4

Investment Treaty Protections

against Conflict-​Related Injuries

A.  Introduction

Historically, the international investment regime was primarily based on the law
of state responsibility for injuries to foreigners under customary international law
and treaties of friendship, commerce, and navigation (FCN). The advent of modern
international investment law traces its roots to the second half of the twentieth
century; in particular, the 1990s when states increasingly began to conclude bilat-
eral and multilateral investment treaties, the number of which has now surpassed
3,000. Investment treaties are entered into between two or more states who under-
take to provide certain reciprocal treatment to investors holding the nationality
of the other contracting state(s). The main objective of the treaties is to promote
foreign investment by enhancing the protections thereof and enabling foreign in-
vestors to seek remedies when treaty provisions have been breached.
Investment treaties contain several provisions that foreign investors can use
to claim damages for losses sustained in the course of armed conflict. Broadly
speaking, conflict-​related claims can be divided in two groups:  first, claims for
losses that investors have suffered as a direct consequence of a forcible action in
the conflict, either by the host state’s organs or by non-​state actors (e.g. destruc-
tion or seizure of the investment facility). Second, claims for losses sustained by
investors indirectly during a conflict or in its immediate aftermath (e.g. as a result
of the ensuing economic and political measures). Accordingly, relevant investment
treaty provisions can also be divided in two sometimes overlapping groups: those
that provide remedies for injuries that are a direct consequence of actions taken in
conflict, such as physical damage and destruction of property (full protection and
security and armed conflict clause), and those that cover losses that are a conse-
quence of measures indirectly related to conflict (fair and equitable treatment and
prohibition of expropriation).
The application of investment treaty provisions in times of armed conflict has
been mostly ignored in investment law scholarship.1 The provisions most relevant

1 For the first relevant contribution on this topic, see C Schreuer, ‘The Protection of Investments in
Armed Conflict’ in F Baetens (ed), Investment Law Within International Law: Integrationist Perspectives
(CUP 2011) 3.

The Protection of Foreign Investment in Times of Armed Conflict. First Edition. Jure Zrilič. © Jure Zrilič 2019.
Published 2019 by Oxford University Press.
 Full Protection and Security  89

for the context of conflict, namely the standard of full protection and security and
the armed conflict clause, have been under-​theorized despite the great deal of in-
consistencies and controversies their application has generated in practice. This
chapter sets out to fill this doctrinal gap by examining the scope and the content of
these provisions, focusing in particular on their problematic aspects, and analyse
the relationship between them. In addition, it also looks into the application of pro-
visions on expropriation, and fair and equitable treatment (FET) in a conflict and
post-​conflict setting.

B.  Full Protection and Security

The provision most commonly invoked by investors who sustain injury due to
a violent interference is the full protection and security (FPS) clause. The treaty
standard has its origin in the rule of customary international law aiming at pro-
tecting the property of aliens through claims of state responsibility and the invo-
cation of diplomatic protection.2 Similar clauses imposing an obligation to protect
aliens and their property existed already in the seventeenth-​and eighteenth-​
century peace treaties entered into by Great Britain,3 and in early US commer-
cial treaties,4 where it was made clear that they encompassed protection against a
range of threats, in particular from attacks during revolutions and war, and from
government confiscations. The clause became a common element of FCN treaties
of the nineteenth and early twentieth centuries and it usually provided for ‘the
most complete’,5 or ‘the fullest measure of ’6 protection and security. During that
volatile time, and in the interwar period, the standard was often invoked before
post-​conflict claims tribunals that helped flesh out its content.7 A variation of the
standard also featured in the US FCN treaties concluded after the Second World
War,8 which led to its adoption in different attempts to regulate the protection of
foreign property in a multilateral international treaty, notably, in the influential

2 See Chapter 2. For a historical overview, see G Foster, ‘Recovering “Protection and Security”: The
Treaty Standard’s Obscure Origins, Forgotten Meaning, and Key Current Significance’ (2012) 45 VJTL
1095, 1116; T Weiler, The Interpretation of International Investment Law (Martinus Nijhoff 2013)
59–​127.
3 See e.g. Treaty of Peace between France and Great Britain, 13 November 1655, Art I, repro-
duced in F Davenport (ed), Two European Treaties Bearing on the History of the United States and Its
Dependencies, 1650–​1697 (Washington Carnegie 1929) 40, 46; Treaty between Great Britain and Tunis,
5 October 1662, renewed by Treaty of 1751, Art VII, 1 Hertslet’s Commercial Treaties 157.
4 See e.g. Treaty of Amity and Commerce, US–​Prussia (1785) Art XVIII; Treaty of Amity, Commerce
and Navigation, US–​Great Britain (1794) Arts II, XIV.
5 See e.g. Treaty of Amity, Commerce and Navigation, US–​Mexico (1931) Art III; Treaty of
Commerce and Navigation, US–​Japan (1894) Art I; Treaty of Friendship, Commerce and Navigation,
US–​Argentine (1853) Arts II, VIII.
6 See e.g. Italy–​Venezuela FCN, in Sambiaggio Case (Italy v Venezuela) (1903) 10 RIAA 499.
7 See Chapter 2 B.
8 K Vandevelde, Bilateral Investment Treaties: History, Policy, and Interpretation (OUP 2010) 244.
90  Investment Treaty Protections

1959 Abs-​Shawcross Convention on Investments Abroad, and in the 1967 OECD

Draft Convention on the Protection of Foreign Property.9
In view of the above, it comes as no surprise that provisions granting protection
and security for foreign investment have become an integral part of most invest-
ment treaties and vary little in their wording.10 Article 4 of the German–​Mali bi-
lateral investment treaty (BIT) is representative of this: ‘Investments by nationals
or companies of either Contracting Party shall enjoy full protection and security
in the territory of the other Contracting Party’.11 Other formulations range from
‘most constant protection and security’,12 and ‘the complete and unconditional
legal protection’,13 to a rather more vaguely worded phrase ‘adequate protection
and security’,14 or simply ‘protection and security’.15 While these differences seem
unremarkable, they can in fact prove important when interpreting the scope of
protection afforded under the clause.
Arbitral tribunals considered the application of the FPS clause in a variety of
situations where investors suffered harm due to physical violence. The clause has
been most frequently invoked in situations of physical violence of low intensity and
limited scale, including the forceful seizure of a hotel by employees,16 the seizure or
usurpation of investment by a third party with the help of government forces,17 so-
cial demonstrations and violent disturbances at the investor’s premises,18 demon-
strations by employees,19 and harassment, seizure, or other action by government
authorities.20 Less frequent has been the application of the clause in situations of
collective violence of high intensity and wider scope. In two cases, investors sus-
tained losses in the midst of widespread riots and lootings,21 while in a few cases

9 ‘Abs-​Shawcross Draft Convention on Investments Abroad’ (1960) 9 J Public L 116, Art I (Abs-​
Shawcross Convention); ‘OECD Draft Convention on the Protection of Foreign Property’ (1968) 7 ILM
117, Art 1(a) (1967 OECD Convention).
10 According to UNCTAD, the FPS provision has been included in 84 per cent of 2,571 mapped
investment treaties. UNCTAD, International Investment Agreements Navigator <https://​
investmentpolicyhub.unctad.org/​IIA/​mappedContent> accessed 18 December 2018.
11 Germany–​Mali BIT (1977) Art 4.
12 Energy Charter Treaty (1994) Art 10; Thailand–​Vietnam BIT (1991) Art 3(2); Japan–​Turkey BIT
(1992) Art 5.
13 Russia–​Ukraine BIT (1998) Art 2.
14 Indonesia–​Yemen BIT (1998) Art 2(2); Indonesia–​Algeria BIT (2000) Art 2(2).
15 US–​Zaire BIT (1984) Art II(4). Some use ‘protection’ only, e.g. China–​Syria BIT (1996) Art 3.
16 Wena Hotels Ltd v Arab Republic of Egypt ICSID Case no ARB/​98/​4, Award, 8 December 2000,
para 84.
17 Amco Asia Corp and Others v The Republic of Indonesia Award, 20 November 1984, 1 ICSID Rep
413; Tatneft v Ukraine UNCITRAL, Award, 29 July 2014, para 428; Joseph Houben v Republic of Burundi
ICSID Case no ARB/​13/​17, Award, 20 May 2013, paras 167, 178; Bernhard von Pezold and Others v
Zimbabwe ICSID Case no ARB/​10/​15, Award, 28 July 2015, paras 585, 597.
18 Tecnicas Medioambientales Tecmed SA v The United Mexican States ICSID Case no ARB(AF)/​00/​2,
Award, 29 May 2003, paras 175–​77; Copper Mesa Mining v Republic of Ecuador, PCA no 2012-​2, Award,
15 March 2016, paras 6.80–​84.
19 Noble Ventures Inc v Romania ICSID Case no ARB/​01/​11, Award, 12 October 2005, para 16.
20 Eureko BV v Poland Ad hoc Arbitration, Partial Award, 19 August 2005, para 236.
21 American Manufacturing & Trading, Inc (AMT) v Republic of Zaire ICSID Case no ARB/​93/​1,
Award, 21 February 1997, para 6.08; Pantechniki SA Contractors & Engineers v The Republic of Albania
ICSID Case no ARB/​07/​21, Award, 30 July 2009, para 82.
 Full Protection and Security  91

claims were filed for losses incurred during a security crisis amounting to a revolu-
tion.22 The application of the FPS in a situation of insurgency and civil war was ad-
dressed prominently in AAPL v Sri Lanka,23 although more instances of invocation
have been reported in pending cases emerging from the civil war in Libya and the
conflict in Crimea.24
While the provision has not attracted as much scholarly attention as other in-
vestment treaty provisions,25 it has given rise to many controversies. Questions
that have yielded conflicting views in scholarship and jurisprudence, and are par-
ticularly relevant to the application of the rule in conflict situations, concern the
scope of the provision and the standard of care required thereunder. Both are ana-
lysed in the next two sections.

1.  Scope of the Protection

With respect to the scope of FPS, two aspects are relevant for the present discus-
sion. The first, which has spawned divisive opinions, concerns the delimitation be-
tween physical and legal protection. The second, overlooked by scholars but no less
controversial, concerns the limits of the FPS obligation with regard to conduct of
state organs.

(a) Physical v legal protection
Tribunals have adopted conflicting views as to the scope of the protection and se-
curity obligation. While some have limited the standard to physical protection
only,26 others have held that the standard extended to legal security (in particular
with respect to quality of the legal system and functioning of the judicial mech-
anisms).27 When a treaty provision is specific enough that it provides only for

22 Ampal v Arab Republic of Egypt ICSID Case no ARB/​12/​11, Decision on Liability and Heads of
Loss, 21 February 2017, para 240; LESI SpA and ASTALDI SpA v People’s Democratic Republic of Algeria
ICSID Case no ARB/​05/​3, Award, 12 November 2008, para 181.
23 Asian Agricultural Products Ltd (AAPL) v Republic of Sri Lanka ICSID Case no ARB/​87/​3, Award,
27 June 1990, para 85(b).
24 See Chapter 1, nn 6–​8.
25 See Foster, ‘Recovering Protection and Security’ (n 2); C Schreuer, ‘Full Protection and Security’
(2010) 1(2) JIDS 6; G Cordero Moss, ‘Full Protection and Security’ in A Reinisch (ed), Standards of
Investment Protection (OUP 2008) 131; H Zeitler, ‘Full Protection and Security’ in S Schill (ed),
International Investment Law and Comparable Public Law (OUP 2010) 184.
26 See e.g. Suez v Argentine Republic ICSID Case no ARB/​03/​17, Decision on Liability, 30 July 2010,
para 173; Saluka Invs BV v Czech Republic UNCITRAL, Partial Award, 17 March 2006, paras 483–​84;
BG Group plc v Republic of Argentina UNCITRAL, Final Award, 24 December 2007, paras 324–​26;
Rumeli Telekom v Kazakhstan ICSID Case no ARB/​05/​16, Award, 29 July 2008, paras 662–​68; Oxus Gold
v Uzbekistan UNCITRAL, Award, 17 December 2015, paras 830–​32; von Pezold (n 17) para 596.
27 See e.g. AES Summit Generation Ltd v Republic of Hungary ICSID Case no ARB/​07/​22, Award, 2
September 2010, para 13.3.2; CME Czech Republic BV v Czech Republic UNCITRAL, Partial Award, 13
September 2001, paras 159–​60, 613; Lauder v Czech Republic UNCITRAL, Final Award, 3 September
2001, para 314; Biwater Gauff Ltd v United Republic of Tanzania ICSID Case no ARB/​05/​22, Award, 24
July 2008, para 729; Azurix Corp v the Argentine Republic ICSID Case no ARB/​01/​12, Award, 14 July
92  Investment Treaty Protections

physical protection,28 or when it explicitly extends beyond it,29 determining the

scope is straightforward. However, even when the provision is drafted in vague
terms (e.g. using qualifiers ‘full’ or ‘constant’, or simply ‘protection and security’),
the practical implications of deciding whether legal security is covered therein may
be less important when there is the possibility of relying on the FET provision,
which is included in most investment treaties and covers the legal aspects of se-
curity.30 Unsurprisingly, many recent tribunals avoided discussing whether the
FPS extends beyond physical safety by deciding allegations pertaining to legal se-
curity under the FET heading only.31
While commentators appear similarly divided on the issue,32 a few have argued
that FPS require a wider reading.33 The argument that this follows from the ap-
plication of interpretive tools codified in the Vienna Convention on the Law of
Treaties (VCLT) is not convincing.34 First, interpretation in line with the ordinary
meaning of the terms as found in a dictionary, and object and purpose of the treaty
is of little assistance, as it renders the meaning so broad and general that FPS easily
replaces all other investment treaty substantive obligations and becomes an om-
nipotent standard embodying the general spirit of a treaty.35 Contextual interpret-
ation brings more clarity, especially the fact that FPS commonly appears in the
same article as a FET standard that primarily includes elements of legal security.
A reading according to which there is a significant substantive overlap between two
distinct standards included in the same provision betrays textual logic, renders the

2006, para 406; Compañía de Aguas del Aconquija SA and Vivendi Universal SA v Argentine Republic
ICSID Case no ARB/​97/​3, Award, 20 August 2007, para 7.4.15.

28 See e.g. Netherlands–​Vietnam BIT (1995) Art 3(2); Albania–​Netherlands BIT (1995) Art 3(2).
29 See e.g. Macedonia–​Ukraine BIT (1998) Art 3(1); Croatia–​Libya BIT (2002) Art 2(5).
30 The tribunal in Suez distinguished between fair and equitable treatment, which covers stability and
legal security, and full protection and security, which covers physical protection only. Suez (n 26) paras
171–​72.
31 See e.g. PSEG Global v Turkey ICSID Case no ARB/​02/​5, Award, 19 January 2007, paras 257–​59;
Copper Mesa (n 18) paras 6.80–​82; Tatneft (n 17) para 429. Some tribunals decided that claims re-
garding legal safety were not justified and thus it was not necessary to discuss the scope of the FPS
provision. See e.g. Rusoro Mining Ltd v Bolivarian Republic of Venezuela ICSID Case no ARB(AF)/​12/​5,
Award, 22 August 2016, para 547; Peter Allard v Barbados PCA Case no 2012-​06, Award, 27 June 2016,
paras 531–​52.
32 For views critical of extensive interpretation, see e.g. C McLachlan, L Shore, and M Weiniger,
International Arbitration:  Substantive Principles (2nd edn, OUP 2017) 335; J Salacuse, The Law of
Investment Treaties (2nd edn, OUP 2015) 236–​38; M Sornarajah, The International Law on Foreign
Investment (4th edn, CUP 2017) 427; Zeitler, ‘Full Protection and Security’ (n 25) 190.
33 Weiler, Interpretation (n 2) 104–​05; Foster, ‘Recovering Protection and Security’ (n 2); Schreuer,
‘Full Protection and Security’ (n 25) 10; A Newcombe and L Paradell, Law and Practice of Investment
Treaties (Kluwer 2009) 314.
34 cf Foster (n 2) 1150 (referring to VCLT Arts 31 and 32).
35 Oxford Dictionary defines ‘to protect’ as ‘to keep safe from harm or injury’, which, taking into
account the commonly understood object and purpose of investment treaties (encouragement of in-
vestment by providing adequate protections), could be construed as protection against any action or
omission that could harm investors. Oxford Dictionary < http://​www.oxforddictionaries.com/​> ac-
cessed 10 May 2018.
 Full Protection and Security  93

inclusion of both standards futile (in contradiction to effet utile interpretation),36

and, as seen above, tends to lead to resolution of relevant claims under only one of
the standards (notably FET).
Some commentators have further argued that the meaning of FPS in modern in-
vestment treaties should be informed by customary international law according to
which protection and security obliges states not only to act with due diligence in pro-
tecting aliens and their property, but also to ‘make available an adequate legal system,
featuring such protections as appropriate remedial mechanisms, due process, and a
right to compensation for expropriation’.37 They buttressed their understanding of the
customary duty by referring to historical writings, arbitral practice, and early FCN
treaties where notions of ‘protection’ and ‘security’ were sometimes used broadly and
all-​encompassing, reflecting the international minimum standard of treatment.38
This view is oblivious to the temporal element of interpretation and the fact that the
linguistic usage of a term ‘protection’ during that era differed from its usage in post-​
Second World War FCN treaties and modern investment treaties in which new legal
standards (such as FET) became commonplace.39 Using that obsolete understanding
to determine the meaning of modern FPS provisions would be contrary to inter-
pretation in good faith.40 Moreover, it has become widely accepted in international
jurisprudence and treaty practice that FPS and FET both derive from customary
international minimum standards, whereby the former covers physical protection
only, while the latter concerns certain elements of legal protection (including denial
of justice).41
Since injuries to investors in volatile times are often caused by the use of force,
the application of the standard in the context of conflict will normally not be con-
troversial. In fact, a few tribunals have held that FPS is meant to cover exactly the

