pre></div>
<div class="figure">
<span style="display:block;" id="fig:perf"></span>
<img src="ESG_p_files/figure-html/perf-1.png" alt="Performance comparison. We plot the index values (S&P 500 and iShares MSCI USA ESG Select ETF), onward from January 28th, 2005 (inception date of the ESG portfolio). The series are normalized so that their initial value is one." width="672"><p class="caption">
FIGURE 4.2: Performance comparison. We plot the index values (S&P 500 and iShares MSCI USA ESG Select ETF), onward from January 28th, 2005 (inception date of the ESG portfolio). The series are normalized so that their initial value is one.
</p>
</div>
<p>At first sight, the first-order conclusion is that there is not much difference between the two series. In the first half of the sample, the lines are hardly distinguishable. Between 2012 and 2019, the S&P 500 seems to outperform marginally, but 2020 has eroded part of this superiority (see below). In Table <a href="Perf.html#tab:perf2">4.3</a>, we compute a few performance metrics.</p>
<div class="sourceCode" id="cb10"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span class="kw"><a href="https://rdrr.io/r/base/library.html">library</a></span><span class="op">(</span><span class="va">tseries</span><span class="op">)</span>
<span class="kw"><a href="https://rdrr.io/r/base/library.html">library</a></span><span class="op">(</span><span class="va"><a href="http://haozhu233.github.io/kableExtra/">kableExtra</a></span><span class="op">)</span>
<span class="va">returns</span> <span class="op"><-</span> <span class="va">prices</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span> <span class="co"># returns</span>
<span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/mutate.html">mutate</a></span><span class="op">(</span>MSCI_ESG <span class="op">=</span> <span class="va">MSCI_ESG</span><span class="op">/</span><span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/lead-lag.html">lag</a></span><span class="op">(</span><span class="va">MSCI_ESG</span><span class="op">)</span> <span class="op">-</span> <span class="fl">1</span>,
SP500 <span class="op">=</span> <span class="va">SP500</span><span class="op">/</span><span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/lead-lag.html">lag</a></span><span class="op">(</span><span class="va">SP500</span><span class="op">)</span> <span class="op">-</span> <span class="fl">1</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/r/stats/na.fail.html">na.omit</a></span><span class="op">(</span><span class="op">)</span>
<span class="va">prices</span> <span class="op"><-</span> <span class="va">prices</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span> <span class="fu"><a href="https://rdrr.io/r/stats/na.fail.html">na.omit</a></span><span class="op">(</span><span class="op">)</span> <span class="co"># remove missing points</span>
<span class="va">ret</span> <span class="op"><-</span> <span class="op">(</span><span class="va">prices</span><span class="op">[</span><span class="fu"><a href="https://rdrr.io/r/base/nrow.html">nrow</a></span><span class="op">(</span><span class="va">prices</span><span class="op">)</span>, <span class="fl">2</span><span class="op">:</span><span class="fl">3</span><span class="op">]</span> <span class="op">/</span> <span class="va">prices</span><span class="op">[</span><span class="fl">1</span>, <span class="fl">2</span><span class="op">:</span><span class="fl">3</span><span class="op">]</span><span class="op">)</span> <span class="op">^</span> <span class="op">(</span><span class="fl">252</span><span class="op">/</span><span class="fu"><a href="https://rdrr.io/r/base/nrow.html">nrow</a></span><span class="op">(</span><span class="va">prices</span><span class="op">)</span><span class="op">)</span><span class="op">-</span><span class="fl">1</span> <span class="co"># returns</span>
<span class="va">vol</span> <span class="op"><-</span> <span class="fu"><a href="https://rdrr.io/r/base/apply.html">apply</a></span><span class="op">(</span><span class="va">returns</span><span class="op">[</span>,<span class="fl">2</span><span class="op">:</span><span class="fl">3</span><span class="op">]</span>, <span class="fl">2</span>, <span class="va">sd</span><span class="op">)</span> <span class="op">*</span> <span class="fu"><a href="https://rdrr.io/r/base/MathFun.html">sqrt</a></span><span class="op">(</span><span class="fl">252</span><span class="op">)</span> <span class="co"># volatility</span>
