1 Critical

Success Factors (CSFs) for Ensuring Bankable

2 Completion Risk in PFI/PPP Mega Projects
3
4
5 Hakeem A. Owolabi1, Lukumon O. Oyedele2 C, Hafiz A. Alaka3, Saheed O. Ajayi4,
6 Olugbenga O. Akinade5, Muhammad Bilal6
7
8
9 Affiliations
10 A1 - Dr Hakeem A. Owolabi BSc., MSc., PG.Cert., Ph.D.
11 Associate Professor – Project Analytics and Digital Enterprise
12 Big Data Enterprise and Artificial Intelligence Laboratory, University of West of the England
13 Bristol, Frenchay Campus, Coldhabour lane, Bristol, United Kingdom. BS16 1QY.
14 E-mail: Hakeem.Owolabi@uwe.ac.uk
15
16 C Corresponding Author and Address
17 A2 - Professor Lukumon O. Oyedele BSc., MSc., LLM., MBA., Ph.D.
18 Assistant Vice Chancellor
19 Digital Innovation and Enterprise, University of West of England, Bristol, Frenchay Campus,
20 Coldhabour lane, Bristol, United Kingdom. BS16 1QY
21 E-mail: Ayolook2001@yahoo.co.uk; L.Oyedele@uwe.ac.uk
22
23 Affiliations
24 A3 - Dr Hafiz A. Alaka BSc., MSc., HND., Ph.D.
25 Senior Lecturer in Civil Engineering Management
26 Faculty of Energy, Construction & Environment, Coventry University, Priority Street, Coventry,
27 United Kingdom. CV1 5FB. Email: Hafiz.alaka@coventry.ac.uk
28
29 Affiliations
30 A4 - Dr Saheed O.Ajayi BSc., MSc., Ph.D
31 Senior Lecturer, School of Built Environment and Engineering, Leeds Beckett University
32 Leeds, City Campus, Leeds - United Kingdom. LS1 3HE. Email: Saheed.ajayi@leedsbeckett.ac.uk
33
34 Affiliations
35 A5 - Dr Olugbenga O. AKinade BSc., MSc., Ph.D
36 Associate Professor – Big Data Application Development
37 Big Data Enterprise and Artificial Intelligence Laboratory, University of West of the England
38 Bristol, Frenchay Campus, Coldhabour lane, Bristol, United Kingdom. BS16 1QY.
39 E-mail: olugbenga.akinade@uwe.ac.uk
40
41 Affiliations
42 A6 - Dr Muhammad Bilal BSc., MSc., Ph.D
43 Associate Professor – Big Data Application Development
44 Big Data Enterprise and Artificial Intelligence Laboratory
45 University of West of the England, Bristol, Frenchay Campus,
46 Coldhabour lane, Bristol, United Kingdom. BS16 1QY.
47 E-mail: muhammad.bilal@uwe.ac.uk

1
48 Abstract

49 This study investigates project financiers’ perspectives on the bankability of completion risk in Private
50 Finance Initiatives and Public Private Partnerships (PFI/PPP) mega projects. Using a mixed methodology
51 approach, focus group discussions with financier stakeholders in UK’s PFI/PPP industry were used to
52 identify 23 criteria relevant for evaluating completion risk in funding applications. These criteria were put
53 in a questionnaire survey to wider audiences of financiers of PFI/PPP projects in the UK. Series of
54 statistical tests were performed, including Reliability Analysis, Kruskal-Wallis Non-Parametric Test,
55 Descriptive Statistics, Principal Rank Agreement Factor (PRAF) and Regressions Analysis. After
56 identifying 21 reliable criteria influencing the bankability of completion risk, the general agreement of
57 three major financier stakeholders (Senior Lenders, Equity Financiers and Infrastructure Financiers) on
58 all the criteria were examined through Kruskal-Wallis test and PRAF. A regression model, constructed
59 and validated with input from another team of expert financiers, revealed five key criteria influencing the
60 bankability of completion risk in PPP mega projects. These include (1) Construction contractor with years
61 of experience of successful completion of mega projects, (2) Construction Contractor’s financial strength,
62 (3) Existence of Tried-and Test Technology for the construction of project, (4) Availability of Independent
63 Technical Consultant (ITC) and (5) Existence of Fixed Price Turn Key (FPTK) construction contract.
64 The research findings will provide PFI/PPP contractors and clients with valuable strategies for satisfying
65 financiers’ requirements in delivering large-scale Infrastructure PPP projects.
66
67 Keywords: Bankability; Risk; Public Private Partnership (PPP); Private Finance Initiatives (PFI); Mega
68 Projects; Financiers’ Perspective.