36 McLachlan et al, Substantive Principles (n 32) 335.

37 Foster, ‘Recovering Protection and Security’ (n 2) 1130; Weiler, Interpretation (n 2) 61.
38 Foster (n 2) 1116–​49; Weiler (n 2) 127.
39 The FET concept was introduced to the US FCN treaties after the Second World War. See K
Vandevelde, The First Bilateral Investment Treaties: U.S. Postwar Friendship, Commerce, and Navigation
Treaties (OUP 2017) 399.
40 VCLT Art 31(1). Regarding temporal interpretation, see Dispute Regarding Navigational and
Related Rights (Costa Rica v Nicaragua) (Judgment) [2009] ICJ Rep 213, 242–​43, paras 63–​66 (noting
that when the parties used generic terms in a treaty, they ‘must be presumed, as a general rule, to have
intended those terms to have an evolving meaning’, especially when a treaty has been entered into
for a long period). See also H Waldock, ‘Sixth Report on the Law of Treaties’ in ILC, Yearbook of the
International Law Commission, Vol II, 1966, UN Doc A/​CN.4/​186/​Add.1–​7, 96, para 7.
41 See e.g. Saluka (n 26)  para 484; BG Group (n 26)  para 324; Enron Corp v Argentine Republic
ICSID Case no ARB/​01/​3, Award, 22 May 2007, para 286. See also 2012 US Model BIT, Art 5(2); North
American Free Trade Agreement (1994) Art 1105 (NAFTA). A number of most recently concluded in-
vestment treaties have expressly limited FPS to police protection as reflected by customary international
law, see Canada–​EU Comprehensive Economic and Trade Agreement (2016) Art 8.10(5); Canada–​
Republic of Guinea BIT (2015) Art 6; Rwanda–​Morocco (2016) Art 2(2); Israel–​Japan BIT (2017) Art
4; Rwanda–​United Arab Emirates BIT (2017) Art 4(3); ASEAN–​Hong Kong, China SAR Investment
Agreement (2017) Art 5; Republic of Korea–​Republics of Central America FTA (2018) Art 9.5.
94  Investment Treaty Protections

kinds of losses that foreign investors may suffer during violent conflict situations.42
Thus, the tribunal in Saluka v Czech Republic stated that ‘the standard applies es-
sentially when the foreign investment has been affected by civil strife and physical
violence’.43 Similarly, the tribunal in Eastern Sugar v Czech Republic held that the
standard provides protection against ‘mobs, insurgents, rented thugs and others
engaged in physical violence’.44
When the scope of FPS provisions is not expressly extended to legal safety, it
may become important to define what constitutes physical interference. The tri-
bunal in Siemens v Argentina pointed to this issue by raising a question of ‘how
the physical security of an intangible asset would be achieved’, if FPS was inter-
preted restrictively.45 The tribunal concluded that the only answer was to interpret
FPS broadly in order to encompass legal security also.46 For the reasons described
above, this solution is not satisfying. Importantly, the question requires inter-
pretive re-​orientation:  what matters is not whether FPS provides legal security,
but rather the scope of physical protection. The Siemens tribunal appears to have
held that only tangible assets could be accorded physical protection, necessitating a
broader reading of FPS.47 According to this view, physical protection denotes pro-
tection against forceful interference that may cause physical damage to the person
or property of an investor. This interpretation can leave an important category of
investment unprotected, since consequences of interference with intangible assets
do not necessarily manifest in physical damage. The problem was further raised in
Saluka case, where the tribunal accepted that the guarantee of physical protection
would apply with respect to police searches of an investor’s premises and seizure
of documents, but left open the question as to whether the scope of the guarantee
would also extend to a national authority’s suspension of the trading of shares or
a police order prohibiting the transfer of shares.48 Since the freezing of assets of
enemy aliens is common in situations of conflict, the question of whether they con-
stitute a breach of the FPS standard could likely be raised in conflict-​related cases.
Another topical example is harm caused by cyber-​attacks. Does the FPS clause
offer protection in case of attacks against investors’ computer networks and digital
infrastructure? If a cyber-​attack causes physical damage (e.g. by creating ham-
mering in oil pipelines which causes them to break),49 the answer would appear
a straightforward yes. However, many cyber-​attacks do not produce physical

42 Saluka (n 26) para 483; Rumeli (n 26) para 668; Eastern Sugar v Czech Republic SCC Case no 088/​
2004, Partial Award, 27 March 2007, para 203; PSEG (n 31) paras 258–​259; Suez (n 26) para 179.
43 Saluka (n 26) para 483.
44 Eastern Sugar (n 42) para 203.
45 Siemens AG v Argentine Republic ICSID Case no ARB/​02/​8, Award, 17 January 2007, para 303.
46 ibid para 308.
47 Similarly, Azurix (n 27) para 408; Vivendi (n 27) para 7.4.15.
48 Saluka (n 26) paras 486–​93.
49 For example, see recent cyber-​attacks against US gas pipelines. P Muncaster, ‘US Gas Pipelines
Hit by Cyber-​Attack’ (Info-​Security Magazine, 4 April 2018) <https://​www.infosecurity-​magazine.com/​
news/​us-​gas-​pipelines-​hit-​by-​cyberattack/​> accessed 10 December  2018.
 Full Protection and Security  95

damage (e.g. causing the computer websites to crash or manipulating important

information located on computer servers), thus the question of whether they are
captured under the guarantee of physical protection could become vital. The more
convincing position would seem to be that the protection is not limited only to
traditional threats of violence capable of generating physical harm. What should
be determinative is not the nature of the outcome of the breach but rather the na-
ture of the conduct constituting it. Such an interpretation is sufficiently flexible so
as to cover all categories of investment and adaptable to new forms of threat such
as computer network attacks, while still excluding the breaches typically associated
with legal security (concerning quality of the legal and judicial system). Following
this view, FPS clauses would provide protection against any forceful intervention
with the ability to generate damage, including interferences that are coercive in
nature and do not result in physical damage but rather impair the functionality of
investment.50 Whether the threshold of application has been reached could only be
determined on a case-​by-​case analysis of the nature and degree of interference, its
severity, and consequences.

(b) Protection against interference of state organs

Under the FPS standard, the host state is obliged to provide protection and se-
curity against the actions of both its own organs and third parties.51 In the latter
case, which is discussed in detail in the next section, the provision imposes an
obligation to protect investors against harmful activities carried out by third
persons, that is private actors as well as international subjects such as other
states or insurrectional movements. The duty to protect encompasses the obli-
gation to prevent violent acts of non-​state actors as well as the obligation to ap-
prehend and punish those responsible for the harm.52 In these situations, state

50 This interpretation is in line with recent developments in the field of cyber warfare, which re-
flect a preference for a flexible definition of ‘damage’ caused by cyber-​attacks, thus enabling applica-
tion of the jus in bello framework. See ‘Tallinn Manual on the International Law Applicable to Cyber
Warfare: Prepared by the International Group of Experts at the Invitation of the NATO Cooperative
Cyber Defence Centre of Excellence’ (CUP 2013)  Rule 5, comment 5 (noting that the cyber op-
eration does not need to result in physical damage to objects or injuries to individuals; it is enough
that it produces a ‘negative effect’); ICRC, ‘Cyber Warfare and International Humanitarian Law: The
ICRC’s Position’ 2  <https://​www.icrc.org/​en/​doc/​assets/​files/​2013/​130621-​cyber-​warfare-​q-​and-​a-​
eng.pdf> accessed 18 December 2018; M Schmitt, ‘Computer Network Attack and the Use of Force in
International Law: Thoughts on a Normative Framework’ (1999) 37(3) Colum J Trans L 885; T Morth,
‘Considering Our Position: Viewing Information Warfare as a Use of Force Prohibited by Article 2(4) of
the U.N. Charter’ (1998) 30(2/​3) Case W Res J Intl L 567.
51 Schreuer, ‘Full Protection and Security’ (n 25) 2–​5; Vandevelde, History (n 8) 243. For divergent
views among arbitral tribunals, see nn 206–​07.
52 See Wena (n 16) paras 82, 84; Frontier Petroleum Services v Czech Republic UNCITRAL, Award,
12 November 2010, para 423; Parkerings-​Compagniet AS v Republic of Lithuania, ICSID Case no ARB/​
05/​8, Award, 11 September 2007, para 355. Pissilo-​Mazzeschi argued that, once the harmful event has
occurred, the FPS covers only the duty to investigate such an event, and pursue and apprehend those
responsible for harm. Further administrative and judicial steps (trial and execution of sentence) fall
within the ‘denial of justice’ and are not measured by the state’s due diligence. See R Pissilo-​Mazzeschi,
96  Investment Treaty Protections

liability can emerge from the failure to exercise due diligence in protecting
investors.53
In contrast, when it comes to actions of its own organs, the impairment to the
investor is caused directly by state organs or through their contribution.54 The
responsibility emerges from the breach of the obligation to refrain from such
wrongful interference and is attached to the host state through the attribution of
the acts of its organs.55 Despite the seemingly clear-​cut distinction between private
violence and the violence of state organs, two questions in particular have yielded
unsatisfactory responses in doctrine and arbitral jurisprudence. The first concerns
the use of the due diligence rule in situations when harm has been caused by state
organs. The second deals with the limitation of the FPS provision in regard to a
state’s failure to refrain from interference.

(i) Due diligence in lieu of attribution

With respect to the first question, scholars have overwhelmingly argued that the
duty of due diligence never applies in situations when harm was caused by state
organs themselves.56 According to this view, the conduct of state organs does not
need to be measured against the due diligence standard since it is already directly
attributed to the state. While this may be true in most cases, there appears to be an
exception to this rule, highlighted in the AAPL case, in which the tribunal found
Sri Lanka liable for failing to take precautions in the planning of its military oper-
ation.57 Following this position, the host state has a due diligence duty to protect
investors from the harm that might result from the actions of its own organs, when
the latter are not prima facie wrongful and involve a heightened risk for incidental
injuries. This would be the case if the state undertook a harmful or inherently risky
but justifiable activity that included diligent planning and measures to avert or
minimize the harm, for example, military targeting. While there is no general obli-
gation to abstain from legitimate targeting operations, diligence is required in the

‘The Due Diligence Rule and the Nature of the International Responsibility of States’ (1992) 35 German
YB Intl L 9, 29–​30. See Sewell Case (US v Mexico) (1930) 4 RIAA 626, 632; Chase Case (US v Mexico)
(1928) 4 RIAA 337, 339; Massey Case (US v Mexico) (1927) 4 RIAA 155, 162; Kennedy Case (US v
Mexico) (1927) 4 RIAA 194, 196–​97; Way Case (US v Mexico) (1928) 4 RIAA 391; Minnie East Case (US
v Mexico) (1930) 4 RIAA 646; Morton Case (US v Mexico) (1929) 4 RIAA 428.

53 Zeitler, ‘Full Protection and Security’ (n 25) 187. See also US Diplomatic and Consular Staff in
Teheran (US v Iran) [1980] ICJ Rep 3.
54 See e.g. Biwater (n 27) para 730; Frontier (n 52) para 261; Parkerings (n 52) para 355.
55 See the general rules on attribution in Articles on Responsibility of States for Internationally
Wrongful Acts, Arts 4–​11.
56 Pissilo-​Mazzeschi, ‘Due Diligence’ (n 52) 23–​25; Zeitler, ‘Full Protection and Security’ (n 25) 191;
E De Brabandere, ‘Host States’ Due Diligence Obligations in International Investment Law’ (2015)
42(2) Syracuse J Intl & Com 320, 333–​34, 337; R Dolzer and C Schreuer, Principles of International
Investment Law (2nd edn, OUP 2012) 162.
57 AAPL (n 23). For a discussion of this aspect of due diligence, see Section 4 C.4.b and, in particular,
Chapter 6 C.2.
 Full Protection and Security  97

planning and execution thereof with the view to avoid or minimize the damage.
Finding the state responsible would thus depend on the assessment of whether the
state organs were diligent or not, that is on whether they adequately evaluated the
risks and took necessary measures for minimizing thereof to the extent possible.
Moreover, as will be discussed in Chapter 6,58 a higher degree of care is required
from a state in carrying out such hazardous activities than in providing protection
against harmful activities of non-​state actors.
The risk-​based approach that uses due diligence rather than attribution of an
immediately harmful act for ascertaining state responsibility can be justified on
practical, policy, and legal grounds. First, the attribution of losses emerging from
hazardous or harmful operations, such as military attacks, can be challenging due
to inherent difficulties in collecting evidence and establishing causality. Due dili-
gence fills this vacuum by shifting the responsibility to the party that was more in
control of the risk and was thus in a position to take certain precautions. Second,
such interpretation incentivizes parties to engage better in ex ante risk assess-
ments with regard to potential incidental losses, and in this way facilitates pru-
dence. Third, it also enables tribunals to avoid making potentially controversial
pronouncements on the legitimacy and legality of state actions while still finding it
liable on a different, less contentious, ground.
Legal support for this understanding of the FPS can be found in some historical
cases, in which it was held that state responsibility could be generated by the lack
of due diligence in the execution of the lawful measures by state organs,59 as well as
in the 1967 OECD Convention, which in its comments stipulated that ‘most con-
stant protection and security’ includes the obligation ‘to exercise due diligence as
regards actions of public authorities’, rather than just private actors.60 Lastly, such
interpretation is in line with the application of due diligence in international hu-
manitarian law,61 international human rights law,62 international environmental

58 Chapter 6 C.
59 See e.g. In re Rizzo (1955) 22 ILR 317, 322; Ousset (1945) 22 ILR 312, 314, cited in J Crawford,
Brownlie’s Principles of International Law (8th edn, OUP 2012) 559.
60 1967 OECD Convention, Notes and Comments to Article 1 (n 9) 9.
61 See Additional Protocol I to Geneva Conventions, Art 57, and Chapter 2 C.1.a. More generally,
the 2016 ICRC Commentary appears to confirm that a state must exercise due diligence with respect
to its own organs (soldiers). ICRC Commentary to Geneva Convention I (2016) Art 12, paras 1360–​
61 (specifying that the obligation to protect wounded and sick requires the exercise of due diligence
in preventing harm caused by, among others, party’s own soldiers). See also Commentary to Geneva
Convention I (2016) Art 15, para 1499; Commentary to Geneva Convention II (2017) Art 12, para
1407; Hague Convention VIII relative to Laying of Automatic Submarine Contact Mines (1907) Arts
3 and 4. See also Armed Activities on the Territory of Congo (DRC v Uganda) [2005] ICJ Rep 252 (para
246), 253 (para 250) (finding Uganda responsible for breaching the due diligence duty with respect to
actions of its own military forces).
62 See Study Group on Due Diligence in International Law, ‘First Report’ (International Law
Association, London 2014) 22. However, as shown in Chapter 2, human rights courts have discussed
the state’s duty to take precautions and exercise requisite care in forceful operations not as part of a posi-
tive obligation of due diligence, but rather as part of the proportionality analysis in the assessment of a
negative obligation to refrain (an approach discussed in the next section).
98  Investment Treaty Protections

law,63 the developments in the regulation of cyberspace,64 as well as the emerging

scholarship on risk in international law.65

(ii) The limits of obligation to refrain

In all other instances of harm caused directly by state organs, the due diligence
rule does not play any role. The question, largely ignored in investment treaty jur-
isprudence and scholarship alike, is what then are the limits of the FPS with re-
spect to acts of state organs interfering with foreign investment? Does any failure
to abstain from physical interference constitute a breach of FPS? If not, what are
the limitations and how can they be legally conceptualized? As will be shown in
the following sections, some investment treaties expressly provide that a physical
interference with an investment can be justified on the ground of security reasons
or due to necessity inherent in situations of armed conflict.66 Furthermore, some
investment treaties specifically limit a host state’s obligation under FPS with a ref-
erence to its domestic laws.67 Even in the absence of such exceptions and limita-
tions, however, states still have discretion to exercise certain powers with the view
to protect their public order and security interests. This has been confirmed in ar-
bitral jurisprudence.
In Saluka v Czech Republic, the tribunal held that the suspension of trading with
investor shares was reasonable and justifiable by the state’s legitimate concerns re-
lating to securities markets and consequently no breach of FPS could be found.68
In Biwater v Tanzania, the tribunal found that the removal of investor manage-
ment from the offices, their subsequent deportation from the country, and seizure
of premises amounted to breach of the FPS clause.69 In its defence, the government,
among others, argued that no more force had been used than lawful and necessary
for carrying out orders of domestic organs, and that the measures were motivated
by legitimate concerns about the safety of the water and sewerage system.70 While