<span class="va">ratio</span> <span class="op"><-</span> <span class="va">ret</span><span class="op">/</span><span class="va">vol</span> <span class="co"># Sharpe ratio (proxy)</span>
<span class="va">mdd</span> <span class="op"><-</span> <span class="fu"><a href="https://rdrr.io/r/base/apply.html">apply</a></span><span class="op">(</span><span class="va">prices</span><span class="op">[</span>,<span class="fl">2</span><span class="op">:</span><span class="fl">3</span><span class="op">]</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span> <span class="fu"><a href="https://rdrr.io/r/stats/na.fail.html">na.omit</a></span><span class="op">(</span><span class="op">)</span>, <span class="fl">2</span>, <span class="va">maxdrawdown</span><span class="op">)</span> <span class="co"># max drawdown</span>
<span class="va">var5</span> <span class="op"><-</span> <span class="fu"><a href="https://rdrr.io/r/base/apply.html">apply</a></span><span class="op">(</span><span class="va">returns</span><span class="op">[</span>,<span class="fl">2</span><span class="op">:</span><span class="fl">3</span><span class="op">]</span>, <span class="fl">2</span>, <span class="kw">function</span><span class="op">(</span><span class="va">v</span><span class="op">)</span> <span class="fu"><a href="https://rdrr.io/r/stats/quantile.html">quantile</a></span><span class="op">(</span><span class="va">v</span>, probs <span class="op">=</span> <span class="fl">0.05</span><span class="op">)</span><span class="op">)</span> <span class="co"># Value-at-Risk</span>
<span class="fu"><a href="https://r-spatial.github.io/sf/reference/tibble.html">tibble</a></span><span class="op">(</span>Index <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"MSCI ESG"</span>, <span class="st">"S&P 500"</span><span class="op">)</span>,
Return <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/numeric.html">as.numeric</a></span><span class="op">(</span><span class="va">ret</span><span class="op">)</span>,
Volatility <span class="op">=</span> <span class="va">vol</span>,
Ratio <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/numeric.html">as.numeric</a></span><span class="op">(</span><span class="va">ratio</span><span class="op">)</span>,
MaxDrawdown <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/numeric.html">as.numeric</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="op">(</span><span class="va">prices</span><span class="op">[</span><span class="va">mdd</span><span class="op">$</span><span class="va">MSCI_ESG</span><span class="op">$</span><span class="va">to</span>,<span class="fl">2</span><span class="op">]</span><span class="op">-</span><span class="va">prices</span><span class="op">[</span><span class="va">mdd</span><span class="op">$</span><span class="va">MSCI_ESG</span><span class="op">$</span><span class="va">from</span>,<span class="fl">2</span><span class="op">]</span><span class="op">)</span><span class="op">/</span><span class="va">prices</span><span class="op">[</span><span class="va">mdd</span><span class="op">$</span><span class="va">MSCI_ESG</span><span class="op">$</span><span class="va">from</span>,<span class="fl">2</span><span class="op">]</span>,
<span class="op">(</span><span class="va">prices</span><span class="op">[</span><span class="va">mdd</span><span class="op">$</span><span class="va">SP500</span><span class="op">$</span><span class="va">to</span>,<span class="fl">3</span><span class="op">]</span><span class="op">-</span><span class="va">prices</span><span class="op">[</span><span class="va">mdd</span><span class="op">$</span><span class="va">SP500</span><span class="op">$</span><span class="va">from</span>,<span class="fl">3</span><span class="op">]</span><span class="op">)</span><span class="op">/</span><span class="va">prices</span><span class="op">[</span><span class="va">mdd</span><span class="op">$</span><span class="va">SP500</span><span class="op">$</span><span class="va">from</span>,<span class="fl">3</span><span class="op">]</span><span class="op">)</span><span class="op">)</span>,
ValueatRisk <span class="op">=</span> <span class="va">var5</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu">kableExtra</span><span class="fu">::</span><span class="fu"><a href="https://rdrr.io/pkg/knitr/man/kable.html">kable</a></span><span class="op">(</span>caption <span class="op">=</span> <span class="st">'Performance indicators.'</span><span class="op">)</span></code></pre></div>
<div class="inline-table"><table class="table table-sm">
<caption>
<span id="tab:perf2">TABLE 4.3: </span>Performance indicators.