2
69 Background

70 Private Finance Initiatives and Public Private Partnerships (PFI/PPP) in mega projects has received
71 increased global attention since the last decade (Kennedy, 2015, Sainati et al., 2017; Owolabi et al., 2018).
72 With increasing scope and size of civil engineering infrastructures, project finance has gradually entered
73 the “tera era” where projects worth trillions of dollars ($) are being delivered across Europe, America and
74 some emerging economies (Flyvbjerg, 2014). According to Flyvbjerg (2014), the annual total global
75 spending on mega projects currently ranges between US$6 trillion to US$9trillion (representing 8% of
76 global GDP). Mega projects are described as multi-billion dollar large-scale projects, involving multiple
77 stakeholders within governments and private sectors (Giezen et al., 2015). From sectors such as energy
78 to water, mining, information technology, urban regeneration, etc., these new-breed of capital-intensive
79 projects are seen as the promise of the future (Boateng et al., 2015; Grabovy and Orlov, 2016). However,
80 like most complex and large-scale infrastructure projects, a major concern for stakeholders, especially
81 project financiers on PPP megaprojects is the bankability of completion risk (Fithali and Ibrahim, 2015;
82 Moser, 2016). By bankability here, we refer to the willingness of lenders to finance a project after due
83 consideration of its risks and returns (Delmon, 2015).
84

85 Completion risk, which also refers to project delay or time overrun in many studies, may be described as
86 the risk that a project may not be completed to time, specification and within agreed budget (Gatzert and
87 Kosub, 2016; Budaya, 2018; Song et al., 2018). According to the February 2016 report of McKinsey
88 Consulting on global construction productivity, completion risk remains the key driver of cost overrun in
89 most construction and engineering projects, with 77% of mega projects delayed by at least 40% of the
90 time. Similar report from KPMG’s 2015 Global Construction Industry Survey also suggested that, only a
91 quarter of construction projects, out of a sample of 109 construction organisations came within 10 percent
92 of their initial deadlines; with delay dispute claims averaging a staggering US$46million (Lepage, 2017).
93 In the context of PPP mega-projects, the recent European Court of Auditors’ report of 2018 also gave a
94 damning verdict of excessive schedule delay in most EU-led PPP projects; with seven out of nine mega-
95 projects (worth €7.8billion) exceeding deadlines by up to 52months and resulting in massive cost
96 overrun.
97

98 From project financiers’ perspective, the adverse impact of delay in PPP projects can be damaging and
99 far-reaching (Domingues and Zlatkovic, 2015). According to Morrison (2016), asides the effect of cost
100 overrun, completion risk can result in difficult issues such as delay in realisation of project’s operating

3
101 revenue, longer debt service repayment period and distorted financing arrangements with project lenders.
102 Other implications of delay in PPP include liquidated and ascertained damages; accumulated interest on
103 project loans, undue lock-down of lenders’ investment among others (Hodge and Greve, 2017; Owolabi
104 et al., 2018). As such, given the high-risk profile of most PPP mega-projects especially at the construction
105 phase (see Fig. 1 for Risk Profile of PPP Projects during Project Life Cycle), the limited recourse nature
106 of its financing (Aladağ and Işik, 2017), vis-à-vis bank’s relatively limited in-house technical skills needed
107 for accurate estimation of project delay during funding appraisal (Chowdhury et al., 2015; Kumar et al.,
108 2018), a key decision for lenders which is often overlooked in most PPP literature is, how do financers’
109 evaluate and determine whether the risk of project incompletion is acceptable/bankable to them?
110 (Özdemir, 2015).
111