63 See e.g. Certain Activities Carried Out By Nicaragua in the Border Area (Costa Rica v Nicaragua)
and Construction of a Road in Costa Rica Along the San Juan River (Nicaragua v Costa Rica) (Judgment)
[2015] ICJ Rep 665, 706 (para 104), 720–​21 (paras 153–​57); Pulp Mills on the River Uruguay (Argentina v
Uruguay) (Judgment) [2010] ICJ Rep 14, 79 (para 197), 83 (para 204); Responsibilities and Obligations of
States Sponsoring Persons and Entities with Respect to Activities in the Area, ITLOS Case no 17 (Advisory
Opinion, 1 February 2011) para 131.
64 Tallinn Manual (n 50)  Rule 5, comments 1, 3; M Schmitt, ‘In Defense of Due Diligence in
Cyberspace’ (2015) 125 Yale LJ Forum 68, 70; K Bannelier-​Christakis, ‘Cyber Diligence:  A Low-​
Intensity Due Diligence Principle for Low-​Intensity Cyber Operations? (2014) 14 Baltic YB Intl L 1, 4, 8.
65 See e.g. S Townley ‘The Rise of Risk in International Law’ (2016) 18(2) Chicago J Intl L 594; M
Ambrus et al (eds), Risk and the Regulation of Uncertainty in International Law (OUP 2017).
66 For analysis of advanced armed conflict clauses and security exceptions, see Section 4 C.3. and
Chapter 5 B, respectively.
67 See e.g. Australia–​Egypt BIT (2001) Art 3(3); Ethiopia–​Sweden BIT (2004) Art 2(4). According
to UNCTAD Navigator, such references have been found in less than 7 per cent of investment treaties.
UNCTAD Navigator (n 10).
68 Saluka (n 26) paras 490, 505.
69 Biwater (n 27) para 731.
70 ibid para 721.
 Full Protection and Security  99

the tribunal failed to address these arguments in any meaningful way, its ultimate
description of the contested state measures as ‘unnecessary and abusive’ implies
that a different outcome not resulting in a breach of FPS would have been possible,
had the factual circumstances been different.71 More pertinent to the context of
armed conflict, in AAPL v Sri Lanka, it was uncontested that a state had a sovereign
right to use force to prevent or suppress resurrection and regain control over its ter-
ritory, and that such use of force could negatively affect foreign investors without
giving rise to state responsibility.72
While the reasoning of these awards leaves much to be desired, what is common to
them is that they recognize that in certain circumstances states have a right to phys-
ically interfere with foreign investment without violating the FPS clause. To argue
otherwise would mean that one of the state’s most important sovereign prerogatives,
that is, its monopoly on the use of force, would be significantly constrained. While the
tribunals have failed to legally conceptualize this limitation to FPS in a clear manner,
two approaches can be discerned. According to the first approach, which seemed to be
followed by the Saluka and Biwater tribunals, a state can interfere with the investment
without violating FPS as long as its conduct is reasonable. In other words, a state will
not be liable if there is a reasonable connection between a state’s actions (e.g. suspen-
sion of shares, removal of management) and a legitimate public concern pursued by
those actions. The use of reasonableness to justify the departure from a state’s obliga-
tion to refrain from interfering with investment must be distinguished from the duty
to exercise reasonable care (i.e. due diligence) in protecting investors, which is dis-
cussed in the next section. In the latter case, reasonableness is used to bring some flexi-
bility when measuring a state’s effort to protect investment. In contrast, the obligation
to refrain from interfering is not an obligation of effort—​reasonableness merely estab-
lishes a connection between a state’s act and a state’s legitimate interest, thus creating
a bar against ‘unnecessary and abusive’ actions. The downside of this approach lies in
the vague nature of the principle of reasonableness, which plays an important part in
loosening of the due diligence obligation, but provides little guidance in the context of
the state’s interference with investment.
The second approach, which is echoed (but not expressly endorsed) in the rea-
soning of the AAPL award, seeks legal justification for such interference in the
doctrine of police powers. The police powers doctrine concerns the exercise of a
state’s sovereign right to constrain certain individual rights and economic inter-
ests (e.g. right to property) with the view to protect and promote public policy
objectives such as public safety, security, and health.73 While originating from US

71 ibid para 731.
72 AAPL, Award (n 23) para 85B; AAPL, Dissenting Opinion (n 23) 592.
73 See the definition in the Black’s Law Dictionary (6th edn, West Publishing Co 1990) which articu-
lates the state’s police powers as the ‘power of the State to place restraints on the personal freedom and
property rights of persons for the protection of the public safety, health, and morals or the promotion of
the public convenience and general prosperity’.
100  Investment Treaty Protections

jurisprudence and legal doctrine,74 the concept gradually found its way into inter-
national law through the draft codifications of the international responsibility of
states,75 and has become increasingly acknowledged in modern investment case
law,76 scholarship,77 and even some investment treaties,78 often as a reflection
of customary international law.79 Since the doctrine has been mostly analysed
with respect to indirect expropriation as a justification for governments’ non-​
discriminatory regulatory measures, the view has emerged that it cannot be ap-
plied as a defence in regard to other investment treaty standards, such as FPS, or as
a justification for measures targeting a specific investor (rather than general regu-
latory measures).80
This position is not legally persuasive. First, many investment tribunals had no
problem referencing or applying ‘police powers’ with respect to measures targeting
individual investors as long as they were not discriminatory.81 Second, there ap-
pears to be no compelling reason to limit the doctrine only to indirect expropri-
ation, in particular if one accepts the view that it reflects customary international
law.82 This is supported by historical cases in which the doctrine was applied to jus-
tify destruction of property in emergency situations.83 More controversially, at the
turn of the twentieth century, an ‘internationalized’ version of police power was

74 Brown v Maryland, 25 U.S. (12 Wheat) 419, 442–​43 (1827). See S Legarre, ‘The Historical
Background of the Police Power’ (2007) 9 U Pa J Const L 745.
75 FV Garcia Amador, ‘Fourth Report on State Responsibility’ in Yearbook of the International Law
Commission, 1959, Vol II, UN Doc A/​CN.4/​SER.A/​1959/​Add.1, 11–​12, paras 43, 46, 133; Harvard Draft
Convention on the International Responsibility of States for Injuries to Aliens (1961) Art 10(5) in ILC,
Yearbook of the International Law Commission, 1969, Vol II, UN Doc A/​CN.4/​SER.A/​1969/​Add.1, 142
(1961 Harvard Draft); American Law Institute, Restatement (Second) of the Law of Foreign Relations of
the United States, 1965, s 197(1)(a); American Law Institute, Restatement (Third) of the Law of Foreign
Relations of the United States, 1986, s 712.
76 The doctrine has been heeded by several investment tribunals, e.g. Marvin Feldman v Mexico
ICSID Case no ARB(AF)/​99/​1, Award, 16 December 2002, para 103; Saluka (n 26) para 255; Lauder
(n 27) para 198; Tecmed (n 18) para 115; Chemtura Corp v Canada, UNCITRAL, Award, 2 August
2010, para 266; Quiborax SA and Non Metallic Minerals v Bolivia ICSID Case no ARB/​06/​2, Award,
16 September 2015, para 202; Renée Rose Levy de Levi v Peru ICSID Case no ARB/​10/​17, Award, 26
February 2014, para 476; Philip Morris et al v Uruguay ICSID Case no ARB/​10/​7, Award, 8 July 2016,
para 295.
77 See e.g. Newcombe and Paradell L, Law and Practice (n 33) 358; A Pellet, ‘Police Powers or the
State’s Right to Regulate’ in M Kinnear et al (eds), Building International Investment Law—​The First 50
Years of ICSID (Kluwer Law International 2016) 449, 451; J Viñuales, ‘Sovereignty in Foreign Investment
Law’ in Z Douglas et al (eds), Foundations of International Investment Law (OUP 2014) 329, 344; C Titi,
‘Police Powers Doctrine and International Investment Law’ in A Gattini et al (eds), General Principles of
Law and International Investment Arbitration (Brill 2018) 323.
78 See e.g. COMESA Common Investment Area Agreement (2007) Art 20(08).
79 Saluka (n 26) para 262; Philip Morris (n 76) para 301. See also Newcombe and Paradell, Law and
Practice (n 33) 358; Pellet, ‘Police Powers’ (n 77) 449; Viñuales, ‘Sovereignty’ (n 77) 329, 344.
80 See Suez (n 26) para 148. See also Titi, ‘Police Powers’ (n 77); Pellet, ‘Police Powers’ (n 77) 457.
81 Viñuales, ‘Sovereignty’ (n 77) 333–​35 (explaining how tribunals in Tecmed, Chemtura, and Saluka
recognized that the doctrine could be applied to targeted measures).
82 Viñuales, ibid 332–​34, 344.
83 See e.g. Dickson Car Wheel Co (United States v Mexico) (1931) 4 RIAA 669, 681–​82; Bischoff Case
(1903) 10 RIAA 420. For more authorities, see M Paparinskis, The International Minimum Standard and
FET (OUP 2013) 224, n 50.
 Full Protection and Security  101

used by the US to safeguard America’s safety and protect US investment abroad

by means of military interventions in Latin America.84 Thus in 1904, Theodore
Roosevelt famously asserted that the US may have to resort ‘to the exercise of an
international police power’ and thus intervene in a state that fails to maintain order
and pay its obligations.85 While today such invocation would undoubtedly consti-
tute abuse of the police powers doctrine, these examples show how historically it
was interlinked with the notions of security and safety. Furthermore, in US con-
stitutional practice as well as modern investment treaty practice, police power has
been associated with the right to take measures for protecting public order against
riots, insurrections, and other types of conflict.86
Support for the application of police powers to FPS can be also found in soft-​law
documents which investment tribunals typically cite when applying the concept.87 For
example, the 1961 Harvard Draft states that ‘deliberate destruction of or damage to the
property of an alien’ and ‘uncompensated taking of property of an alien or a depriv-
ation of the use or enjoyment of property of an alien which results from . . . [the State’s
action] in the maintenance of public order . . . or from the valid exercise of belligerent
rights . . .’ is not wrongful.88 Accordingly, the doctrine applies beyond indirect expro-
priation as to encompass other measures, general or specific, that may deprive the in-
vestor of its property (including by means of confiscation or destruction thereof). The
express reference to maintenance of public order and exercise of belligerents’ rights
leaves no doubt as to the appropriateness of applying the doctrine in the context of
different types of conflict.
Finally, several investment tribunals have evaluated the exercise of state’s police
powers by using a proportionality analysis.89 While ‘police powers’ create a zone
within which a state can pursue certain actions, proportionality sets the borders
of this zone. This is evocative of the international humanitarian law (IHL) frame-
work, in which military necessity in a similar manner permits belligerents the

84 J von Bernstorff, ‘The Use of Force in International Law before World War I: On Imperial Ordering
and the Ontology of the Nation-​State’ (2018) 29(1) EJIL 233, 250.
85 T Roosevelt, Annual Message to Congress (6 December 1904) <ourdocuments.gov/​doc.php?flash=
true&doc=56&page=transcript> accessed 18 December 2018.
86 Some US BITs state that public order covers ‘measures taken pursuant to a Party’s police power to
ensure public health and safety’. See letters of submittal of the US–​Jamaica BIT; US–​Lithuania BIT; US–​
Mongolia BIT. See also K Gudgeon, ‘United States Bilateral Investment Treaties: Comments on Their
Origin, Purposes, and General Treatment Standards’ (1986) 4(1) Berkeley J Intl L 105, 121.
87 See n 75.
88 1961 Harvard Draft, Arts 9 and 10(5).
89 As will be argued in Chapter  6 C.3, the AAPL award reflects such application of proportion-
ality as part of the assessment of a state’s forceful conduct. Generally, see Azurix (n 27) paras 311–​12;
Tecmed (n 18) para 122; Burlington Resources Inc v Ecuador ICSID Case no ARB/​08/​5, Decision on
Liability, 14 December 2012, para 504; Philip Morris (n 76) paras 295, 305; Bear Creek Mining Corp
v Peru ICSID Case no ARB/​14/​21, Award, 30 November 2017, paras 453, 458 (and other decisions
cited in n 604 in the Bear Creek award). See also C Henckels, Proportionality and Deference in Investor-​
State Arbitration: Balancing Investment Protection and Regulatory Autonomy (CUP 2015) 83–​171; G
Bücheler, Proportionality in Investor-​State Arbitration (OUP 2015) 129.
102  Investment Treaty Protections

exercise of certain rights, the limits of which are determined by proportionality.90

Thus, in both legal systems, proportionality has been used as a legal constraint to
state conduct otherwise permitted in certain exceptional circumstances.
As a rule of customary international law, the application of ‘police powers’ does
not necessarily depend on its express inclusion in an investment treaty.91 It can be
applied either directly or as a principle guiding the interpretation of the investment
treaty standard.92 Despite the lack of express reference to ‘police powers’ in con-
nection with the application of FPS in investment treaties and modern case law, it
is safe to conclude that FPS does not impose an absolute obligation to refrain from
interference with foreign investment. Which of the above-​described approaches
will prevail in practice, however, is yet to be seen.

2.  The Standard of Due Diligence

There is a general consensus in modern arbitral jurisprudence that the guarantee of

FPS does not create strict or absolute liability for harm caused by private persons,
but rather provides a general obligation for the host state to exercise due diligence
in the protection of foreign investment.93 In other words, measures that states take
to protect investors against non-​state actors are not evaluated based on whether
the investment was effectively protected in the end, but rather on the basis of the
state’s conduct,94 that is that the host state acted diligently and took the measures
of vigilance necessary in order to protect investors from forcible interference. This
understanding reflects the customary rule discussed above and has been con-
firmed by the International Court of Justice (ICJ).95 This does not mean that the ap-
plication of the standard is unproblematic. The tension in the perspectives between
developed and developing countries at the end of the nineteenth century as to the
degree of liability of a host state has subsisted and spilled over into a debate on what
exactly is the content of due diligence in the FPS clause. The central question is
whether the diligent conduct should be measured against an objective standard or
against the background of the local conditions of the host state.

90 See Chapter 2 C.1.a.
91 However, for the avoidance of doubt, it would be advisable to include a reference to ‘police powers’
in future investment treaties (without limiting its application to indirect expropriation).
92 Saluka (n 26) paras 62, 82, 136, 270–​76; Too v Greater Modesto Insurance Associates Award, 29
December 1989, 23 IUSCT Rep 378; Philip Morris (n 76) paras 287–​350.
93 See e.g. AAPL (n 23) paras 53, 85B; Saluka (n 26) para 484; Tecmed (n 18) para 177; Frontier (n
52) para 270; Lauder (n 27) para 308; Wena (n 16) para 84; El Paso Energy v Argentina ICSID Case no
ARB/​03/​15, Award, 31 October 2011, para 523; Allard (n 31) paras 543–​44; Houben (n 17) para 161; von
Pezold (n 17) para 596; Copper Mesa (n 18) para 6.81; Ampal (n 22) para 241.
94 C Economides, ‘Content of the Obligation: Obligations of Means and Obligations of Result’ in J
Crawford et al (eds), The Law of International Responsibility (OUP 2010) 371, 378 (arguing that the obli-
gation to protect foreigners is an obligation of conduct).
95 Elettronica Sicula SPA (US v Italy) (Merits) [1989] ICJ Rep 15, 65, para 108.
 Full Protection and Security  103

A number of commentators have observed that a due diligence rule reflected a

sliding scale of state liability, ranging from subjective to objective standards.96 If
due diligence is seen as a subjective standard, the state’s conduct is measured solely
on the basis of its existing capabilities and expectations in a given situation. The
standard reflects the outdated theory of the absolute non-​responsibility of states
for losses incurred by aliens in situations of conflict, and has not been espoused by
modern investment tribunals.
On the other hand, according to the objective standard, whether the host state
fulfilled its due diligence obligation is tested against the conduct of a modern, rea-
sonably organized government under similar circumstances.97 All states have to
adhere to the same standard, regardless of their individual circumstances. This is
the approach followed by the AMT tribunal that explicitly recognized the standard
of vigilance as an ‘objective obligation’ and did not take Zaire’s situation into ac-
count when considering whether it had failed to protect the investor.98 Similarly,
some other arbitrators leaned towards the objective nature of due diligence by
linking it to the ‘parameters inherent in a democratic state’,99 ‘the canons of good
governance’, and a ‘reasonably well-​organized modern State’.100 Some have sug-
gested that the tribunal in AAPL v Sri Lanka also heeded this approach.101 The
more accurate view, however, appears to be that the AAPL tribunal departed from
the pure objective standard by limiting the host state’s duty to prevent the harm to
what could ‘be reasonably expected’ in a given situation.102 Reasonableness brings
an element of subjectivity to the due diligence standard as its content will vary
depending on the local situation of the host state.
Similar references to what is ‘reasonable under the circumstances’ when as-
sessing the host state’s due diligence obligation have been made in some other re-
cent cases.103 While reasonableness is a vague term and its content will depend
on the context of each individual case, it arguably provides certain leeway to ac-
commodate a state’s conditions in the assessment equation. These cases thus mark
a shift towards a more tempered approach, combining subjective and objective