</caption>
<thead><tr>
<th style="text-align:left;">
Index
</th>
<th style="text-align:right;">
Return
</th>
<th style="text-align:right;">
Volatility
</th>
<th style="text-align:right;">
Ratio
</th>
<th style="text-align:right;">
MaxDrawdown
</th>
<th style="text-align:right;">
ValueatRisk
</th>
</tr></thead>
<tbody>
<tr>
<td style="text-align:left;">
MSCI ESG
</td>
<td style="text-align:right;">
0.0868963
</td>
<td style="text-align:right;">
0.1864341
</td>
<td style="text-align:right;">
0.4660966
</td>
<td style="text-align:right;">
-0.2637472
</td>
<td style="text-align:right;">
-0.0175106
</td>
</tr>
<tr>
<td style="text-align:left;">
S&P 500
</td>
<td style="text-align:right;">
0.0915205
</td>
<td style="text-align:right;">
0.1951129
</td>
<td style="text-align:right;">
0.4690642
</td>
<td style="text-align:right;">
-0.3371726
</td>
<td style="text-align:right;">
-0.0181815
</td>
</tr>
</tbody>
</table></div>
<p>All values are arguably close, and no difference in any metric would pass a test of statistical significance. The broad market index has a marginal superiority in returns, but it is bested across both risk measures. The volatility-adjusted average return is even slightly higher for the ESG index. Overall these results are most in line with those of Section <a href="Perf.html#mixed">4.3</a>: in the long run, its is hard to find evidence (in this small sample) of outperformance in one way or the other.</p>
<p>Researchers and practitioners often seek to determine if ESG exposure acts as a hedge in bad times. To shed some light on this question, we zoom in on two sub-periods of the sample, namely the years 2008–2009, and 2020. These are shown in Figure <a href="Perf.html#fig:perf3">4.3</a>, where the series are scaled to start at unit value.</p>
<div class="sourceCode" id="cb11"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span class="va">g2</span> <span class="op"><-</span> <span class="va">prices</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/filter.html">filter</a></span><span class="op">(</span><span class="va">date</span> <span class="op">></span> <span class="st">"2019-12-31"</span>, <span class="va">date</span> <span class="op"><</span> <span class="st">"2021-06-30"</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/mutate.html">mutate</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/across.html">across</a></span><span class="op">(</span>.cols <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="fl">2</span>,<span class="fl">3</span><span class="op">)</span>, <span class="va">norm_</span><span class="op">)</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/tidyr/man/pivot_longer.html">pivot_longer</a></span><span class="op">(</span><span class="op">-</span><span class="va">date</span>, names_to <span class="op">=</span> <span class="st">"Index"</span>, values_to <span class="op">=</span> <span class="st">"Value"</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggplot.html">ggplot</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/aes.html">aes</a></span><span class="op">(</span>x <span class="op">=</span> <span class="va">date</span>, y <span class="op">=</span> <span class="va">Value</span>, color <span class="op">=</span> <span class="va">Index</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/geom_path.html">geom_line</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggtheme.html">theme_light</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/scale_manual.html">scale_color_manual</a></span><span class="op">(</span>values <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"#0DCD64"</span>, <span class="st">"#0D70CD"</span><span class="op">)</span>, labels <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"MSCI ESG"</span>, <span class="st">"S&P 500"</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/theme.html">theme</a></span><span class="op">(</span>legend.position <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="fl">0.75</span>, <span class="fl">0.2</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">xlab</a></span><span class="op">(</span><span class="st">"Date"</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">ylab</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_blank</a></span><span class="op">(</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/theme.html">theme</a></span><span class="op">(</span>text <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_text</a></span><span class="op">(</span>size <span class="op">=</span> <span class="fl">14</span><span class="op">)</span>, aspect.ratio <span class="op">=</span> <span class="fl">0.8</span><span class="op">)</span>