Project Risk Construction Phase of the Project

Pre- Construction Phase Post-Construction Phase

112
113 Project Timeline
114
115 Fig.1 Risk Profile of PPP Project during Project Life Cycle
116
117 Recent review of PPP literature has uncovered a dearth in studies on completion risk evaluation, especially
118 from project financiers' perspectives regarding completion risk. For instance, whilst many studies have
119 explored risk assessment and modelling in PPP, most views have often focused on client, project sponsors
120 and contactors’ perspectives (kennedy, 2015; Amidu, 2017; Song et al., 2017; Budayan, 2018), with
121 limited concern for bankability of risks (Fathali and Ibrahim, 2015; Moser, 2016). Although, Critical
122 Success Factors (CFS) for PPP is also a common theme within this research domain, however, articles on
123 CSFs often emerge with the aim of identifying generic drivers of PPP in different climes, without in-depth
124 attention to completion risk evaluation and its impact on financiers’ investments (Wibowo and Alfen,
125 2015; Osei-Kyei, and Chan, 2015; Liu et al., 2016; Chou and Pramudawardhani, 2015; Osei-Kyei and

4
126 Chan, 2017). Other similar studies on PPP have also concentrated on examining comparative analysis of
127 PPP performances across nations like China, Australia, UK, Indonesia including Singapore and Turkey
128 among others (Chou and Pramudawardhani, 2015; Liu et al., 2016; Van den Hurk et al., 2016). In addition,
129 existing studies on schedule delay in PPP have been described as too fixated on identifying causative
130 factors of time and cost overrun and are believed to be too deterministic in approach (Owolabi et al., 2018;
131 Kokkaew and Chiara, 2010; Kokkaew and Wipulanusat, 2014). According to Ortiz-Pimiento and Diaz-
132 Serna (2018), current perspectives on delay in PPP projects are mostly contextualised to different
133 countries and often emerge from the perspectives of other PPP practitioners except project financiers.
134 Although, there appears a growing increase in the studies on mega-projects (Giezen et al., 2015; Kennedy,
135 2015; Larsen et al., 2015; Aladağ and Işik, 2017), most of the literature are either centred on exploring
136 Mega-project as a concept (Flyvbjerg, 2014; Mok et al., 2015; Hannan and Sutherland, 2015), not focused
137 on PPP contexts (Boateng et al., 2015; He et al., 2015) or concentrating on sector-specific performance
138 evaluation as well as complexities associated with such large-scale projects (Hannan and Sutherland,
139 2015; He et al., 2015; Aladağ and Işik, 2017; Lundrigan et al., 2015). In most instances, literature on mega
140 projects have prioritised investigating few isolated case studies of projects without much attention to the
141 financial impact of the delay on project financiers (Hannan and Sutherland, 2015; Lundrigan et al., 2015;
142 Brooks and Rich, 2016).
143
144
145 Nevertheless, despite the contributions of the above studies, there is currently a clear and noticeable gap
146 in knowledge, indicating that most studies have overlooked project financiers’ perspectives to the pre-
147 contract evaluation of completion risk in PPP mega-projects, especially as it affects the efforts to raise the
148 much-needed debt capital that is critical for its successful delivery. This study therefore emerged as a very
149 significant contribution to the literature within engineering and construction PPP domain. The study
150 addresses practitioners’ concerns over lack of clarity regarding lenders views on critical risk
151 and other factors influencing financiers’ decisions when determining whether risks are
152 bankable/acceptable in a PPP funding deal. This lack of insight from lenders’ frame of mind
153 has been highlighted as one of the key reasons why many laudable potential PPP projects have
154 not seen the light of the day due to poor financial structuring (Moser, 2015; Amidu, 2017). But,
155 more importantly, with the unceasing dismal reputation of the construction industry on time
156 and cost performance, especially in mega-projects. As well as the increasing loss of motivation
157 for long-term infrastructure financing by many project lenders, better understanding of
158 bankability of risks and its structuring are critical for construction and engineering
159 practitioners, for convincing financiers and winning funding approval PPP projects.