96 I  Brownlie, State Responsibility (Clarendon Press 1983)  162, 168; Newcombe and Paradell,
Law and Practice (n 33) 310; H Bray, ‘SOI–​Save Our Investments! International Investment Law and
International Humanitarian Law’ (2013) 14(3) JWIT 578; De Brabandere, ‘Due Diligence’ (n 56) 353.
See also AAPL (n 23) para 77.
97 See A Freeman, Responsibility of States for Unlawful Acts of their Armed Forces (AW Sijthoff 1957)
277–​78 (stating that due diligence ‘is nothing more nor less than the reasonable measures of prevention
which a well-​administered government could be expected to exercise under similar circumstances’);
Dolzer and Schreuer, Principles (n 56) 162 (remarking that ‘lack of resources to take appropriate action
will not serve as an excuse for the host state’).
98 AMT (n 21) paras 6.05–​6.06, 6.11, 6.14. According to the tribunal, the government’s duty was to
take ‘every measure necessary to protect’ the investor, rather than what would be ‘reasonably’ expected.
99 Tecmed (n 18) para 177.
100 Suez, Separate Opinion of Pedro Nikken (n 26) para 20.
101 De Brabandere, ‘Due Diligence’ (n 56) 356; Bray, ‘SOI’ (n 96) 591; ILA, First Report (n 62) 10.
102 AAPL (n 23) para 85B.
103 See e.g. Lauder (n 27) para 308; CME (n 27) para 353; El Paso (n 93) para 523.
104  Investment Treaty Protections

standards. The so-​called modified objective approach acknowledges that the host
state is ‘required to exercise an objective minimum standard of due diligence’, how-
ever, the state’s conduct is assessed against what could reasonably be expected of
the state in the given circumstances and in the light of its resources.104 The standard
of vigilance required from the state thus depends on a number of factors, such as
the intensity of the violence and the resources that can be diverted for the purposes
of protection.105 The importance of the standard in conflict situations was high-
lighted in Pantechniki v Albania in a celebrated paragraph:

A failure of protection and security is to the contrary likely to arise in an unpre-

dictable instance of civic disorder which could have been readily controlled by
a powerful state but which overwhelms the limited capacities of one which is
poor and fragile . . . it seems difficult to maintain that a government incurs inter-
national responsibility for failure to plan for unprecedented trouble of unprece-
dented magnitude in unprecedented places.106

In a similar vein, the tribunal in LESI v Algeria rejected the FPS claim because the
government had taken reasonable and proportionate security measures during the
revolution. The tribunal stressed that the duty of due diligence is an obligation of
means and that its assessment depends on the circumstances such as the security
situation in a host state and the specificity of a region where the investment is lo-
cated and which may affect the ability to provide adequate protection.107
The reasoning in these cases implies that the more intense the conflict and the
poorer and more fragile the host state, the more likely it is that the state will not be
found liable for losses incurred by non-​state actors. As articulated by Newcombe
and Paradell: ‘An investor investing in an area with endemic civil strife and poor
governance cannot have the same expectation of physical security as one investing
in London, New York or Tokyo.’108 One could thus observe that the content of due
diligence has evolved and is imbued with more flexibility which accommodates
states’ political and economic realities. The standard which takes into account
the situation of the host state (including the character and the extent of the con-
flict) and the means available to it, has received support by commentators,109 and

104 See Newcombe and Paradell, Law and Practice (n 33)  310; I Brownlie, Principles of Public
International Law (6th edn, OUP 2003) 504.
105 Pantechniki (n 21) para 82.
106 ibid para 77.
107 LESI (n 22) para 181. It should be noted that the tribunal applied a basic armed conflict clause as
lex specialis to exclude the application of FPS, which is criticized below. Effectively, however, the tri-
bunal was analysing the government’s due diligence. See Section 4 C.4.a.
108 Newcombe and Paradell, Law and Practice (n 33) 310.
109 ibid; Brownlie, Principles (n 104); McLachlan et  al, Substantive Principles (n 32)  331; De
Brabandere, ‘Due Diligence’ (n 56); N Gallus, ‘The Influence of the Host State’s Level of Development on
International Investment Treaty Standard of Protection’ (2005) 6 JWIT 711; Sornarajah, International
Law (n 32) 162.
 Full Protection and Security  105

mirrors the application of due diligence in the law of state responsibility for injury
to foreigners under customary international law discussed in Chapter 2.
The flexible nature of the due diligence standard does not imply that a state fa-
cing poverty along with a severe security crisis can never be found liable for the
breach of an FPS provision. Each case and every episode of an alleged failure to
protect investment must be assessed on their own facts, and it is entirely possible
that even in the midst of a prolonged armed conflict, the circumstances and the
available resources are such to permit state forces to prevent or mitigate losses.110
For example, in Ampal v Egypt, the tribunal found that Egypt failed to protect the
investor’s pipeline system from terrorist attacks in the course of the Arab Spring
revolution.111 The award has been criticized by some commentators for embracing
an absolute, uniform due diligence standard and disregarding Egypt’s extraor-
dinary circumstances at the time of attacks.112 The close reading of the award sug-
gests that the claims by the commentators are exaggerated. Namely, the tribunal,
by referencing the Pantechniki award, expressly acknowledged the relative nature
of due diligence under the FPS provision and stated that the standard ‘must be as-
sessed according to the particular circumstances in which the damage occurs’.113
By inspecting each violent incident separately, the tribunal paid respect to dif-
ferent types of circumstances that inform the assessment of whether a due diligence
duty was met. Thus, it held that the government had not incurred responsibility for
the losses emanating from the first four terrorist attacks as they could not have
been prevented due to the unprecedented nature of violent events (foreseeability
and magnitude).114 It noted, however, that they should have served as a warning
to the government ‘that further attacks might be carried out if security measures
were not taken and implemented’.115 It further took into account that, in particular
as of the date of attack number five, the well-​resourced Egyptian armed forces were
present in high numbers in the area of the pipeline (availability of resources and
state’s capacity), and that the attacks did not include mob violence or the use of
heavy arms (degree of violence).116 While it acknowledged the it could be burden-
some to secure an almost 200 km long pipeline (topographical circumstances), it

110 See e.g. Houben (n 17) paras 160–​64 (the tribunal paid due regard to the state’s individual circum-
stances, while still finding the state responsible for not utilizing the available resources to remove the
squatters from the investor’s premises).
111 Ampal (n 22) para 290.
112 See e.g. A Yacoub, ‘The Case of Ampal v Egypt: What Are the Parameters of the Due Diligence
Standard?’ (Cambridge International Law Journal, 16 November 2018) <http://​cilj.co.uk/​2018/​11/​16/​
the-​case-​of-​ampal-​v-​egypt-​what-​are-​the-​parameters-​of-​the-​due-​diligence-​standard/​> accessed 18
December 2018; I Ryk-​Lakhman, ‘Foreign Investments as Non-​Human Targets’ in B Baade et al (eds),
International Humanitarian Law in Areas of Limited Statehood—​Adaptable and Legitimate or Rigid and
Unreasonable? (Nomos 2018) 171.
113 Ampal (n 22) paras 241, 244, 284.
114 ibid paras 285, 289.
115 ibid para 289.
116 ibid paras 808–​09.
106  Investment Treaty Protections

also observed that attacks concentrated on the same closely located segments of the
pipeline,117 and that there was no evidence that the authorities took any ‘concrete
steps’ to protect the investment in reaction to terrorist activities.118 Moreover, in
one instance, the security forces were called to help stop the saboteurs from laying
explosives on the pipeline, but refused to do so despite their availability, proximity
of their location to the pipeline, and sufficient time for an adequate action.119
While one may disagree with the tribunal’s assessment of the evidence in the
Ampal case, it clearly follows from the award’s reasoning that there was no devi-
ation from the relative nature of the due diligence standard. Despite inconsistent
application of the FPS provision in the early investment treaty practice, and con-
tinuing criticism in scholarship, the more persuasive view is that a state’s exercise
of due diligence in providing physical security against non-​state actors is guided
according to the relative standard that takes into account the relevant state’s
circumstances.

C.  Armed Conflict Clause

The provision that can be found in most investment treaties and refers directly
to situations of conflict has been known in investment scholarship as the ‘war
clause’.120 This term appears to be a misnomer, particularly because such provi-
sions, as a rule, refer not only to war but also to other situations of violence, such as
insurrections, revolutions, riots, civil disturbances, or similar events. A more pre-
cise term would thus be an armed conflict clause, whereby ‘armed conflict’ is used
as an overarching term encompassing the situations enumerated in the respective
clause. As mentioned in Chapter 1, some BITs have explicitly embraced this ap-
proach and use ‘armed conflict’ as a generic term covering all situations of conflict
listed in the clause.121
Armed conflict clauses are among the least examined provisions of investment
treaties.122 This is perhaps unsurprising given the fact that tribunals mostly re-
jected the application of the clause. The provision was invoked in cases involving

117 ibid paras 793–​99.

118 ibid para 290.
119 ibid para 288.
120 Schreuer, ‘Protection of Investments’ (n 1) 12; Newcombe and Paradell, Law and Practice (n 33)
499; O Mayorga, ‘Arbitrating War: Military Necessity as a Defence to the Breach of Investment Treaty
Obligations’ (Policy Brief, August 2013) Program on Humanitarian Policy and Conflict Research
Harvard University.
121 Philippines–​Pakistan BIT (1999) Art V; Philippines–​Myanmar BIT (1998) Art V. Some invest-
ment treaties use the term ‘hostilities’ to cover different type of situations, from revolution to insurrec-
tions. See e.g. Japan–​Republic of Korea BIT (2002) Art 11.
122 For a limited discussion, see Schreuer, ‘Protection of Investments’ (n 1) 12–​16; F Perez-​Aznar,
‘Investment Protection in Exceptional Situations:  Compensation-​for-​Losses Clauses in IIAs’ (2017)
32(3) ICSID Rev 696.
 Armed Conflict Clause  107

armed conflict,123 invasion and seizure of commercial farms,124 economic crisis,125

and extreme weather conditions.126 Since they specifically address the situation of
conflict and their analysis by tribunals has been inconsistent and often inaccurate,
they merit deeper analysis.
The provision provides foreign investors with certain guarantees with respect
to the payment of compensation for losses suffered in times of conflict. The scope
of these guarantees is twofold: basic and advanced. The basic clause provides for
a non-​discriminatory treatment as to the payment of post-​conflict indemnities,
while the advanced clause creates a ground for remedying losses inflicted by gov-
ernment forces, and introduces exceptions to state responsibility. Before describing
the content of these two types of a clause, its scope will be first outlined.

1.  Scope of the Clause

The wording of armed conflict clauses across different investment treaties is

similar, while demonstrating some variation as to the situations they cover. For ex-
ample, some treaties entail a more detailed list of violent situations (e.g. the Libya–​
Portugal BIT adds ‘disobedience’ or ‘disturbances’127 and the 2008 UK Model BIT
adds ‘revolt’ and ‘riots’128), some are shorter and use more general terms (e.g. the
2012 US Model BIT is streamlined to cover ‘armed conflict and civil strife’129), and
some extend the scope of the clause to situations that are not necessarily related
to violence (e.g. the 2010 Austria Model BIT also covers ‘acts of God or force ma-
jeure’,130 while the 2004 Canada Model BIT adds ‘natural disasters’131 to the list).

123 AAPL (n 23) paras 65–​70; AMT (n 21) para 6.12; LESI (n 22).

124 Bernardus Funnekotter and Others v Republic of Zimbabwe ICSID Case no ARB/​05/​6, Award, 22
April 2009, para 104; von Pezold (n 17) para 592.
125 CMS Gas Transmission Company v Argentine Republic ICSID Case no ARB/​01/​8, Award, 12 May
2005, para 375; LG&E Energy Corp v Argentine Republic ICSID Case no ARB/​02/​1, Decision on Liability,
3 October 2006, para 243; Enron (n 41) para 320; Sempra Energy International v Argentine Republic
ICSID Case no ARB/​02/​16, Award, 28 September 2007, para 362; National Grid plc v Argentine Republic
UNCITRAL, Award, 3 November 2008, para 253; El Paso (n 93) para 559; BG Group (n 26) para 382;
Total SA v Argentine Republic ICSID Case no ARB/​04/​1, Decision on Liability, 27 December 2010, para
230; Impregilo SpA v Argentine Republic ICSID Case no ARB/​07/​17, Award, 21 June 2011, para 341; Suez
(n 26) para 271; AWG Group Ltd v Argentine Republic UNCITRAL, Decision on Liability, 30 July 2010,
para 271; EDF International SA and Others v Argentine Republic ICSID Case no ARB/​03/​23, Award, 11
June 2012, paras 1157, 1162.
126 Consortium RFCC v Morocco v Kingdom of Morocco ICSID Case no ARB/​00/​6, Award, 22
December 2003.
127 Libya–​Portugal BIT (2003) Art 7.
128 2008 UK Model BIT Art 4.
129 2012 US Model BIT Art 5(4). Similarly, NAFTA Art 1105(2).
130 2010 Austria Model BIT Art 8(1).
131 2004 Canada Model BIT Art 12(1). It would seem that the absence of the express reference to nat-
ural disasters would preclude the application of a clause to such situations. For example, Consortium
RFCC tribunal rejected the application of the clause, which was confined to situations of conflict, to
extreme weather conditions, noting that they do not constitute an event analogous to armed conflict.
Consortium RFCC (n 126) para 80.
108  Investment Treaty Protections

These differences in terminology are mostly not material. What matters is that the
provision establishes a certain threshold for a conflict situation that needs to be
met for the provision to apply. The concept of conflict that determines the area
of operation of the clause is broader than the IHL-​based ‘armed conflict’, but nar-
rower than forcible interferences covered by the FPS clause. Thus, armed conflict
clauses do not apply to isolated incidents specifically targeting investors, like pri-
vate harassment and occupations of investors’ premises. Instead, they are designed
to address losses arising directly out of forcible actions related to situations of
such proportion as to constitute an emergency. Given the breadth of the situations
envisaged in such clauses, it should not be too problematic for a tribunal to as-
sess prima facie if the clause applies. However, in some situations this may not be
straightforward.
For example, does the clause cover losses arising out of acts of looting, van-
dalism, and mere banditry? The answer would likely depend on whether such
an act was situated within the context of a wider situation of disorder covered
by the clause.132 Riots typically involve vandalism and the destruction of private
and public property. Acts of vandalism and theft not carried out during times of
such civil disorder would, however, be unlikely to trigger the applicability of the
clause. In AMT v Zaire, the investor was subjected to two separate occasions of
looting by the soldiers of the Zairean armed forces. The destruction and theft of
the investor’s property was part of the widespread looting experienced by Zaire
in September 1991 and January 1993. The tribunal held that the armed conflict
clause was applicable as the situation in question was covered by the wording ‘riot
or act of violence’.133 Even when the clauses are worded more economically (e.g.
‘armed conflict and civil strife’),134 such acts of looting are covered therein. The
same should apply to strikes—​if general, appearing on a massive scale, driven by
broader discontent with a social and economic situation in the state and resulting
in widespread violence, they should fall under the scope of the clause. On the other
hand, isolated strikes and protests of a smaller scale would probably not reach the
applicability threshold.
Another question is whether armed conflict clauses also apply to terrorist acts.
Many clauses are non-​exhaustive in listing the violent situations, stating instead
that they apply to any similar events (e.g. the US–​Ukraine BIT). Would a terrorist
act be interpreted as such an event? At first glance, there is a notable difference,
namely situations usually listed in the clauses are forms of collective violence and
mass incidents, whereas a terrorist attack can be planned and executed by a small
group of people or even an individual. The number of people involved should not

132 The government in Zimbabwe unsuccessfully relied on the armed conflict clause as a defence, ar-
guing that its application was triggered by the state of emergency resulting from invasion by settlers and
war veterans of foreign investors’ farms. See Funnekotter (n 124) para 104; von Pezold (n 17) para 592.
133 AMT (n 21) para 6.12.
134 2012 US Model BIT Art 5(4).
 Armed Conflict Clause  109

be a determinative factor, however. What matters more is the potential for mass im-
pact that such an attack can have and the emergency situation to which it amounts.
In fact, terrorist acts are frequently included as a type of civil emergency in national
governments’ strategies and on national risk registers.135 If armed conflict clauses
list a ‘state of emergency’ as one of the violent situations they cover, terrorist attacks
may likely be covered as well. The same could apply to the scarcely worded clauses.
This broad interpretation was supported by Scott Gudgeon, a key negotiator of
US BITs, who explained in the commentary of the US Model BIT that even the
wording ‘war or civil disturbance’ should be understood to include terrorism.136 To
avoid any doubt as to the broad scope of the clause, a specific reference to an ‘act of
terrorism’ has been included in some of the early US BITs.137