<span class="va">g1</span> <span class="op"><-</span><span class="va">prices</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/filter.html">filter</a></span><span class="op">(</span><span class="va">date</span> <span class="op">></span> <span class="st">"2007-12-31"</span>, <span class="va">date</span> <span class="op"><</span> <span class="st">"2010-01-01"</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/mutate.html">mutate</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/across.html">across</a></span><span class="op">(</span>.cols <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="fl">2</span>,<span class="fl">3</span><span class="op">)</span>, <span class="va">norm_</span><span class="op">)</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/tidyr/man/pivot_longer.html">pivot_longer</a></span><span class="op">(</span><span class="op">-</span><span class="va">date</span>, names_to <span class="op">=</span> <span class="st">"Index"</span>, values_to <span class="op">=</span> <span class="st">"Value"</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggplot.html">ggplot</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/aes.html">aes</a></span><span class="op">(</span>x <span class="op">=</span> <span class="va">date</span>, y <span class="op">=</span> <span class="va">Value</span>, color <span class="op">=</span> <span class="va">Index</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/geom_path.html">geom_line</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggtheme.html">theme_light</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/scale_manual.html">scale_color_manual</a></span><span class="op">(</span>values <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"#0DCD64"</span>, <span class="st">"#0D70CD"</span><span class="op">)</span>, labels <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"MSCI ESG"</span>, <span class="st">"S&P 500"</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/theme.html">theme</a></span><span class="op">(</span>legend.position <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="fl">0.65</span>, <span class="fl">0.8</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">xlab</a></span><span class="op">(</span><span class="st">"Date"</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">ylab</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_blank</a></span><span class="op">(</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/theme.html">theme</a></span><span class="op">(</span>text <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_text</a></span><span class="op">(</span>size <span class="op">=</span> <span class="fl">14</span><span class="op">)</span>, aspect.ratio <span class="op">=</span> <span class="fl">0.8</span><span class="op">)</span>
<span class="va">g1</span> <span class="op">+</span> <span class="va">g2</span></code></pre></div>
<div class="figure">
<span style="display:block;" id="fig:perf3"></span>
<img src="ESG_p_files/figure-html/perf3-1.png" alt="Focus on 2008–2009 and 2020–2021. We plot the index values, from January 1st, 2008 to December 31st, 2009 (left panel) and from January 1st, 2020 to the end of June 2021 (right panel). The series are normalized so that their initial value is 1" width="672"><p class="caption">
FIGURE 4.3: Focus on 2008–2009 and 2020–2021. We plot the index values, from January 1st, 2008 to December 31st, 2009 (left panel) and from January 1st, 2020 to the end of June 2021 (right panel). The series are normalized so that their initial value is 1
</p>
</div>
<p>In 2020 (right panel), the two curves move closely together until April, which means that the sustainable tilt did not immunize the ESG portfolio against the crash.<a class="footnote-ref" tabindex="0" data-toggle="popover" data-content='<p>Again, researchers disagree on this matter. <span class="citation"><a href="Bib.html#ref-diaz2020reconsidering" role="doc-biblioref">V. Diaz, Ibrushi, and Zhao</a> (<a href="Bib.html#ref-diaz2020reconsidering" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-singh2020covid" role="doc-biblioref">Singh</a> (<a href="Bib.html#ref-singh2020covid" role="doc-biblioref">2020</a>)</span> report some hedging benefits from ESG exposure, while <span class="citation"><a href="Bib.html#ref-folger2020esg" role="doc-biblioref">Folger-Laronde et al.</a> (<a href="Bib.html#ref-folger2020esg" role="doc-biblioref">2020</a>)</span>, <span class="citation"><a href="Bib.html#ref-demers2020esg" role="doc-biblioref">Demers et al.</a> (<a href="Bib.html#ref-demers2020esg" role="doc-biblioref">2021</a>)</span> and <span class="citation"><a href="Bib.html#ref-mahmoud2020anatomy" role="doc-biblioref">Mahmoud and Meyer</a> (<a href="Bib.html#ref-mahmoud2020anatomy" role="doc-biblioref">2020</a>)</span> do not.</p>'><sup>32</sup></a> To a certain extent, this is also true for the subprime crisis in 2008.</p>