5
160
161 Additionally, whilst this study acknowledges that bankability varies and may involve broader macro-
162 economic conditions such as economic and political stability of project’s host nation, legal and regulatory
163 conditions, including more generic factors such as reliable public sector, experienced private sector party,
164 smart financing structure, etc. However, this study is only limited to investigating how completion risk in
165 mega PPP projects can be made bankable/acceptable to project lenders at the financial engineering and
166 appraisal stage, by focusing on specific bankability requirements (See Fig. 2 below for the Main Focus of
167 the Study). Hence, the central hypothesis behind this study is that, “there are some critical
168 bankability criteria that strongly influence financiers’ decision when evaluating the risk of
169 incompletion in PPP mega-project deals”. “And that, perspectives on these critical factors may
170 vary across different financier participants.”
171

172
173 Fig.2 Main Focus of the Study
174
175 Therefore, the overall aim of this study is to examine the perspectives of project financiers’ in the UK on
176 the essential criteria for evaluating bankability of completion risk in PFI/PPP megaprojects. Based on the
177 above aim, the objectives of the study include:
178 1. To identify top-ranked criteria influencing the bankability of completion risk in funding
179 applications for PPP megaprojects.

6
180 2. To compare perceptions and understand patterns of agreement on the identified bankability
181 criteria among various financial stakeholder groups (senior lenders, infrastructure financiers, and
182 equity financiers).
183 3. To identify the key criteria influencing the bankability of completion risk in funding applications
184 for PPP megaprojects based on the perception of the three stakeholders.
185 This paper is laid out in the following order. The next section of the paper is the literature review section
186 and examines completion risk and its drivers in PPP mega projects. This is then followed by the
187 methodology section, which employs mixed methodological approach (Focus group and questionnaire
188 survey to UK project lenders and other project finance experts) towards examining the phenomenon.
189 Immediately after the methodology section is the qualitative data analysis; which was carried out using
190 thematic analysis. This is then followed by quantitative data analysis of questionnaires distributed to
191 project lenders and other project finance experts in the UK. Following the data analysis section is the
192 discussion of major findings within the study. The implications of the research findings for construction
193 and engineering practitioners, especially those involved in PFI/PPP projects were also discussed. The final
194 section concludes the paper.
195
196 Completion Risk in PFI/PPP Mega Projects and Bankability

197 Risk analysis and management is an essential part of decision-making process for funding Private Finance
198 Initiatives and Public Private Partnerships (PFI/PPP) projects (Aladağ, and Işik, 2017). Al Bahar et al.
199 (1990) define risk as: "The exposure to the chance of occurrences of events which may adversely or
200 favourably affect project objectives as a consequence of uncertainty”. According to Moser (2016),
201 although, every human activity is, to an extent, characterised by various forms of risks. However,
202 modernisation has brought the delivery of more complex and large-scale projects, thereby resulting in
203 increasing potential for risks to project stakeholders (Delmon, 2015). Going by these perspectives, one of
204 the most critical risks in PPP projects is the risk that a project may not be completed, in spite huge capital
205 investments involved (Xu et al., 2015). To most project participants, especially the financiers, funding a
206 project with unbankable completion risk represents a plunge down the abyss (Moser, 2016).
207
208 Speaking generally, the riskiest stage of project undertakings in PPP arrangements is the construction
209 phase (Budayan, 2018; Owolabi et al., 2018). According to Owolabi et al. (2018), various forms of risk
210 events often account for the high-risk profile of PPP projects at the construction stage. These risks in most
211 cases pose threats to project completion. Studies such as Amoatey et al. (2015); Larsen et al. (2015); Liu
212 et al. (2016); Budayan, (2018); Owolabi et al. (2018) among others have identified factors that may cause