2.  Basic Armed Conflict Clause

Most investment treaties contain a basic armed conflict clause that imposes on a
host state a specific non-​discrimination obligation, typically with respect to the
payment of indemnities for losses sustained by investors in a situation of con-
flict.138 Unlike the FPS provision, the basic armed conflict clause was not featured
in the FCN treaties or in drafts of multilateral investment treaties. The exceptions
are the provisions in the Resolution of the Institute of International Law (1927)
and the Basis of Discussion drawn by the Preparatory Committee of the Hague
Conference for the Codification of International Law (1929) which provided for a
national treatment as to the payment of post-​conflict indemnities.139
The clause was first introduced in German investment treaties after the Second
World War as a reaction to losses the German investors sustained abroad.140 It
was probably devised by Hermann Abs, who played a pivotal role in the design
of post-​war German economic policy and was part of the delegation negotiating
reparations for the German assets confiscated in the US during the war.141 The

135 See e.g. UK National Risk Register of Civil Emergencies (11 July 2013).
136 Gudgeon, ‘United States Bilateral Investment Treaties’ (n 86) 127.
137 See e.g. US–​Cameroon (1989) Art IV; US–​Senegal (1990) Art IV; US–​Panama (1991) Art V; US–​
Bangladesh (1989) Art IV.
138 According to UNCTAD, 90 per cent of mapped investment treaties contain such clauses.
UNCTAD Navigator (n 10).
139 Draft on International Responsibility of States for Injuries in their Territory to the Person or
Property of Foreigners, Prepared by the Institute of International Law (1927) Art VII in ILC, Yearbook
of the International Law Commission, 1956, Vol II, UN Doc A/​CN.4/​SER.A/​1956/​Add.1, 227; Bases of
Discussion Drawn up in 1929 by the Preparatory Committee of the Conference for the Codification of
International Law (The Hague, 1930) No 21(4) and No 22(b) in ILC, Yearbook of the International Law
Commission, 1956, Vol II, UN Doc A/​CN.4/​SER.A/​1956/​Add.1, 222 (Bases of Discussion).
140 It is included in the first BIT between Germany and Pakistan (1959) Art 3(3). See also Vandevelde,
History (n 8) 311.
141 ‘Abs, Hermann Josef ’ (Declassified and Released by Central Intelligence Agency, Nazi War Crimes
Disclosure Act, 2001). <https://​www.cia.gov/​library/​readingroom/​docs/​ABS%2C%20HERMANN%20
J._​0051.pdf> accessed 18 December 2018. Interestingly, the Abs-​Shawcross Convention, drafted
110  Investment Treaty Protections

only partial success in obtaining the settlement could explain the motivation be-
hind this treaty innovation. The language from German BITs was soon adopted in
Italian BITs, followed by France, the UK, and the US shortly thereafter, eventually
becoming a common provision in most investment treaties.142
The basic clauses guarantee foreign investors that the host state will provide
them with national treatment, most favoured nation (MFN) treatment, or both
with respect to measures such as restitution, indemnification, or compensation for
losses occurring in situations of armed conflict. For example, Article 3(3) of the
US–​Ukraine BIT provides:

Nationals or companies of either Party whose investments suffer losses in the ter-
ritory of the other Party owing to war or other armed conflict, revolution, state of
national emergency, insurrection, civil disturbance or other similar events shall
be accorded treatment by such other Party no less favorable than that accorded to
its own nationals or companies or to nationals or companies of any third country,
whichever is the most favorable treatment, as regards any measures it adopts in
relation to such losses.

The prevailing view among commentators is that basic clauses do not create sub-
stantive rights to restitution or compensation, that is they do not oblige a host state
to compensate an investor for a covered loss.143 This view was confirmed in several
arbitral awards resulting from the Argentine economic crisis and was articulated
clearly by the tribunal in CMS v Argentina who stated that the provision is meant:

. . . to provide a floor treatment for the investor in the context of the measures
adopted in respect of losses suffered in the emergency, not different from that ap-
plied to national or other foreign investors. [It] does not derogate from the Treaty
rights but rather ensures that any measures directed at offsetting or minimizing
losses will be applied in a non-​discriminatory manner.144

Some BITs explicitly limit such measures to ‘restitution, indemnification, com-

pensation or other valuable consideration’.145 In other words, if the state does not

in the same period, did not include the armed conflict clause. The reason for this could be that Abs and
Shawcross wanted to keep the draft simple and avoid any specific, politically sensitive provisions (espe-
cially in the aftermath of war) that could complicate obtaining a multilateral consensus.

142 Only a minority of BITs do not contain the provision, e.g. the Argentina–​Mexico BIT (1996).
143 C Brown, Commentaries on Selected Model Investment Treaties (OUP 2013) 164; Newcombe and
Paradell, Law and Practice (n 33) 315, 500; Gudgeon, ‘United States Bilateral Investment Treaties’ (n
86) 129, Schreuer, ‘Protection of Investments’ (n 1) 12; Salacuse, Law of Investment Treaties (n 32) 369;
K Vandevelde, United States Investment Treaties:  Policy and Practice (Kluwer 1992) 212; UNCTAD
Report, ‘Bilateral Investment Treaties 1995–​2006: Trends in Investment Rule Making’ (2006) 55.
144 CMS Gas (n 125) para 375. For other cases, see n 125.
145 See e.g. Germany–​Syria BIT (1980) Art 4(3); Greece–​Syria BIT (2003) Art 6(1).
 Armed Conflict Clause  111

compensate either its national investors or investors from third countries for losses
covered by such clauses, foreign investors have no right to compensation under
the title of this provision.146 On the other hand, if the host state decides, as a matter
of domestic law or policy (e.g. as an act of grace) to provide payment to its own or
third-​country investors, it should do the same, and on equal terms, for foreign in-
vestors protected under the investment treaty.
This view has been contested in the case law.147 The tribunal in AAPL v Sri Lanka
interpreted the clause broadly and held that it contained a renvoi to the entire body
of customary international law.148 It considered that the national and MFN treat-
ment did not refer only to matters of compensation, but also to the determination
of a host state’s liability. In other words, it construed the provision as to impose
a substantive obligation on the host state to pay compensation to investors for
conflict-​related losses. This enabled the tribunal to incorporate the customary rule
of due diligence into the investment treaty and on that basis found the host state
liable. This interpretation is problematic for several reasons and has been rightly
criticized by the dissenting arbitrator as ‘fundamentally erroneous’.149 First, the
provision in question was titled ‘Compensation for losses’ and, second, it expli-
citly referred to ‘restitution, indemnification, compensation or other settlement’.150
Extending the treatment beyond these measures was not in line with the wording
of the provision and contradicted its purpose. Even when the clause uses a more
general wording, as in the above-​cited US–​Ukraine BIT, and refers to ‘any meas-
ures [the host State] adopts in relation to such losses’, this does not provide national
or MFN treatment as to the question of liability. Namely, the clause refers to meas-
ures that are adopted only after the losses have been incurred, that is after the point
at which the liability arose. Such measures are beneficial to those who have already
sustained losses since they purport to mitigate or repair them.151 Thus, the basic
clause arguably excludes the question of liability even in cases when it is not expli-
citly limited to the question of compensation.
The clause covers situations where the liability is not established but the host
state nonetheless decides to implement measures, such as payments according
to its own national policies. This interpretation reflects the historical practice of

146 An odd exception to this rule seems to be created, intentionally or through hasty drafting, in some
of the Italian BITs which impose a duty on a host state to offer an adequate compensation for conflict-​
related losses, regardless of who caused them. See e.g. Syria–​Italy BIT (2003) Art 4; Armenia–​Italy BIT
(2003) Art 4; Morocco–​Italy (2000) Art 4; Bangladesh–​Italy (1994) Art 4. See also Consortium RFCC
case, in which the tribunal acknowledged that Art 4(1) of the Morocco–​Italy BIT established a strict,
objective liability. Consortium RFCC (n 126) para 56.
147 See e.g. AAPL (n 23); AMT (n 21) para 6.14; LESI (n 22) para 175. See Section 4 C.4.a.
148 AAPL (n 23) paras 65–​70.
149 ibid, Dissenting Opinion, 586.
150 UK–​Sri Lanka BIT (1980) Art 4(1).
151 The clause applies only with respect to state’s corrective or compensatory measures taken in reac-
tion to already incurred losses. See Impregilo (n 125) paras 341–​43; El Paso (n 93) para 559; BG Group (n
26) para 382; Total (n 125) para 229; Enron (n 41) para 320.
112  Investment Treaty Protections

paying voluntary awards of indemnities for conflict-​related losses, for which no

legal liability was incurred. While it was sometimes argued that host states were
responsible for indemnifying foreigners if they had indemnified their own na-
tionals,152 often states arbitrarily limited the classes of the beneficiaries, which
sometimes excluded foreigners from being paid voluntary indemnities.153 Since
those national compensation programmes were discretionary and no inter-
national custom emerged that would provide a national or MFN treatment with
respect to such payments,154 the investment treaty regime went further and cre-
ated a binding obligation upon a host state to provide such non-​discriminatory
treatment. Arguably, this obligation is not covered by the general national or MFN
treatment clauses that refer to establishment and treatment of foreign investment,
and the activities associated with it.155 Payment of compensation falls outside that
scope which explains the rationale for including basic armed conflict clauses in in-
vestment treaties.156

3.  Advanced Armed Conflict Clause

Much less common are advanced armed conflict clauses,157 which cover the
same situations, but go a step further. They elevate the protection of foreign in-
vestors by granting them a substantive right to restitution or compensation for
losses incurred by the host state’s forces or authorities through requisitioning or
unnecessary destruction of an investor’s property. The origin of the provision can
be traced to the late eighteenth-​century treaties of amity and commerce and early

152 See e.g. the claims against Belgium after the battle of Antwerp in 1830. While the Belgian gov-
ernment compensated its own citizens for the losses suffered during the battle, no such measures were
adopted with respect to foreign nationals. In the end, the claims were settled on an equitable basis.
Reported in Study by the Secretariat, ‘ “Force majeure” and “Fortuitous event” as Circumstances
Precluding Wrongfulness: Survey of State Practice, International Judicial Decisions and Doctrine’ in
ILC, Yearbook of the International Law Commission, 1978, Vol II, UN Doc A/​CN.4/​SER.A/​1978/​Add.1,
61, 106. See also C Eagleton, The Responsibility of States in International Law (New York University Press
1928) 152; J Moore, A Digest of International Law (Vol 6, Washington GPO 1906) 892 (reporting that
the US government urged the Brazilian government to pay indemnities for injuries sustained by the
American investor in the course of insurrection of 1893 merely on the basis that the payment for similar
losses had been made to Brazilian companies).
153 E Borchard, Diplomatic Protection of Citizens Abroad (Banks Law Publishing 1915) 279, 280.
154 For a different view, see Perez-​Aznar, ‘Compensation-​for-​Losses Clauses’ (n 122) 714 (arguing
that the basic armed conflict clause reflects customary international law).
155 See Aroa Mines Case (Great Britain v Venezuela) (1903) 9 RIAA 402, 407.
156 Newcombe and Paradell have argued that such measures qualified as ‘aid’, and unless the dis-
tinction between similarly situated investors was arbitrary, it was not prohibited under the customary
international law. Newcombe and Paradell, Law and Practice (n 33) 225. A similar view was held by
arbitrator Huber in Spanish Zone of Morocco (Great Britain v Spain) (1924) 2 RIAA 615, 625. On the
other hand, Salacuse has argued that the clause adds no new protection to treaties already containing a
national treatment or MFN clause. Salacuse, Law of Investment Treaties (n 32) 369–​70.
157 According to UNCTAD, they appear in around 33 per cent of investment treaties. UNCTAD
Navigator (n 10).
 Armed Conflict Clause  113

nineteenth-​century FCN treaties which sometimes required that the property

seized during wartime is restored to the owner,158 or compensated.159 A version of
the clause appeared also in the Basis of Discussion No 21 drafted by the Preparatory
Committee of the Hague Conference.160 However, it was not incorporated in other
drafts addressing state responsibility for injuries to foreign property and it was not
included in the US post-​Second World War FCN treaties.
The clause was introduced to modern investment treaties by the UK.161 One can
but suspect that Lord Shawcross, whose familiarity with the laws of war was well
established (he was a British prosecutor at the Nuremberg War Crimes tribunal),
played some part in this through his role as a director of Shell which at the time ac-
tively participated in consultation with the British government in drafting the UK
model investment treaty.162 The provision has since become part of the majority of
UK BITs, and also features in some other investment treaties, including the 2012
US Model BIT and the Energy Charter Treaty.163 An example is found in Article
5(2) of the UK–​Ukraine BIT:
Without prejudice to paragraph (1) [basic clause] of this Article, investors of one
Contracting Party who in any of the situations referred to in that paragraph suffer
losses in the territory of the other Contracting Party resulting from
a) requisitioning of their property by its forces or authorities, or
b) destruction of their property by its forces or authorities which was not caused
in combat action or was not required by the necessity of the situation,
shall be accorded restitution or adequate compensation. Resulting payment
shall be freely transferable.

The advanced standard mirrors the principles of the law of war on the protec-
tion of private property. As discussed in Chapter 2, the rule that compensation is
due when armed forces seize an alien’s private property, or when the destruction
thereof was not incidental to combat action and not required by military neces-
sity, developed in the practice and jurisprudence of the nineteenth century and
was later codified in the Hague and Geneva Conventions. Some scholars have thus
argued that the clause is superfluous since investors are already protected under
the laws of war.164 This assumption seems to be inaccurate as there are benefits

158 Treaty of Friendship, Commerce and Navigation, Mexico–​United Kingdom (1826) Art XII;
Treaty of Friendship, Commerce and Navigation, Colombia–​US (1824) Art 24.
159 Treaty of Amity and Commerce, US–​France (1778) Art 22; Treaty of Amity, US–​Prussia (1799)
Art XXIII.
160 Bases of Discussion (n 139).
161 The first BIT to introduce the clause was the UK–​Egypt BIT in 1975 in Art 4(2).
162 J Bonnitcha, L Poulsen, and M Waibel, The Political Economy of the Investment Treaty Regime
(OUP 2017) 188.
163 See e.g. UK–​China (1986) Art 4(2); UK–​Bolivia BIT (1988) Art 4(2); UK–​Sri Lanka BIT, Art 4(2).
164 See Gudgeon, ‘United States Bilateral Investment Treaties’ (n 86) 128.
114  Investment Treaty Protections

that investors could derive from the advanced clause that are not available to them
under the IHL framework.
First, although historically the customary law of armed conflict prohib-
ited the seizure of private property without payment of compensation or its
return,165 practice during the two World Wars deviated from this rule. Thus,
during the First World War, the Allied powers allowed their nationals to confis-
cate the private investments of enemy nationals for their own use.166 The Treaty
of Versailles did not penalize these illegal practices, but instead authorized the
Allies ‘to retain and liquidate all the property and interests of German subjects
or companies under their control in their territory’, pending Germany’s pay-
ment of war reparations.167 The confiscations of the private assets of German
nationals and companies resumed in the Second World War with the aim of
securing the payment of reparations and preventing Germany from becoming
a global superpower.168 In the light of these practices, the controversial view
emerged that international law no longer prohibited the confiscation of private
enemy property during times of war.169 By expressly prescribing the payment of
compensation for the requisition of investment property or its restitution, ad-
vanced armed conflict clauses in investment treaties could be thus seen as clari-
fying and re-​establishing the customary protections of private property that had
existed prior to the First World War. Indeed, this was the intention of the British
government in drafting the first clause,170 and is confirmed in some investment
treaties in express terms.171
Second, and as discussed in Chapter 2, Additional Protocol II to the Geneva
Conventions (AP II), which regulates non-​international armed conflicts, contains
no provision on the protection of private property, which makes it less clear to what
extent foreign investors are protected against requisition and destruction taking
place during civil wars. What is clear, however, is that IHL rules do not protect in-
vestors in internal disturbances that fall below the threshold of armed conflict. In
contrast, advanced armed conflict clauses extend protections to ‘lesser’ types of

165 See e.g. the Upton Case (US v Venezuela) (1903–​1905) 9 RIAA 235–​36.
166 See Borchard’s ‘Introduction’ in J Gathings, International Law and American Treatment of Alien
Enemy Property (American Council on Public Affairs 1940) v–​vi. See also Civilian Claims:  Eritrea’s
Claims 15, 16, 23, 27–​23 (Partial Award of 17 December 2004) (2004) 26 RIAA 195, 236, para 128.
167 Treaty of Versailles (1919) Art 297(b) and (e).
168 See e.g. Paris Reparation Agreement (15 March 1946) 555 UNTS 69, Part I, Art 6(A). See also M
McDougal and F Feliciano, ‘International Coercion and World Public Order: The General Principles of
the Law of War’ (1958) 67 Yale L J 771; H Van Houtte et al, Post-​War Restoration of Property Rights under
International Law (CUP 2008) 275–​82.
169 See e.g. Contra Prince Salm-​Salm v The State of the Netherlands and the Nederlands Beheers-​
Instituut District Court of The Hague, 28 June 1954; Court of Appeal of The Hague, 8 November 1956
and Supreme Court, 21 June 1957 (1957) 24 ILR 893, 895; Van Houtte (n 168).
170 E Denza and S Brooks, ‘Investment Protection: United Kingdom Experience’ (1987) 36 ICLQ 908,
911–​12.
171 See e.g. Armenia–​Austria BIT (2001) Art 6(2) (noting that the provision reflects ‘common
international law’).
 Armed Conflict Clause  115

conflict, such as revolts and revolutions. Third, the investment treaty framework
provides investors with the right to seek remedy directly from the host state, a priv-
ilege that is not available under IHL.
While the clause clarifies the substantive protection to investors, it also brings
important benefits for a host state by specifying situations in which a state is ex-
empt from liability for destruction of an investor’s property, namely when the loss
was caused in ‘combat action’ or due to ‘necessity of the situation’.172 While this
specific exception is reflective of IHL practice, caution should be exercised in the
interpretation of these concepts in the investment law context. The following two
sections will explain how they were addressed by the AAPL tribunal, elucidating
their meaning and highlight their problematic aspects.