<p>However, the interesting pattern is probably revealed <em>after</em> the crises. It is in the aftermath of the crashes that differences materialize, to the benefit of the ESG index. It is as if, after being burnt by an extreme event, investors redirect flows toward more sustainable assets. This is consistent with some conclusions of <span class="citation"><a href="Bib.html#ref-dyck2019institutional" role="doc-biblioref">Dyck et al.</a> (<a href="Bib.html#ref-dyck2019institutional" role="doc-biblioref">2019</a>)</span>. Nevertheless, zooming out back to Figure <a href="Perf.html#fig:perf">4.2</a> reveals that this effect fades. After 2009, the S&P 500 out-performed until 2019 (see Figure <a href="Perf.html#fig:perf3">4.3</a>), as if the appeal of ESG stocks decreases with the time span after a crash.</p>
<div class="sourceCode" id="cb12"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span class="va">g2</span> <span class="op"><-</span> <span class="va">returns</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/mutate.html">mutate</a></span><span class="op">(</span>year <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/lubridate/man/year.html">year</a></span><span class="op">(</span><span class="va">date</span><span class="op">)</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/tidyr/man/pivot_longer.html">pivot_longer</a></span><span class="op">(</span><span class="op">-</span><span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="va">date</span>, <span class="va">year</span><span class="op">)</span>, names_to <span class="op">=</span> <span class="st">"Index"</span>, values_to <span class="op">=</span> <span class="st">"return"</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/group_by.html">group_by</a></span><span class="op">(</span><span class="va">year</span>, <span class="va">Index</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://r-spatial.github.io/sf/reference/tidyverse.html">summarise</a></span><span class="op">(</span>avg_return <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/mean.html">mean</a></span><span class="op">(</span><span class="va">return</span><span class="op">)</span><span class="op">*</span><span class="fl">252</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggplot.html">ggplot</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/aes.html">aes</a></span><span class="op">(</span>x <span class="op">=</span> <span class="va">year</span>, y <span class="op">=</span> <span class="va">avg_return</span>, fill <span class="op">=</span> <span class="va">Index</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/geom_bar.html">geom_col</a></span><span class="op">(</span>position <span class="op">=</span> <span class="st">"dodge"</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggtheme.html">theme_light</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/scale_manual.html">scale_fill_manual</a></span><span class="op">(</span>values <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"#0DCD64"</span>, <span class="st">"#0D70CD"</span><span class="op">)</span>, labels <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"MSCI ESG"</span>, <span class="st">"S&P 500"</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/theme.html">theme</a></span><span class="op">(</span>legend.position <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="fl">0.45</span>, <span class="fl">0.2</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">xlab</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_blank</a></span><span class="op">(</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">ylab</a></span><span class="op">(</span><span class="st">"Annualized Return"</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/theme.html">theme</a></span><span class="op">(</span>text <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_text</a></span><span class="op">(</span>size <span class="op">=</span> <span class="fl">14</span><span class="op">)</span><span class="op">)</span>
<span class="va">g1</span> <span class="op"><-</span> <span class="va">returns</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/mutate.html">mutate</a></span><span class="op">(</span>year <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/lubridate/man/year.html">year</a></span><span class="op">(</span><span class="va">date</span><span class="op">)</span>,