7
213 project incompletion, including extreme or poor weather condition, poor design of project, cost overrun,
214 delayed access to project site, etc. (See Table 1. Below for factors that may influence project incompletion
215 at the construction stage).
216
217
218
219 Considering the nature of these risks factors and the huge uncertainty they bring into projects' construction
220 processes, financiers are often much more careful in providing financial backing, even if the project is
221 lucrative from a commercial point of view (Mills, 2010). In addition, the poor reputation of the
222 construction industry for coping with construction-related risks suggests the need for more rigorous
223 financing considerations from the financiers' point of view (Zou et al., 2007; Le-Hoai et al., 2008).
224 However, in spite numerous researches on completion risk analysis in PPP projects (Kokkaew and Chiara,
225 2014; Bing et al., 2005; Owolabi et al., 2018; Zhang, 2007; Tam and Fung, 2008), financiers’ perspectives
226 on key criteria influencing bankability of completion risk PPP megaprojects remain unexplored. For
227 instance, in a recent review literature on delay in PPP projects, Budayan (2018) examined the
228 perception of consultants, project sponsors and public sector on causes of delay in BOT projects
229 in Turkey, by relying on Analytical Hierarchical Process (AHP). The study identified “certainty
230 in political and governmental issues” and “reduction in design changes” as key factors to
231 consider for minimising completion risk in Turkish PPP projects. Similarly, Song et al. (2017)
232 identified factors responsible for completion risk and early termination of PPP contracts in
233 China, with “government decision error” and “government payment default” seen as the most
234 factors influencing PPP project completion in China. Also, in another related study, Owolabi
235 et al. (2018) examined a big data analytics approach to predicting completion risk in large
236 portfolio of PPP projects by comparing the predictive power and accuracy of five big data
237 algorithms. These include, Linear Regression, Random Forest, Support Vector Machine,
238 Regression Trees, and Deep learning, with the study suggesting Random forest as the best
239 algorithm. Other related studies such as Larsen et al. (2015); Amoatey et al. (2015); Perera et
240 al. (2016), Ortiz-Pimiento and Diaz-Serna, (2018) and Kokkaew and Wipulanusat (2014) have
241 also examined other issues relating to delay in PPP projects. However, despite the significant
242 contributions of the above literature on delay in PPP literature, most of these studies have not
243 emerged from project financiers’ perspectives.
244
245 Similarly, Osei-Kyei and Chan (2015) in a study on PPP in Ghana, conducted a review of
246 literatures on CSFs for implementing PPP projects. The study uncovered top CSFs for PPP
247 application to include risk allocation and sharing, strong private consortium, political support,

8
248 Table 1: Factors Influencing Completion Risk in Mega PFI/PPP Projects
No Factors Influencing Completion Risk in Mega PFI/PPP Literature Sources
1 Defective design of project Projects Davis et al. (1989); Burati et al. (1992); Gransberg and Molenaar (2004).
2 Projects’ cost overrun Kaming et al., (1997); Dikmen et al., (2007); Flyvbjerg et al., (2004); Semple et al. (1994)
3 Ground conditions (geology/ground water) Sanger and Sayles (1979); Van Staveren (2006); Fookes et al., (1985); Kangari (1995)
4 Cost/impact of delay Yang and Wei (2010); Odeh and Battaineh (2002); Assaf et al. (1995); Le-Hoai et al. (2008)
5 Building area Ching (2014); Allen and Iano (2011); Tolman (1999)
6 Sub-standard subcontractors Eccles (1981); Odeh and Battaineh (2002); Errasti et al., (2007)
7 challenges with innovation in construction techniques Tatum (1987); Harty (2005); Tatum (1989); Bossink (2004)
8 Extreme or poor weather True (1998); Kaming et al., (1997); Moselhi et al., (1997); Odeh and Battaineh (2002)
9 Delayed access to project site Fan et al. (1989); Mustafa and Al-Bahar (1991); Sun and Meng (2009)
10 Material and equipment shortage Baloi and Price (2003); Kittusamy and Buchholz (2004); Teizer et al. (2010)
12 Site safety and security Mohamed (2002); Tam et al. 2004; Fung et al. (2010); Carter and Smith (2006)
13 Bankruptcy of construction firm El-Sayegh (2008); Russell and Jaselskis (1992); Ling and Hoi (2006); Dissanayaka, and
14 Delay in project start up Kumaraswamy (1999)
Bing et al. (2005); Aibinu and Jagboro (2002); Sun and Meng (2009); Tiong (1990)
15 Poor maintain of construction technology Hendrickson and Au (1989); Rousseau and Libuser (1997); Shen et al. (2007); Tam and Fung
16 Delay or failure to secure necessary planning permits (2008)
Ng and Loosemore (2007); Mezher and Tawil (1998); Ahmed et al. (1999); El-Sayegh (2008).
17 Delayed dispute resolution Robinson and Scott (2009); Javed et al. (2013); Tam et al. (2004)
18 Inaccuracy of construction material estimates Zou et al. (2007); Le-Hoai et al. (2008); Baloi and Price (2003); Shane et al. (2009)
19 Defective work and mistakes Kangari (1995); Dikmen et al., (2007); Flyvbjerg et al., (2004); Kaming et al., (1997); Moselhi et
20 Changes in government regulations/ tax rate changes al., (1997). (2008); Russell and Jaselskis (1992); Kangari (1995); Bossink (2004)
El-Sayegh
21 Natural Disaster Gransberg and Molenaar (2004); Odeh and Battaineh (2002); Assaf et al. (1995)