(a) Combat action
The case AAPL v Sri Lanka concerned a Sri Lankan counter-​insurgency operation
against the Liberation Fighters of Tamil Elaam (the ‘Tamil Tigers’). In January
1987, a Sri Lankan Special Task Force launched a forceful attack on the Serendib
Shrimp Farm, co-​owned by a Hong Kong company called AAPL. According
to the official accounts, the intense combat action resulted in the death of more
than twenty AAPL employees and the destruction of the farm’s property, which
led to the termination of the farm’s operations.173 The claimant alleged that the
Sri Lankan government was liable for the destruction of the investment because it
breached the advanced armed conflict clause in Article 4 of the applicable UK–​Sri
Lanka BIT. The clause provides that the host state has to pay adequate compensa-
tion in cases of:
. . . war or other armed conflict, revolution, a state of national emergency, revolt,
insurrection or riot [causing foreign investors to] suffer losses resulting from
a) requisitioning of their property by [the host State’s] forces or authorities, or
b) destruction of their property by [the host State’s] forces or authorities which
was not caused in combat action or was not required by the necessity of the
situation.

More specifically, the claimant argued that the loss was ‘not caused in combat ac-
tion’, but amounted to ‘the wanton destruction of AAPL’s property and the cold-​
blooded killing of the farm manager and the permanent staff members’ which was
‘clearly not planned pursuant to any combat action’.174 Since the invoked provision
excludes from its scope property destruction caused by the government forces in

172 In some BITs, the clause does not refer to ‘combat action’ See e.g. Austria–​Armenia BIT (2001) Art
6(2); Sweden–​Kazakhstan (2004) Art 5(2); Energy Charter Treaty (1994) Art 12(2); Australia–​Korea
FTA (2014) Art 11.6.
173 Decades later, the testimonies of witnesses provided for a different, more devastating, account. See
Chapter 1, n 1.
174 AAPL (n 23) para 28.
116  Investment Treaty Protections

‘combat action’, the tribunal, presided by Professor Berthold Goldman, had to de-
termine, first, what is meant by ‘combat action’, and second, whether the violent
action in question could be classified as such.
The tribunal decided to interpret the term progressively, ‘according to its natural
and fair meaning as commonly used under prevailing circumstances, ie, within the
context of guerrilla warfare which characterizes the modern civil wars conducted
by insurgents’.175 It accurately noted that the civil wars in recent history have rarely
featured the classical military confrontation ‘between two opposing armed groups
on a battle field where the adversaries engage simultaneously in fighting each other
on the spot’.176 More commonly, they take the form of sporadic attacks by armed
groups and governmental counter-​insurgency actions that can take place in vast
inhabited areas. This led the tribunal to conclude that the military operations on
the AAPL farm qualified as ‘combat action’ and therefore the losses incurred as a
result were not covered by the provision.
Regrettably, the tribunal’s analysis was brief and left much to be desired. In par-
ticular, it would have been useful if the tribunal had clarified the criteria for distin-
guishing combat action from other forms of violent measures. Are there any spatial
and temporal requirements for the conduct to qualify as combat action? Does the
action have to involve direct confrontation between adversaries, or can it qualify as
such even in the absence of the resistance of an armed opposition group? Does it
matter what type of weaponry is used in the action? Does it have to be undertaken
against the backdrop of a civil war, or does it also include law enforcement oper-
ations in less intense internal conflicts?
The text of the armed conflict clause implies that this exception to a state’s li-
ability applies to all types of conflict listed in the basic armed conflict clause (in-
cluding riots and revolts), and to the conduct of both a state’s armed forces as well
as its police authorities. This makes the assessment of the contours of ‘combat ac-
tion’ even more important. In particular, as a consequence of a broad interpretation
of ‘combat action’, the host state could benefit from the exemption even in cases of
isolated anti-​terrorist measures that take place outside of the context of armed con-
flict, or even police action against demonstrators and rioters. It thus seems apposite
to inspect how the term has been interpreted in other legal contexts.
IHL, whose principles have likely influenced the content of advanced armed
conflict clauses, is limited to the regulation of belligerent parties’ conduct in armed
conflict as defined in its treaties and jurisprudence, excluding law enforcement
measures in peacetime. Within that context, however, combat action is broadly
perceived, encompassing different styles of fighting, including guerrilla actions.177
IHL rules equate ‘combat action’ with the term ‘attack’, which is given a broad

175 ibid para 61.

176 ibid.

177 Additional Protocol I to the Geneva Conventions (AP I) Art 44(3).

 Armed Conflict Clause  117

definition in Article 49 of AP I as ‘acts of violence against the adversary, whether in

offence or in defence’.178 Military attacks must be distinguished from ‘military op-
erations’, which is an even broader concept ‘understood to mean any movements,
manoeuvers and other activities whatsoever carried out by the armed forces with
a view to combat’, and typically before the actual combat.179 While belligerents are
bound by a general duty of constant care in the conduct of military operations with
a view to spare the civilian population and civilian objects,180 more specific obliga-
tions are attached to the planning and carrying out of actual attacks, in the course
of which collateral casualties are tolerated as long as they are not excessive in rela-
tion to the concrete and direct military advantage anticipated.181
Since IHL treaties do not provide much guidance as to what constitutes ‘combat
action’ outside the scope of IHL-​defined armed conflict, other frameworks are
worth investigating. As mentioned above, international human rights instruments
do not formally distinguish between military operations and law enforcement ac-
tions. This notwithstanding, it would appear that the distinction between a law
enforcement operation and a real combat action was taken into consideration by
the ECtHR when assessing whether or not the use of force on the part of the re-
spondent state had been lawful.182
A more concrete analogy can be found in domestic torts laws that exempt states
from civil liability for losses incurred in combat actions. A few such cases emerged
in the aftermath of the Second World War. For example, in Adams v Naylor, the
House of Lords opted for a broad interpretation of ‘combat’ which included not
only the actual fighting between belligerent parties, but also other measures under-
taken to defeat the enemy, such as laying down a minefield.183 In Johnson v US, the

178 ICRC Commentary to AP I, Art 49(1), 602, para 1880.

179 ICRC Commentary to AP I, Art 57(1), 680, para 2191. See also ICRC Commentary to Art 51(1),
para 1936, which defines military operations as ‘all the movements and activities carried out by armed
forces related to hostilities’. See also J-​F Queguiner, ‘Precautions under the Law Governing the Conduct
of Hostilities (2006) 88(864) IRRC 793, 797.
180 AP I, Art 57(1).
181 AP I, Art 57(2). At the time of ratification of AP I, several states clarified that in determining
the anticipated military advantage, attacks are considered as a whole, rather than isolated and par-
ticular parts thereof. See e.g. ‘UK Reservation Re Article 51 and 57 of Additional Protocol I’ (2 July
2002)  <https://​ihl-​databases.icrc.org/​ihl/​NORM/​0A9E03F0F2EE757CC1256402003FB6D2?OpenDo
cument> accessed 18 December 2018.
182 See Ahmet Özkan and Others v Turkey App no 21689/​93, Judgment (ECtHR, 6 April 2004) paras
305–​06; Isayeva v Russia App no 57950/​00, Judgment (ECtHR, 24 February 2005)  paras 180–​81
(Isayeva II); Ergi v Turkey App no 23818/​94, Judgment (ECtHR, 28 July 1998); Güleç v Turkey App
no 21593/​93 (ECtHR, 27 July 1998). See also A Gioia, ‘The Role of the European Court of Human
Rights in Monitoring Compliance with Humanitarian Law in Armed Conflict’ in O Ben-​Naftali (ed),
International Human Rights and International Humanitarian Law (OUP 2011) 226.
183 [1946] 2 All ER 241, [1946] AC 543. The case concerned the death of a child killed by a mine that
was placed on the beach by the UK military authorities during the Second World War. The House of
Lords unanimously held that the death resulted from the use of the mine for combat activities and thus
the state was not responsible for the death. The question of whether laying down the mines constitutes
an attack (i.e. combat action) has been raised by the ICRC, which appears to support the view that there
is an attack in as much as a person is ‘directly endangered by a mine laid’. ICRC Commentary to AP I,
Art 49(1), para 1881.
118  Investment Treaty Protections

Court of Appeal interpreted the expression ‘combatant activities’ to ‘include not

only physical violence, but activities both necessary to and in direct connection
with actual hostilities’.184
In the context of modern combat action, the relatively recent case of the Israeli
District Court is interesting. The case of Ploni v The State of Israel concerned an in-
nocent bystander who was injured as a result of the targeted killing operation con-
ducted by the Israeli Defence Forces against an active member of the Islamic Jihad
terrorist organization.185 Since the Israeli Civil Torts Act provides that the state is
exempt from liability for losses incurred in combat action, the Court had to deter-
mine whether targeted killing operations qualified as such. The Court emphasized
the following criteria in its assessment: the context in which the anti-​terrorist op-
eration had been executed (the conflict between Israel and terrorist organizations
in Gaza), the urgency to protect Israel’s security interests (the action was based
on intelligence which indicated that the targeted terrorist was planning a suicide
terrorist attack), and the nature of the operation (an aerial attack, a typical warfare
method). The Court held that the ‘combat action’ did not require that the action
had to be taken against the armed forces of another state, and concluded that the
anti-​terrorist operation in question fell within the scope of the exception. Since
then, the ‘combat action’ definition in the Israeli Civil Torts Act has been extended
to encompass actions designed to combat and prevent terrorism or hostile acts.186
Based on the wording of advanced armed conflict clauses, it appears that the
‘combat action’ exception extends beyond the scope of IHL, providing the host
state with a wider space for the protection of its security interests. While it initially
emerged from the principles governing hostilities in wartime, the clause typically
refers to situations other than war, including riots and even terrorist threats. The
rationale behind the exception is clear: to ensure that a state’s efforts to combat
threats to its national security are not unnecessarily curtailed. However, the broad
interpretation according to which the exception includes measures such as police
actions against demonstrators, or raiding an investor’s property in an attempt to
apprehend terrorists, could easily give rise to abuse. Thus, the preferred interpret-
ative approach entails a careful weighing of different circumstances, including the
backdrop against which the operation was taken, the temporal and spatial nexus
between the state’s action and the broader conflict, the methods and means de-
ployed in the action, and the nature of the organs who undertook the operation, to
name but a few. Such an interpretation is more in line with the ordinary meaning

184 Johnson v United States, 170 F.2d (9th Cir, 1948) 767, 770. See also US v Marks, 187 F.2nd (9th Cir,
1951) 724, 727, 728; Skeels v US, 72 F Supp (W D La, 1947) 372, 374.
185 CA [Civil Action] 1081/​04 Ploni v the State of Israel, cited in Y Shany and I Rosenzweig, ‘Terrorism
and Democracy’ (Issue No 4, April 2009)  <http://​en.idi.org.il/​analysis/​terrorism–​and–​democracy/​
issue–​no–​4/​targeted–​killing–​constitutes–​combat–​action–​ca–​[civil–​action]–​108104–​ploni–​v–​the–​
State–​of–​israel/​–​> accessed 5 April  2016.
186 ibid.
 Armed Conflict Clause  119

of the phrase ‘combat action’, as well as with the object and purpose of investment
treaties.

(b) Necessity of the situation
The second exception to the host state’s liability for the destruction of an investor’s
property under the armed conflict clause is the ‘necessity of the situation’.
Disappointingly, the AAPL tribunal avoided any analysis of the condition by con-
cluding that there was not enough evidence to determine whether the destruction
on the Serendib farm could be attributed to the necessity of the situation. The tri-
bunal, however, indicated that the standard of necessity in question was the same
as the one established in the jurisprudence of the old arbitrations and mixed com-
missions which refused to hold host states responsible for losses incurred during
hostilities when such losses ‘were compelled by the imperious necessity of war’.187
In other words, the tribunal held that the necessity exception in the advanced
armed conflict clause reflected military necessity, an IHL principle described in
Chapter 2.188
The ‘necessity of the situation’ clause provides a justification for a state’s conduct
in specifically outlined exceptional situations. The exception is built into a substan-
tive standard and thus purports to define permissible state actions in conflict situ-
ations.189 It legitimizes the conduct that is otherwise inconsistent with investment
treaty values. In other words, it is a more specific manifestation of a state’s police
powers, creating a zone within which state organs are allowed to take certain meas-
ures with an aim to protect security (not exclusively military) interests.
While commentators, in line with the AAPL award, suggested that the excep-
tion in advanced armed conflict clauses is a codification of the IHL principle of
military necessity,190 this view appears imprecise. This follows from the ordinary
meaning of the investment treaty exception, which refers to the ‘necessity of the
situation’ rather than the jus in bello expression ‘necessity of war’. It is in line with
the broader scope of the armed conflict clause that conceptualizes the category of
conflict in wider terms than IHL. By using the word ‘situation’ instead of ‘war’, the
treaty drafters made the conscious decision that the necessity of the armed con-
flict clause transverses the necessity of humanitarian law. It would be unfair to jus-
tify the destruction of an investor’s property in any situation of internal strife by
invoking a standard of military necessity that bestows a higher degree of discretion
on a military commander than a police officer. While a certain forceful measure
may be considered necessary for achieving a military goal in the context of severe

187 AAPL (n 23) para 63.

188 See Chapter 2 C.1.a.
189 For the comparison and different interpretations of the legal effect of ‘general’ security exceptions,
see Chapter 5 B.3.
190 Mayorga, ‘Arbitrating War’ (n 120); Vandevelde, History (n 8) 310; Newcombe and Paradell, Law
and Practice (n 33) 498.
120  Investment Treaty Protections

hostilities, the same measure could be completely disproportionate and unneces-

sary in a different, less intense context of a law enforcement operation. A more
appropriate view thus seems to be that necessity in armed conflict clauses is a grad-
ated standard that accommodates different types of conflicts.
An analogy can be made with the approach taken by the European Court of
Human Rights (ECtHR) in applying necessity and proportionality in conflict-​
related cases. While an important part of both legal regimes, the principles of
necessity and proportionality are believed to operate differently in international
human rights law and IHL. The ECtHR has assessed the lawfulness of military and
counter-​terrorist measures pursuant to human rights standards by refusing to af-
ford the state’s armed forces a high degree of discretion in applying the necessity
and proportionality standards.191 The case law seems to suggest, however, that the
means necessary for suppressing an insurrection may not be the same as those that
are considered acceptable for apprehending a terrorist192 or suppressing a riot.193
Thus, it would appear that the Court applied the more liberal IHL standard when
the situation in question was one of armed conflict that could qualify as such under
the IHL framework.194
Like the international human rights regime, the investment regime primarily
aims at protecting the individual vis-​à-​vis the state. While neither of the two re-
gimes formally distinguishes between different types of conflict, at least not as a
legal category, both consider them relevant factual circumstances when applying
legal standards, including necessity. Necessity in the armed conflict clause should
thus be used flexibly, moving on a sliding scale of different types of conflict situ-
ations. Only when the situation amounts to armed conflict under IHL would the
standard converge with military necessity. Even then, however, necessity should be
used restrictively to advance the protection of civilians against the effects of attacks
during wartime.195

4.  The Relationship between the Armed Conflict Clause and FPS

Having analysed the content of two provisions most pertinent for investors who
sustained losses in armed conflict, the issue addressed next, and whose treatment

191 Y Shany, ‘Human Rights and Humanitarian Law as Competing Legal Paradigms for Fighting
Terror’ in O Ben-​Naftali (ed), International Human Rights and International Humanitarian Law (OUP
2011) 28–​29. See also Chapter 2 D.1.a.
192 Hamiyet Kaplan and Others v Turkey App no 36749/​97 (ECtHR, Judgment, 13 September
2005) paras  51–​52.
193 Güleç (n 182) paras 70–​71.
194 Isayeva II (n 182) paras 178–​218. Even in those cases, however, the conduct of the armed forces was
subjected to the strict proportionality test. See Chapters 2 and 6 for the discussion on proportionality.
195 See Chapter 6 C.3.
 Armed Conflict Clause  121

in investment jurisprudence has been inconsistent and largely unsatisfactory, con-

cerns the relationship between the FPS provision and armed conflict clauses.