diff <span class="op">=</span> <span class="va">MSCI_ESG</span><span class="op">-</span><span class="va">SP500</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/select.html">select</a></span><span class="op">(</span><span class="va">year</span>, <span class="va">diff</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/group_by.html">group_by</a></span><span class="op">(</span><span class="va">year</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://r-spatial.github.io/sf/reference/tidyverse.html">summarise</a></span><span class="op">(</span>avg_return <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/mean.html">mean</a></span><span class="op">(</span><span class="va">diff</span><span class="op">)</span><span class="op">*</span><span class="fl">252</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggplot.html">ggplot</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/aes.html">aes</a></span><span class="op">(</span>x <span class="op">=</span> <span class="va">year</span>, y <span class="op">=</span> <span class="va">avg_return</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/geom_bar.html">geom_col</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggtheme.html">theme_light</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">xlab</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_blank</a></span><span class="op">(</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">ylab</a></span><span class="op">(</span><span class="st">"ESG - S&P"</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/theme.html">theme</a></span><span class="op">(</span>text <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_text</a></span><span class="op">(</span>size <span class="op">=</span> <span class="fl">14</span><span class="op">)</span><span class="op">)</span>
<span class="va">g2</span> <span class="op">+</span> <span class="va">g1</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/patchwork/man/plot_layout.html">plot_layout</a></span><span class="op">(</span>heights <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="fl">2</span>, <span class="fl">1</span><span class="op">)</span><span class="op">)</span></code></pre></div>
<div class="figure">
<span style="display:block;" id="fig:perf4"></span>
<img src="ESG_p_files/figure-html/perf4-1.png" alt="Annual returns. In the upper panel, we plot the average daily returns multiplied by 252, on a year-by-year basis. The lower panel shows the difference between the two indices (MSCI ESG minus S&P 500)." width="672"><p class="caption">
FIGURE 4.4: Annual returns. In the upper panel, we plot the average daily returns multiplied by 252, on a year-by-year basis. The lower panel shows the difference between the two indices (MSCI ESG minus S&P 500).
</p>
</div>
<p>To illustrate the shifting relative risk of ESG portfolios, we compute the volatility ratio between the S&P 500 and the ESG index in Figure <a href="Perf.html#fig:vol">4.5</a>. Realized volatilities are computed as the standard deviation of the daily returns over the past 60 trading days. While the S&P 500 is more often the most volatile index, the ESG portfolio does run through some pockets of superior risk, especially between 2016 and 2019.</p>
<div class="sourceCode" id="cb13"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span class="kw"><a href="https://rdrr.io/r/base/library.html">library</a></span><span class="op">(</span><span class="va">RcppRoll</span><span class="op">)</span>
<span class="va">vol</span> <span class="op"><-</span> <span class="va">returns</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/mutate.html">mutate</a></span><span class="op">(</span>MSCI_ESG <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/RcppRoll/man/RcppRoll-exports.html">roll_sd</a></span><span class="op">(</span><span class="va">MSCI_ESG</span>, <span class="fl">60</span>, fill <span class="op">=</span> <span class="cn">NA</span><span class="op">)</span>,
SP500 <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/RcppRoll/man/RcppRoll-exports.html">roll_sd</a></span><span class="op">(</span><span class="va">SP500</span>, <span class="fl">60</span>, fill <span class="op">=</span> <span class="cn">NA</span><span class="op">)</span>,
vol_ratio <span class="op">=</span> <span class="va">SP500</span><span class="op">/</span><span class="va">MSCI_ESG</span>,
col <span class="op">=</span> <span class="va">vol_ratio</span> <span class="op">></span> <span class="fl">1</span>,
vol_SP <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/if_else.html">if_else</a></span><span class="op">(</span><span class="va">col</span>, <span class="va">vol_ratio</span>, <span class="fu"><a href="https://rdrr.io/r/base/numeric.html">as.numeric</a></span><span class="op">(</span><span class="cn">NA</span><span class="op">)</span><span class="op">)</span>,