249

9
250 community/public support and transparent procurement. In another related study, Liu et al.
251 (2016) conducted a comparative analysis of critical success factors (CSF) influencing the
252 efficiency and effectiveness of the tendering process for PPPs in Australia and China. Using
253 literature review, interviews and survey, the study unravelled robustness of business case
254 development, quality of project brief among others, as key factors determining efficient and
255 effective PPP tendering process. Wibowo and Alfen (2015); Chou and Pramudawardhani
256 (2015) and Osei-Kyei and Chan (2017) have also all identified critical drivers of PPP in
257 Indonesia, Ghana, Singapore and Taiwan respectively. However, despite the efforts of these
258 various studies, project financiers’ perspectives to completion risk in mega PPP deals remain
259 a noticeable gap in literature, which many studies have overlooked, and is therefore being
260 considered in this study.

261 Methodology
262 To ensure in-depth understanding of the research phenomenon while also facilitating its wider
263 applicability, this study adopted exploratory sequential mixed methodology approach to research. With
264 this strategy, initial exploration of the phenomenon through qualitative research approach was followed
265 with a quantitative approach. According to Creswell and Clark (2017), a sequential mixed method is
266 suitable where a phenomenon is yet to be conceptualised, adequately explored in the literature or is being
267 examined in a context whose research questions are unknown. In this regard, the qualitative phase of the
268 study involved focus group interviews with experienced financier stakeholders involved in Private
269 Finance Initiatives and Public Private Partnerships (PFI/PPP) megaprojects in the UK. This exploratory
270 approach was adopted to identify a broad range of criteria influencing the bankability of completion risk
271 and to confirm the generalisability of the criteria. The focused interviews also enabled the research team
272 to explore in-depth understanding and perceptions of key financial stakeholders, i.e., senior lenders, equity
273 financiers, infrastructure financiers, and hedge fund managers on the factors influencing bankability of
274 completion risk in PFI/PPP funding applications. Considering the need for information-rich participants
275 (i.e. financiers with prior experience in PFI/PPP project financing deals), the study employed purposive
276 sampling strategy to select the interview participants. Patton (1990) described purposive sampling method
277 as a non-probabilistic sampling with which the researcher carefully selects information-rich cases or
278 participants by relying on well-thought out selection criteria. This sampling method allows the researcher
279 to use his or her judgement to make decisions on the suitability of research participant, based on their
280 richness in terms of information, the information need of the research and the nature if the research
281 questions (Suri, 2011).
282