(a) Basic armed conflict clause v FPS

The relationship between the FPS provision and the basic armed conflict clause
is unproblematic as they pursue different objectives: the former creates a ground
for state liability, while the latter only ensures that investors are not discriminated
against when it comes to the payment of conflict-​related compensation or indem-
nities, regardless of state responsibility. This understanding has been confirmed
by a number of arbitral tribunals.196 The provisions thus co-​exist without any par-
ticular interaction.
Some tribunals, however, held different positions—​none very persuasive with
their reasoning. The way the AAPL tribunal construed the clause as a renvoi to the
international minimum standard of treatment has been already critiqued above.
The tribunal in AMT v Zaire similarly treated the clause as an additional ground
for finding of the host state’s liability. According to the tribunal, the purpose of the
clause was merely to reinforce the FPS clause.197 It is puzzling how the tribunal
arrived at this conclusion, in view of the fact that the interpreted treaty provision
expressly referred to the non-​discriminatory treatment as regards making of resti-
tution and payment of compensation.
A different but no less confusing interpretation was given to a clause in another
conflict-​related case, LESI v Algeria. Here, the tribunal held that the situation of
revolution triggered the application of the basic armed conflict clause which, as
lex specialis, excluded the application of a more general FPS provision.198 In the
tribunal’s view, the state was consequently only obliged to provide investors with
a national or MFN treatment rather than a ‘superior’ FPS.199 It is easy to see how
the tribunal’s interpretation was led by the peculiar wording of Article 4.5 of the
Algeria–​Italy BIT which, unlike standard basic armed conflict clauses, does not ex-
pressly link the non-​discriminatory treatment to the payment of compensation.200
However, the tribunal failed to interpret the clause in its context; namely the pre-
ceding paragraph of the same provision, to which the armed conflict clause re-
fers (‘bénéficient, de la part de ce dernier’), concerns payment of compensation. It

196
See n 125.
197
AMT (n 21) para 6.14.
198 LESI (n 22) para 177.

199 ibid paras 174–​75.

200 The tribunal cited Art 4.5 of the Algeria–​Italy BIT:

Les nationaux ou personnes morales de l’un des Etats contractants dont les investissements
auront subi des pertes dues à la guerre ou à tout autre conflit armé, révolution, état d’urgence
national ou révolte survenus sur le territoire de l’autre Etat contractant, bénéficient, de la part
de ce dernier, d’un traitement non moins favorable que celui accordé à ses propres nationaux
ou personnes morales ou à ceux de la nation la plus favorisée.
ibid, para 173. In Italian version, the provision appears under paragraph 6.
122  Investment Treaty Protections

follows from this that the basic armed conflict clause in question, despite its quirky
wording common to early Italian BITs,201 limits the non-​discriminatory treatment
to the payment of conflict-​related compensation and does not establish a ground
for finding a state’s liability for those losses.

(b) Advanced armed conflict clause v FPS

Less certain is the added value of the advanced clause when compared to FPS.
Both standards provide protection against physical violence in times of conflict.
In contrast to FPS provisions, the application of advanced armed conflict clauses
is limited to cases in which investment losses are caused by a host state’s ‘forces
or authorities’. This condition may complicate the successful application of the
clause, as it is often unclear during hostilities whether damage was caused by state
or non-​state actors,202 or it may be difficult to prove that acts of members of a state
organ are actually attributable to a host state. The latter challenge was presented to
the tribunal in AMT v Zaire, which had to decide if looting and destruction of the
investor’s property by unpaid Zairean soldiers were the acts committed by the state
forces or authorities within Article IV(2) of the Zaire–​US BIT.203 While the tribunal
refused to apply the advanced clause and ultimately decided the claim on other
legal grounds, it was inclined to conclude that the acts of soldiers were not attrib-
utable to the government since they ‘acted individually without any one being able
to show either that they were organized or that they were under order’.204 A more
detailed and persuasive analysis was offered in a Separate Opinion, in which the ar-
bitrator Golsong argued that according to the rules of attribution under the law of
state responsibility, acts of soldiers who were wearing official uniforms, using army
weapons and vehicles in the course of lootings, and brought the stolen goods back
to the army premises, should indeed be treated as those of the state, and hence the
armed conflict clause ought to have been applied.205
If one subscribes to the view that FPS provisions only cover injuries caused by
non-​state actors, the advanced clause would fill the gap and provide an additional
protection against the physical interference of a host state’s forces. Although some
tribunals supported this narrow interpretation of the FPS,206 the prevailing and

201 See n 146.
202 AAPL (n 23).
203 AMT (n 21) paras 6.15–​19.
204 ibid para 7.08.
205 ibid, Statement of Individual Opinion of Heribert Golsong, para 15. This view reflects ultra vires
acts regulated under Art 7 of the ARS. Golsong supported his arguments with several authorities, e.g.
the Caire Case in which the French–​Mexican Claims Commission found that a murder of an alien by
members of Mexican forces was attributable to Mexico, despite the fact that the perpetrators exceeded
the limits of their competences. The Commission stressed that what was relevant in considering attribu-
tion was that the perpetrators in the course of their acting at least appeared as state officials, or they used
‘powers or measures appropriate for their official character’. See Caire Case (France v Mexico) (1929) 5
UNRIAA 516, 531; Velásquez Rodríguez v Honduras Judgment of the IACtHR of 29 July 1988 para 170.
206 See e.g. El Paso (n 93) para 524; Ulysseas v Ecuador UNCITRAL, Final Award, 12 June 2012, para
272; Eastern Sugar (n 42) paras 203–​07.
 Armed Conflict Clause  123

more convincing view is that the standard provides protection against both non-​
state and state actors.207 This substantive overlap calls for a clarification of the rela-
tionship between the two provisions when included in the same investment treaty.
The AAPL tribunal treated the FPS provision as a more general law vis-​à-​vis
the armed conflict clause. The tribunal appears to have agreed with the view of the
dissenting arbitrator, who held that the FPS provision had been displaced by the
armed conflict clause that was specifically meant to regulate conflict situations.208
Arbitrator Asante argued that the advanced armed conflict clause as lex specialis
‘must prevail as the definitive and exhaustive source of liability in respect of the
conduct of the armed forces of the host State’.209 In other words, if the investor sus-
tained losses due to the host state’s interference in the context of conflict, this provi-
sion would be the only source for the determination of liability. This interpretation
is accurate only in part. It is hard to object to the view that an advanced armed
conflict clause is a more special norm given the different levels of specialty perme-
ating it. The specialty of the provision is reflected in five elements: (1) the type of
situations in which the provision is applicable (only in defined violent situations);
(2) the type of subject that can be the perpetrator of the injury (only the host state’s
forces); (3) the type of injuries for which the provision can be invoked (requisition
and destruction of private property); (4) defined situations from which a state can
escape liability (destruction of property in combat action or when required by ne-
cessity); and (5) the standard of compensation (different standard and methods of
payment of compensation).
This is in stark contrast to a very succinct and broad wording of a FPS provision.
The specificity of the clause indicates that the parties to the treaty wanted the pro-
vision to primarily regulate the situations enclosed therein. This is understandable
due to the politically sensitive nature of the clause as it defines the limits of the
host state’s freedom in protecting its most vital interests. The added value of the
provision for the host state is that, unlike the FPS provision, it expressly provides
for exceptions to the liability for the conduct of the state’s forces in the course of
conflict. It would thus appear safe to conclude that when conflict-​related injuries to
the investor are incurred by the host state’s forces, the primary recourse for remedy
would be the advanced armed conflict clause.
However, contrary to Asante’s argument, it is submitted that the FPS provision
is not made redundant by the armed conflict clause. As a general rule, it continues
to provide protection in cases not encompassed in the advanced armed conflict
clause, in particular in situations when damage has been caused by non-​state
actors. More importantly, the advanced armed conflict clause does not exhaust the

207 See e.g. Biwater (n 27) para 730; Frontier (n 52) para 261; Parkerings (n 52) para 355.
208 AAPL, Dissenting Opinion (n 23) 584.
209 ibid 585. A similar argument was advocated by arbitrator Golsong in AMT case. See Statement of
Individual Opinion (n 21) para 6.
124  Investment Treaty Protections

obligation of the host state to exercise due diligence in protecting investors with
respect to its own organs. The fact that the investor’s property was destroyed in a
combat action or due to the necessity of the situation precludes the finding of li-
ability under the advanced armed conflict clause. However, this should not mean
that the state cannot be held liable for failing to protect investors under the FPS
clause. The due diligence obligation is not limited by the exceptions defined in
armed conflict clauses. It is a broader concept, which, as explained above, also en-
tails an obligation to undertake certain precautionary measures against potential
losses that may emerge in a combat action or out of necessity.210 This can apply
also, or particularly, in situations when it is impossible to ascertain who caused the
injury, but is clear that the damage could have been prevented or minimized had
the state exercised its duty of care, for example by issuing warnings to investors be-
fore launching the forceful operation. While such were the facts in the AAPL case,
the majority did not feel comfortable rejecting Asante’s view that the armed con-
flict clause completely displaced the FPS provision. The tribunal’s indecision led
to the legally inaccurate reliance on the basic armed conflict clause to arrive at the
customary international law standard of due diligence, while the same could have
been achieved by the straightforward application of the FPS provision.

D.  Other Relevant Protections

While FPS provisions and armed conflict clauses are specifically designed to ad-
dress losses sustained due to forcible interference, other treaty provisions may also
provide investors with a remedy, in particular the prohibition of expropriation
without compensation and the FET standard.211 The following sections provide
a brief analysis of these provisions in the light of their relevance in the context of
conflict-​related injuries.

1.  Expropriation

The general rule of customary international law that has evolved over time and
is encapsulated in most treaties allows expropriation under the condition that it
is accompanied by compensation, is non-​discriminatory, pursues a public pur-
pose, and is in accordance with due process.212 Expropriation can be either direct

210 See Section 4 B.1.b. and Chapter 6 C.2.

211 In pending cases emerging from conflicts in Egypt and Ukraine, these provisions have been re-
portedly invoked by investors. See Chapter 1, nn 6–​8.
212 For an overview, see J Cox, Expropriation in Investment Treaty Arbitration (OUP 2019);
Newcombe and Paradell, Law and Practice (n 33) 321–​98; Dolzer and Schreuer, Principles (n 56) 98–​
126; McLachlan et al, Substantive Principles (n 32) 359–​412; Salacuse, Law of Investment Treaties (n
32) 313–​57. See e.g. 2012 US Model BIT Art 6; Canada–​Benin BIT (2013) Art 11.
 Other Relevant Protections  125

(a host state openly seizes the property and transfers title to the private property
to itself) or indirect (a host state’s measures have as their effect the deprivation of
an investor’s property or its benefits). The taking of foreign, in particular enemy,
investors’ assets, has been a frequent measure to which states have resorted in con-
flict and post-​conflict periods. Direct expropriation, which takes place during an
armed conflict, is often described as confiscation. During the post-​conflict period,
which often sees a new regime assume power, expropriations may take the form of
nationalizations, encompassing entire industries of the economy.213
Indirect expropriation must be distinguished from governmental regulatory or
targeted measures not requiring compensation. In the context of conflict, a state
may take a variety of measures that could prevent investors from using or enjoying
their property on the grounds of maintaining public order or protecting national
security. For example, it may sequestrate an investor’s property to prevent its use by
an enemy state, it may physically block access to an investor’s premises, freeze the
investor’s assets, deny visas required for technical staff, or introduce war taxes. Do
such measures amount to indirect expropriation?
As a general rule, the protection against expropriation applies to governmental
measures that deprive investors of their ownership or control of, or substan-
tial benefits from investment, regardless of the form of expropriation (including
nationalization, confiscation, sequestration, seizure, attachment, or any other
measure producing similar effects).214 Arbitral tribunals have attempted to draw
the line between indirect expropriation and non-​compensable measures in a
number of cases, often reaching different conclusions.215 Some have focussed solely
on the impact that the measure had on the investment (‘sole effect’ doctrine),216
while others held that the state’s intent and public purpose of the measure should
be taken into account,217 sometimes relying on proportionality as a technique to
balance the opposing positions.218 The last two approaches provide a fair room for

213 Newcombe and Paradell (n 33) 324. A recent example is the nationalization programme in Egypt,
in which the national courts ordered the post-​revolution government to nationalize the investments
that had been privatized by the overthrown regime. See M Fick, ‘Egypt Drags Its Feet in Privatisation
Tussle’ (Reuters, 29 May 2013) <http://​ in.reuters.com/​article/​2013/​05/​29/​egypt–​renationalisation–​
idINL6N0DX07420130529> accessed 21 March 2015.
214 On the scope of expropriation see e.g. ACP-​CEE 2172/​92 of the Council of the European
Communities, ‘Community Position on Investment Protection Principles in the African, Caribbean
and Pacific States’ (Brussels, 3 November 1992) which was drafted to provide advice for the negotiation
of bilateral treaties, cited in Patrick Mitchell v Democratic Republic of Congo ICSID Case no ARB/​99/​7,
Annulment Decision, 1 November 2006, para 52, n 22.
215 To minimize the unpredictability, some investment treaties stipulate the criteria for indirect ex-
propriation, see e.g. 2012 US Model BIT, Annex B, para 4; 2004 Canada Model BIT, Annex B.13(1).
216 See e.g. Southern Pacific Properties Ltd v Arab Republic of Egypt ICSID Case no ARB/​84/​3, Award,
20 May 1992, para 227; Compañia del Desarrollo de Santa Elena, SA v. Costa Rica ICSID Case no ARB/​
96/​1, Final Award, 17 February 2000, para 72; Metalclad Corp v Mexico ICSID Case no ARB(AF)/​97/​1,
Award, 30 August 2000, para 103; Vivendi (n 27) para 7.5.21; Patrick Mitchell v Democratic Republic of
Congo ICSID Case no ARB/​99/​7, Award, 9 February 2004, para 72.
217 Methanex Corp v United States of America UNCITRAL, Award, 3 August 2005, Part IV, Chapter D,
para 15; Saluka (n 26) paras 254–​65, 275–​76; Chemtura (n 76) para 266.
218 See n 89.
126  Investment Treaty Protections

exemption of a state’s measures taken for security reasons from the obligation to
pay compensation. They are in line with another avenue of defence, namely the
police powers doctrine. As explained above,219 the doctrine according to which the
non-​discriminatory measures taken within the state’s general regulatory or admin-
istrative powers with an aim to protect legitimate public objectives do not qualify
as indirect expropriation, has found wide support in investment jurisprudence, 220
often as a reflection of customary international law. Measures that states take in
order to maintain public order and safety have been typically found to fall within
the police powers exception.
In several historical cases, host states that imposed restrictions on foreign in-
vestment in times of conflict were considered to have been merely exercising their
sovereign police powers and thus no compensation was ordered.221 For example,
the Iran–​US Claims Tribunal decided that setting a limitation on the type of cargo
that investors could unload was a reasonable and legitimate measure during a time
of civil unrest.222 The temporary transfer of enemy-​owned investments under
the administration of the government-​appointed custodian on security or safety
grounds could be similarly justified under the doctrine. Here, an important consid-
eration will be the duration of the measure. As emphasized by the Iran–​US Claims
Tribunal, an appointment of temporary management by a host state will amount
to expropriation unless such ‘deprivation is not merely ephemeral’.223 What is
‘ephemeral’ will depend on the facts of each case.224 For example, in Wena v Egypt,
a forceful seizure and possession of the investor’s hotels for nearly a year was con-
sidered ‘more than an ephemeral interference in the use of that property or with the
enjoyment of its benefits’ and thus compensation was due.225 Similarly, in Mitchell
v Congo, the tribunal held that the sealing of the investor’s premises, seizure of its
documents, and the imprisonment of two employees for more than eight months,
which resulted in the loss of the investor’s clients, could not ‘be qualified as being
of an exclusively transitory nature’ and was thus deemed a measure equivalent to
expropriation.226 In considering the temporariness of the measure, rigid interpret-
ation that is oblivious to the purpose of the measure and the context in which it was