vol_MS <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/dplyr/man/if_else.html">if_else</a></span><span class="op">(</span><span class="op">!</span><span class="va">col</span>, <span class="va">vol_ratio</span>, <span class="fu"><a href="https://rdrr.io/r/base/numeric.html">as.numeric</a></span><span class="op">(</span><span class="cn">NA</span><span class="op">)</span><span class="op">)</span><span class="op">)</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span>
<span class="fu"><a href="https://rdrr.io/pkg/tidyr/man/pivot_longer.html">pivot_longer</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="va">vol_SP</span>, <span class="va">vol_MS</span><span class="op">)</span>, names_to <span class="op">=</span> <span class="st">"Index"</span>, values_to <span class="op">=</span> <span class="st">"vol"</span><span class="op">)</span>
<span class="va">vol</span> <span class="op"><a href="https://rdrr.io/pkg/magrittr/man/pipe.html">%>%</a></span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggplot.html">ggplot</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/geom_path.html">geom_line</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/aes.html">aes</a></span><span class="op">(</span>x <span class="op">=</span> <span class="va">date</span>, y <span class="op">=</span> <span class="va">vol</span>, color <span class="op">=</span> <span class="va">Index</span><span class="op">)</span>, show.legend <span class="op">=</span> <span class="cn">FALSE</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/ggtheme.html">theme_light</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/theme.html">theme</a></span><span class="op">(</span>text <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_text</a></span><span class="op">(</span>size <span class="op">=</span> <span class="fl">14</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/geom_abline.html">geom_hline</a></span><span class="op">(</span>yintercept <span class="op">=</span> <span class="fl">1</span>, linetype <span class="op">=</span> <span class="st">"twodash"</span>, size <span class="op">=</span> <span class="fl">1.4</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">xlab</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/element.html">element_blank</a></span><span class="op">(</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/labs.html">ylab</a></span><span class="op">(</span><span class="st">"Volatility ratio (SP500/ESG)"</span><span class="op">)</span> <span class="op">+</span> <span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/scale_manual.html">scale_color_manual</a></span><span class="op">(</span>values <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"#0DCD64"</span>, <span class="st">"#0D70CD"</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/annotate.html">annotate</a></span><span class="op">(</span>geom <span class="op">=</span> <span class="st">"text"</span>, x <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/zoo/man/yearmon.html">as.Date</a></span><span class="op">(</span><span class="st">"2012-01-01"</span><span class="op">)</span>, y <span class="op">=</span> <span class="fl">0.85</span>, label<span class="op">=</span><span class="fu"><a href="https://rdrr.io/pkg/latex2exp/man/TeX.html">TeX</a></span><span class="op">(</span><span class="st">"MSCI ESG more volatile"</span><span class="op">)</span>, color<span class="op">=</span><span class="st">"#0DCD64"</span>, size <span class="op">=</span> <span class="fl">5</span><span class="op">)</span> <span class="op">+</span>
<span class="fu"><a href="https://rdrr.io/pkg/ggplot2/man/annotate.html">annotate</a></span><span class="op">(</span>geom <span class="op">=</span> <span class="st">"text"</span>, x <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/pkg/zoo/man/yearmon.html">as.Date</a></span><span class="op">(</span><span class="st">"2014-01-01"</span><span class="op">)</span>, y <span class="op">=</span> <span class="fl">1.25</span>, label<span class="op">=</span><span class="fu"><a href="https://rdrr.io/pkg/latex2exp/man/TeX.html">TeX</a></span><span class="op">(</span><span class="st">"S&P 500 more volatile"</span><span class="op">)</span>, color<span class="op">=</span><span class="st">"#0D70CD"</span>, size <span class="op">=</span> <span class="fl">5</span><span class="op">)</span></code></pre></div>
<div class="figure">
<span style="display:block;" id="fig:vol"></span>
<img src="ESG_p_files/figure-html/vol-1.png" alt="Volatility ratio. We plot the ratio of the S&P 500 volatility divided by the volatility of the MSCI ESG index. Realized volatilities are computed as the standard deviation of the daily returns over the past 60 trading days." width="672"><p class="caption">
FIGURE 4.5: Volatility ratio. We plot the ratio of the S&P 500 volatility divided by the volatility of the MSCI ESG index. Realized volatilities are computed as the standard deviation of the daily returns over the past 60 trading days.