10
283 As agued by Moustakas (1998), in conducting a robust qualitative enquiry using interviews, a minimum
284 of 5 and maximum of 25 interviews may be suitable. Relying on this perspective, this study conducted
285 five (5) focus group interviews with financiers who boast vast experience in structuring PFI/PPP loans.
286 While the focus group interviews facilitated data collection within a shorter time-frame from participants
287 who inter-subjectively build on one another’s perspectives (Lederman, 1990), exploration of commonly
288 shared views of the participants regarding the phenomenon was also facilitated. A total number of
289 nineteen (19) participants were involved in the five focus group interviews, with all having an average of
290 12.4years of experience in PFI/PPP financing. The focused interviews were moderated by an experienced
291 researcher who was able to explore various perspectives to issues determining the bankability of
292 construction and completion related risks in PFI/PPP project appraisals. The entire focus group interviews
293 lasted an average total of 34.5minutes. Additionally, all the discussions were tape-recorded and
294 transcribed using Nvivo10 software. This software allowed the creation of various nodes which aided the
295 coding of emergent themes from the data transcript. After thorough analysis, the study identified 23
296 relevant bankability criteria used by financiers to decide the bankability of completion risk in PFI mega
297 projects.
298
299 The second phase of the study involved quantitative data collection. As part of the objective of the study,
300 which aimed at confirming the wider applicability of the research findings, the 23 bankability criteria
301 identified through focus group interviews were put together in a questionnaire survey. The survey was
302 designed to generate more reliable findings from wider audiences of project financiers and other subject
303 matter experts in UK’s PFI/PPP industry. Using a random sampling technique, a list of 225 financial,
304 contracting and consulting firms were identified and collated from the PFI/PPP projects’ database
305 provided by the HM Treasury. This list comprised hedge funders, pension fund administrators, project
306 finance consultants, senior lenders, infrastructure financiers, equity investment firms, etc. However,
307 before distributing the questionnaire, the research team conducted a pilot study to ensure the adequacy of
308 the research instrument. The pilot study involved four senior lenders (members of staffs of banks) and one
309 academic in the UK who all volunteered to evaluate the questionnaire. Their average experience in project
310 finance was 6.5years. The two major feedbacks, which include rephrasing of questions and re-scaling of
311 questions not answered as expected, were carried out. In developing the final questionnaire, participants
312 were asked to rank each bankability criterion in the questionnaire based on their perceived significance in
313 influencing financiers’ consideration for completion risk in PFI/PPP mega project appraisal. This was
314 carried out on a five-point Likert scale, where 1 represented “Not Important” and 5, “Most Important”.
315

11
316 After that, a large-scale distribution of the questionnaires was conducted. This was done via email with
317 185 questionnaires distributed to senior lenders, equity investment firms, infrastructure financiers, hedge
318 fund managers, etc. Each questionnaire was accompanied with a letter of introduction/statement of intent
319 to introduce respondents to the study, including its aim and objectives. Several reminder emails, which
320 lasted a period of 1-year, 7months, between January 2016 and July 2017 were sent to the respondents.
321 Out of the 185 questionnaires distributed, 109 were returned, representing 58% rate of return. This rate
322 of return was considered suitable for analysis given the claim by Oyedele (2012) who argued that any
323 survey return rate that is lower than 30 to 40% might be regarded as biased and of little significance.
324 Additionally, six (6) out of the 109 questionnaires returned were found to be incomplete and so were
325 considered unsuitable for analysis. These were immediately removed, leaving us with 103 usable
326 questionnaires from senior lenders, infrastructure financiers, hedge fund managers, equity financiers, etc.
327 Out of the 103 questionnaires, 43 represents senior lenders, 21 were equity financiers, 34 were
328 infrastructure financiers while 5 were hedge fund managers (see Table 2 for Demographics of Survey
329 Respondents)
330 Table 2: Demographics of Survey Respondents
Variables Sample Size
Total Number of Respondents 103
Type of Organisation
§ Senior lenders (Staff Members of banks) 43
§ Infrastructure Financiers 34
§ Equity Financiers 21
§ Hedge Fund managers 5
Years of Experience in PPP Project Finance
§ <1 5
§ 1-5 18
§ 6-10 33
331
332
333 All the participants have an average of 10.9 years in PFI/PPP megaprojects both in the UK and
334 internationally. With the aid of SPSS, the results of the questionnaire survey were analysed. Statistical
335 tests such as, Reliability Analysis, Kruskal-Wallis Non-Parametric Test, Descriptive Statistics, Principal
336 Rank Agreement Factor (PRAF) and Regressions Analysis were carried out on the data.
337 Data Analysis
338 Qualitative Data Analysis