219 See Section 4 B.1.b.

220 See e.g. Marvin Feldman (n 76) para 103; Saluka (n 26) para 255; Lauder (n 27) para 198; Tecmed
(n 18) para 115; Chemtura (n 76) para 266; Quiborax (n 76) para 202; Les Laboratoires Servier, SAS
Biofarma, SAS, Arts et Techniques du Progrès SAS v Poland PCA, UNCITRAL, Final Award, 14 February
2012, para 276; Renée Rose (n 76) para 476; Philip Morris (n 76) para 295.
221 See e.g. Poggioli Case (Italy v Venezuela) (1903) 10 RIAA 669, 691; Parsons Case (Great Britain v
US) (1925) 6 RIAA 165, 166.
222 Sea-​Land Service Inc v Iran (1984) 6 Iran–​USCTR 149.
223 Phelps Dodge v Iran (1986) 10 Iran–​USCTR 130, para 22; Tippetts, Abbett, McCarthym Stratton v
TAMS-​AFFA (1984) 6 Iran–​USCTR 219, 225–​26.
224 Azurix (n 27) para 313; Tecmed (n 18) para 116; SD Myers v Canada UNCITRAL Arbitration, First
Partial Award, 13 November 2000, para 287.
225 Wena (n 16) para 99.
226 Mitchell, Award (n 216) para 72.
 Other Relevant Protections  127

taken should be avoided.227 For example, if a state sequestrates a private investment

with the view to deny potential benefits to the enemy party, such a measure can last
for the whole duration of the conflict, which can be as quick as a few days or as long
as several months or even years. It would be contrary to the practice of war tribu-
nals to consider such measures as indirect expropriation merely because of their
duration.228
In this regard, the Mitchell award,229 in particular, is problematic because the tri-
bunal based its decision solely on the impact the measures had on investment while
ignoring the intent of the government and the public purpose driving the measures
(according to the authorities, the investor had ties with the rebellion against the
government and thus posed a threat to national security), and broader contextual
circumstances (Congo was in a state of war).230 The tribunal concluded that it did
not have enough information to assess whether the government had a power to ex-
ercise its belligerent rights in that situation.231 The more appropriate approach, en-
dorsed in recent arbitral practice, would be to acknowledge Congo’s ‘police powers’
and assess whether they were exercised proportionately (taking account not only of
the duration and economic effect of the measure, but also the legitimacy of the ob-
jective, the ongoing armed conflict, and the emergency in the context of which the
measure was adopted).232 Following this approach, the tribunal would have had to
determine whether the measure was suitable (capable of achieving the objective)
and necessary (whether alternative measures interfering less with the investor’s
right would be possible) before engaging in the balancing exercise (proportionality
stricto sensu) to establish whether the impact of the measure was proportionate to
the aim that the state sought to achieve.
A helpful analogy for further guidance can be drawn from the human rights
case law regarding the non-​conviction-​based confiscation or civil forfeiture which
reflects a state’s police powers. Human rights courts have held that a state authority
can seize an individual’s assets as a precautionary and preventive security measure
as long as it can prove that those assets were likely to be used for facilitating un-
lawful conduct. In order to prevent abusive and arbitrary actions, the human rights
courts have factored certain procedural requirements into the proportionality

227 Achmea BV v Slovakia UNCITRAL, PCA Case no 2008-​13, Final Award, 7 December 2012,
para 289.
228 See Chapter 2 C.1.a, in particular jurisprudence of the Eritrea–​Ethiopia Claims Commission, e.g.
Eritrea’s Civilian Claims (n 166) para 128.
229 The award was later annulled on the grounds that the law firm did not constitute an ‘investment’.
Mitchell, Annulment (n 214) paras 25–​33.
230 Mitchell, Award (n 216) para 74; Mitchell, Annulment (n 214) paras 51, 53, 56.
231 Mitchell, Award (n 216) para 74.
232 For an analysis of how investment tribunals have utilized variants of the proportionality ana-
lysis, see C Henckels, Proportionality and Deference (n 89)  83–​171. See also A  Stone Sweet and J
Mathews, ‘Proportionality Balancing and Global Constitutionalism’ (2008) 47 Colum J Transnatl L
73; E Leonhardsen, ‘Looking for Legitimacy: Exploring Proportionality Analysis in Investment Treaty
Arbitration’ (2012) 3(1) JIDS 95, 112.
128  Investment Treaty Protections

equation.233 In particular, they stressed that such measures had to be adopted by

competent authorities and be supervised by judicial officials in line with due pro-
cess (in Mitchell, the government measures were eventually overruled and the
return of the assets was ordered by the military court);234 that governmental au-
thorities had to provide reasons justifying the measure (the Congolese authorities
provided what they thought were sufficient security reasons justifying the inter-
vention);235 and that there should be no delays in returning the seized assets (this
was a point of disagreement between the parties which the Mitchell tribunal re-
jected as irrelevant).236
Furthermore, in appraising Congo’s measures, the tribunal would have to afford
some deference to the government’s analysis of the situation in which the police
powers were exercised.237 Instead of deferring to the factual and legal assessment of
Congolese military authorities in the course of the armed conflict, the arbitrators
held that they lacked sufficient information to themselves make a retrospective as-
sessment about the legitimacy of the public purpose.238 While such determinations
about the decisions taken in times of security crises are undoubtedly extremely dif-
ficult,239 the tribunal should have reviewed the measure on the basis of information
available to the Congolese authorities on 5 March 1999, the day when the measures
had been taken.240 Given the importance of the protected public interest (national
security), the context in which the measures was taken (war), and arbitrators’ lack
of expertise (none of them was an expert on public international law let alone se-
curity studies), a more deferential approach in ascertaining whether the measures
fell within legitimate police powers would have been more apposite.241 This is not
to suggest that the tribunal should have given the Congolese authorities a blank
check for their action. The fact that the measures included the imprisonment of

233 See Chaparro Álvarez and Lapo Iñiguez v Ecuador Judgment of the IACtHR of 21 November
2017, paras 188, 197, 204, 209; Memoli v Argentina Judgment of the IACtHR of 22 August 2013, para
180; Gogitidze and Others v Georgia App no 36862/​05, Judgment (ECtHR, 12 May 2015) paras 108–​13;
Paulet v The United Kingdom App no 6219/​08, Judgment (ECtHR, 13 May 2014) para 65.
234 Mitchell, Award (n 216) para 62.
235 ibid para 74.
236 ibid para 63.
237 C Henckels, ‘Indirect Expropriation and the Right to Regulate:  Revisiting Proportionality
Analysis and the Standard of Review in Investor-​State Arbitration’ (2012) 15(1) JIEL 223, 240.
238 Mitchell, Award (n 216) para 74.
239 Metalpar SA and Buen Aire SA v Argentine Republic ICISD Case no ARB/​03/​5, Award, 6 June
2008, para 198.
240 The standard that the assessment of a state’s action must be based on the circumstances and know-
ledge of the state or a commander at the time when the action was taken, was confirmed in Nuremberg
trials. See Hostage case (US v List et al) (American Military Tribunal, Nuremberg, 1948) 11 NMT 1253,
1295–​96. See also W Boothby, Law of Targeting (OUP 2012) 172; Y Dinstein, The Conduct of Hostilities
under the Law of International Armed Conflict (3rd edn, CUP 2016) 166; G Corn, ‘Humanitarian
Regulation of Hostilities: The Decisive Element of Context’ (2018) 51(3) VJTL 763, 766; M Newton,
‘Reframing the Proportionality Principle’ (2018) 51(3) VJTL 867, 882; See Chapter 2, n 155.
241 Henckels, ‘Revisiting Proportionality’ (n 237) 245 (observing that more sensitive exercise of def-
erence in situations concerning national security has been endorsed by ECtHR and European Court of
Justice).
 Other Relevant Protections  129

two employees could have importantly informed the proportionality analysis as to

whether police powers were exercised appropriately.242
The last criterion commonly relied on by tribunals when identifying indirect
expropriation is the effect that the measure had on the investor’s reasonable ex-
pectations.243 In other words, investors need to prove that their decision to invest
in the host state was based on a state of affairs which did not anticipate the chal-
lenged governmental measures. Establishing the breach of legitimate and rea-
sonable expectations is particularly difficult when the governmental measure is
taken in the context of a volatile situation. On several occasions, tribunals stressed
that investment treaties do not provide blank insurance for economic and polit-
ical risks.244 In the Iran–​US Claims Tribunal case of Starrett Housing Corp v Iran,
the US housing corporation challenged a series of government actions related to
the Iranian Revolution, including a reduction in the project’s work force owing to
conditions in Iran, strikes and shortages of materials, the collapse of the banking
system, the freezing of bank accounts, harassment of Starrett personnel by armed
Revolutionary Guards, etc. The tribunal rejected the claim that several revolution-​
related governmental measures, either individually or taken together, constituted
expropriation, and noted that:

. . . investors in Iran, like investors in all other countries, have to assume a risk
that the country might experience strikes, lock-​outs, disturbances, changes of
economic and political system and even revolution. That any of these risks ma-
terialised does not necessarily mean that property rights affected by such events
can be deemed to have been taken. A revolution as such does not entitle investors
to compensation under international law.245

In sum, some degree of political risk is inherent to the commercial nature of the in-
vestment, thus the assessment of investment-​backed expectations for the purposes
of determining the breach of the expropriation standard will be reasonably attuned

242 As explained in Chapter 2 D, in the human rights field, in particular, the relevance of protected
interests has been often factored in the assessment of proportionality of the measure. The detention of
investor’s employees (i.e. violation of the right to movement) would thus invite a stricter proportionality
test than solely confiscation of the property. See e.g. De Tommaso v Italy App no 43395/​09, Judgment
(ECtHR, 23 February 2017) Joint Concurring Opinion of Judges Raimondi, Villiger, Šikuta, Keller, and
Kjølbro, para 18. Generally, see J Boucht, The Limits of Asset Confiscation: On the Legitimacy of Extended
Appropriation of Criminal Proceeds (Hart Publishing, Oxford, 2017) 174.
243 See e.g. Marvin Feldman (n 76); Newcombe and Paradell, Law and Practice (n 33) 324. The condi-
tion has been included in some investment treaties, e.g. Canada–​Peru BIT (2006) Annex B.13(1).
244 See e.g. MTD Equity Sdn Bhd and MTD Chile SA v Republic of Chile ICSID Case no ARB/​01/​7,
Award, 25 May 2004, para 178; Maffezini v The Kingdom of Spain ICSID Case no ARB/​97/​7, Award, 13
November 2000, para 64; Waste Management Inc v United Mexican States ICSID Case no ARB(AF)/​00/​
3, Award, 30 April 2004, para 160.
245 Starrett Housing Corp v Iran (Case No 24) (1983) 4 Iran–​USCTR 122, 156.
130  Investment Treaty Protections

with the broader political circumstances of the host state. A similar view has also
been taken with respect to the content of the FET provision.

2.  Fair and Equitable Treatment

The FET standard has been touted as one of the most important provisions in in-
vestment treaties, enabling investors to challenge a wide array of governmental ac-
tions that affect their investment.246 The provision has been invoked by investors
most frequently and with the best record of success.247 Typically, tribunals have
held that FET provides obligations to make all laws, regulations, and policies clear
to foreign investors in advance; to maintain a stable and predictable legal frame-
work; to behave in good faith and respect investors’ legitimate expectations; to
refrain from discrimination; and to provide due process.248 In the context of a con-
flict, a host state’s actions may be challenged with regard to the freezing of assets of
investors coming from an enemy country, the harassment of investors coming from
an enemy country or supporting an enemy regime, and the passing of laws and pol-
icies with a view to adjusting to extraordinary circumstances, among others.
During a conflict, or in the period immediately thereafter, host states often
make the legal and policy changes that are necessary to adjust to their new realities.
Should these legislative measures adversely affect investments, investors may be
poised to invoke the FET provision, arguing that their expectations as to the sta-
bility of the host state’s legal regime have been frustrated. Determining the scope
of the investor’s legitimate expectations will play a critical role. Whether the legit-
imate expectations of the investor will be assessed against the circumstances pre-
vailing in the host state is not completely clear, as demonstrated by the conflicting
conclusions in investment jurisprudence.249 Some tribunals have held that the FET
standard provides a guarantee of a stable legal environment and thus host states

246 For an overview, see R Dolzer, ‘Fair and Equitable Treatment’ (2005) 39 Intl Lawy 87, 90; A
Diehl, The Core Standard in International Investment Protection: Fair and Equitable Treatment (Kluwer
2012); McLachlan et al, Substantive Principles (n 32) 296–​329; Salacuse, Law of Investment Treaties (n
32) 241–​67.
247 According to UNCTAD, the breach of the FET standard has been alleged in 433 cases, and found
in 113 cases. See UNCTAD, Investment Dispute Settlement Navigator, <https://​investmentpolicyhub.
unctad.org/​ISDS/​FilterByBreaches> accessed 18 December 2018.
248 Dolzer and Schreuer, Principles (n 56) 130–​60.
249 Tribunals held that a state’s circumstances do not affect the fair and equitable treatment standard
in e.g. Sempra (n 125) paras 303, 396; Pantechniki (n 21) para 76; Tecmed (n 18) para 154. On the other
hand, tribunals examined the circumstances of the host state in e.g. Bayindir Insaat Turizm Ticaret
Ve Sanayi AS v Islamic Republic of Pakistan ICSID Case no ARB/​03/​29, Award, 27 August 2009, paras
192–​95; Duke Energy Electroquil Partners v Republic of Ecuador ICISD Case no ARB/​04/​19, Award,
18 August 2008, para 340; National Grid (n 125) para 179; Parkerings (n 52) paras 333–​35; Biwater (n
27) para 217. See also the discussion on this subject in N Gallus, ‘The Fair and Equitable Treatment
Standard and the Circumstances of the State’ in C Brown and K Miles (eds), Evolution in Investment
Treaty Law and Arbitration (CUP 2011) 223; U Kriebaum, ‘The Relevance of Economic and Political
Conditions for Protection under Investment Treaties’ (2011) 10 L & Prac of Intl Ct & Trib 383.
 Other Relevant Protections  131

are not allowed to make material changes to law or policy governing foreign in-
vestment.250 Other tribunals have upheld the opposite view. For example, the tri-
bunal in Duke Energy Electroquil Partners v Republic of Ecuador contended that
the investor’s legitimate expectations must be assessed by taking ‘into account all
circumstances, including not only the facts surrounding the investment, but also
the political, socioeconomic, cultural and historical conditions prevailing in the
host State’.251
In reality, much will depend on the type of conduct of the state that is being
challenged. For example, if the ground for invoking the FET standard is the denial
of justice, the fact that the breach occurred in the context of a conflict or a post-​
conflict transition is unlikely to inform the determination of the state’s liability.
As explained by Arbitrator Paulsson in Pantechniki v Albania, regardless of the
circumstances, foreign investors are always ‘entitled to decision-​making which is
neither xenophobic nor arbitrary’.252 Specifically, the ability to abide by this obli-
gation does not hinge on the state’s physical infrastructure, which may indeed be
threatened under extraordinary circumstances, but rather on the ‘human factor of
obedience to the rule of law’ which remains intact even in times of conflict.253
On the other hand, when in times of conflict or transition a host state is prompted
to pass certain measures (e.g. freeze the assets of enemy investors) and make
changes to its laws (e.g. introduce new tariffs or increase taxes), the better approach
is to consider the influence of the broader political and factual circumstances of the
FET. In this vein, the tribunal in Toto v Lebanon held that the investor’s legitimate
expectations as regards the stability of the investment’s environment could not re-
main unaffected by the circumstances in a country in transition (in that period,
there were bomb and terrorist attacks, a war with Israel, and two episodes of in-
tense internal fighting in Lebanon).254 The tribunal thus concluded that ‘the post-​
civil war situation in Lebanon, with substantial economic challenges and colossal
reconstruction efforts, did not justify legal expectations that custom duties would
remain unchanged’.255 Similarly, the tribunal in Bayindir v Pakistan rejected the
investor’s claim that his reasonable and legitimate expectations had been breached,

250 See e.g. Petroleum v Czech Republic UNCITRAL, Final Award, 12 November 2010, para 285;
Tecmed (n 18) para 154.
251 Duke Energy (n 249) para 340.
252 Pantechniki (n 21) para 76.
253 ibid. It is worth reiterating the legal distinction between different types of enforcement activities.
While the state compliance with some enforcement obligations (e.g. investigation and apprehension of
perpetrators as part of the duty to protect) is measured against the flexible standard of due diligence,
the fulfilment of other enforcement obligations (e.g. trial and execution of penalty) is not affected by
external factors. See Pissilo-​Mazzeschi, ‘Due Diligence Rule’ (n 52) 30.
254 Toto Costruzioni Generali SpA v Lebanon ICSID Case no ARB/​07/​12, Award, 7 June 2012, paras
242, 245.
255 ibid para 245.
132  Investment Treaty Protections

by measuring the latter against the ‘volatility of the political conditions prevailing
in Pakistan’.256
In sum, tribunals appear to increasingly support the view that the legal expect-
ations of investors with regard to the scope of protection they are afforded under
the treaty will be assessed in the light of the circumstances during or after the con-
flict. In other words, investors have to adjust their legal expectations to align with
the conditions of the host state. This could enable states to retain some degree of
legal flexibility in dealing with the consequences of an extraordinary burden im-
posed by armed conflict.

E. Preliminary Conclusions

This chapter has shown that the substantive standards purporting to protect in-
vestors’ physical security largely mirror the customary rules on the treatment of
aliens. In the same vein, the tensions between investment expectations, on the
one hand, and the state security interests and their individual circumstances, on
the other hand, echo the nineteenth-​century theories on state responsibility for
conflict-​related injuries to aliens. It has been argued that the treaty protections are
not absolute. While host states are obliged to refrain from forceful interferences,
this obligation is limited by their right to pursue measures necessary for protecting
their security (as ensured by the police powers doctrine, exceptions in advanced
armed conflict clauses etc). Similarly, while they are obliged to protect investors
from the violence of non-​state actors, fulfilment of this obligation is measured
against the relative standard of due diligence. This notwithstanding, inconsistent
and flawed interpretations in many arbitral awards may prompt reliance on other
legal mechanisms with a view to protect states’ freedom to act and regulate in times
of conflict. The next chapter turns the focus to them.

256 Bayindir (n 249) para 193. In a similar vein, the tribunal in Parkerings stated that in evaluating an
investor’s legitimate expectations regarding stability, the political environment at the time of investment
had to be considered. For states in transition, legislative changes are more likely to occur. See Parkerings
(n 52) paras 333–​35.