</p>
</div>
</div>
</div>
<div class="chapter-nav">
<div class="prev"><a href="Investors.html"><span class="header-section-number">3</span> Investors and SRI</a></div>
<div class="next"><a href="Quant.html"><span class="header-section-number">5</span> Quantitative portfolio construction with ESG data and criteria</a></div>
</div></main><div class="col-md-3 col-lg-2 d-none d-md-block sidebar sidebar-chapter">
<nav id="toc" data-toggle="toc" aria-label="On this page"><h2>On this page</h2>
<ul class="nav navbar-nav">
<li><a class="nav-link" href="#Perf"><span class="header-section-number">4</span> ESG investing and financial performance</a></li>
<li>
<a class="nav-link" href="#toy-model"><span class="header-section-number">4.1</span> Toy model</a><ul class="nav navbar-nav">
<li><a class="nav-link" href="#theory-assets-agents-equilibrium"><span class="header-section-number">4.1.1</span> Theory: assets, agents, equilibrium</a></li>
<li><a class="nav-link" href="#numerical-example"><span class="header-section-number">4.1.2</span> Numerical example</a></li>
</ul>
</li>
<li><a class="nav-link" href="#sri-improves-performance"><span class="header-section-number">4.2</span> SRI improves performance</a></li>
<li><a class="nav-link" href="#mixed"><span class="header-section-number">4.3</span> SRI does not impact performance</a></li>
<li><a class="nav-link" href="#sri-is-financially-detrimental"><span class="header-section-number">4.4</span> SRI is financially detrimental</a></li>
<li>
<a class="nav-link" href="#it-depends"><span class="header-section-number">4.5</span> It depends</a><ul class="nav navbar-nav">
<li><a class="nav-link" href="#dimensions"><span class="header-section-number">4.5.1</span> Dimensions</a></li>
<li><a class="nav-link" href="#geography"><span class="header-section-number">4.5.2</span> Geography</a></li>
<li><a class="nav-link" href="#time-varying-link"><span class="header-section-number">4.5.3</span> Time-varying link</a></li>
<li><a class="nav-link" href="#industry"><span class="header-section-number">4.5.4</span> Industry</a></li>
<li><a class="nav-link" href="#ownership"><span class="header-section-number">4.5.5</span> Ownership</a></li>
<li><a class="nav-link" href="#nonlinearities"><span class="header-section-number">4.5.6</span> Nonlinearities</a></li>
</ul>
</li>
<li><a class="nav-link" href="#csr-and-risk"><span class="header-section-number">4.6</span> CSR and risk</a></li>
<li><a class="nav-link" href="#other"><span class="header-section-number">4.7</span> ESG and other financial metrics</a></li>
<li><a class="nav-link" href="#empirical-illustration"><span class="header-section-number">4.8</span> Empirical illustration</a></li>
</ul>
<div class="book-extra">
<ul class="list-unstyled">
</ul>
</div>
</nav>
</div>
</div>
</div> <!-- .container -->
<footer class="bg-primary text-light mt-5"><div class="container"><div class="row">
<div class="col-12 col-md-6 mt-3">
<p>"<strong>Perspectives in sustainable equity investing</strong>" was written by Guillaume Coqueret. It was last built on 2022-07-18.</p>
</div>
<div class="col-12 col-md-6 mt-3">
<p>This book was built by the <a class="text-light" href="https://bookdown.org">bookdown</a> R package.</p>
</div>
</div></div>
</footer><!-- dynamically load mathjax for compatibility with self-contained --><script>
(function () {
var script = document.createElement("script");
script.type = "text/javascript";
var src = "true";
if (src === "" || src === "true") src = "https://mathjax.rstudio.com/latest/MathJax.js?config=TeX-MML-AM_CHTML";
if (location.protocol !== "file:")
if (/^https?:/.test(src))
src = src.replace(/^https?:/, '');
script.src = src;
document.getElementsByTagName("head")[0].appendChild(script);
})();
</script><script type="text/x-mathjax-config">const popovers = document.querySelectorAll('a.footnote-ref[data-toggle="popover"]');
for (let popover of popovers) {
const div = document.createElement('div');
div.setAttribute('style', 'position: absolute; top: 0, left:0; width:0, height:0, overflow: hidden; visibility: hidden;');
div.innerHTML = popover.getAttribute('data-content');
var has_math = div.querySelector("span.math");
if (has_math) {
document.body.appendChild(div);
MathJax.Hub.Queue(["Typeset", MathJax.Hub, div]);
MathJax.Hub.Queue(function() {
popover.setAttribute('data-content', div.innerHTML);
document.body.removeChild(div);
})
}
}
</script>
</body>
</html>