339 In order to analyse the qualitative data collated from focus group interviews, a thematic analytical
340 approach was adopted for the study. Being a content-driven technique, thematic analysis enables
341 exhaustive comparison of all segments of qualitative data to identify relationships and structures among

12
342 recurring themes (Aronson, 1995; Braun et al. 2014). Using Nvivo 10, the focus group interviews with
343 participants were transcribed, while the interview transcripts were printed out and proofread for errors and
344 possible omissions. Thereafter, initial coding of the data was carried out by considering the descriptive
345 terminologies used by interviewees during the focus group discussions. This helps to improve the
346 dependability of the analysis as suggested by Kerr and Beech (2015). The thematic analysis was then
347 carried out using a structured coding scheme to unravel the various issues relating to bankability of
348 completion risk in funding applications for Private Finance Initiatives and Public Private Partnerships
349 (PFI/PPP) megaprojects. The coding scheme focuses on three main areas namely, sources, context and
350 theme category. While the source identifies the discussant, who initiates the transcript segment, the theme
351 category summarises the important issues discussed within the quotation segment. Table 3 below shows
352 the example of the quotation classification based on coding scheme.

353 Table 3: Sample of Classification based on the Coding Scheme

No. Quotation Source Theme Context Theme category

1. “In most cases, big construction Discussant 4 Experienced Construction

firms with vast experience and construction Contractor
financial strength are often the brain contractor should be Competence
behind such projects. But the engaged
important thing is to have a
competent contractor with good track
record.”

2. “There are definitely a host of risk Discussant 17 Construction& Robust Risk

mitigations strategies that can be Completion risk Mitigation
used to sway project financiers. You must be mitigated Strategies
need to identify the right ones for
your negotiations, and it all depends
on how much you intend to convince
the financiers of the viability of the
project".

13
3. “The important issue is, get a good Discussant 13 Much will be Performance-driven
construction contractor, and tie him required of the Penalties and
to a performance contract so that he contractor regarding Incentives
can be held accountable.” performance

4. "In the case of such complex Discussant 1 A good procurement Strong

engineering projects, you need a contract is essential Construction
strong procurement contract to Procurement
deliver within time and budget. Every Contract
single contract clause is essential,
and you need the construction
contractor to agree to some
commitments in terms of risk and the
354 likes.

355 At the end of the qualitative data analysis, the study identified 23 criteria relevant for appraising the
356 bankability of completion risk in PFI/PPP mega project deals (see Table 4 for bankability Criteria for
357 Evaluating Construction Risk in PFI/PPP Loan Applications).
358
359 Completion Risk Bankability Framework
360
361 Based on the identified criteria for evaluating bankability of completion risk in Private Finance Initiatives
362 and Public Private Partnerships (PFI/PPP) mega projects, the study developed a qualitative framework.
363 The framework is thus presented in Fig 3 below.
364
365 Quantitative Data Analysis:
366
367 The quantitative phase of the data analysis was carried out using SPSS. Although few alternative statistical
368 approaches were considered for this study i.e. the use of Significance-Index method in place of Mean-
369 Test for descriptive statistics, Factor Analysis for identifying key underlying structures in the dataset, as
370 against multiple linear regression analysis. However, the researcher was more concerned with adopting
371 approaches that best deliver the objectives of the study. Hence, the quantitative data analytical techniques
372 employed in this study include Reliability Analysis, Descriptive Statistics-Mean Test, Kruskal Wallis,
373 Principal Rank Agreement Factor (PRAF) and Regression Analysis. Below is a brief description of these
374 statistical techniques and the various hypotheses behind their application in the study:

14
375 Table 4: Criteria for Evaluating the Bankability of Construction & Completion Risk in PFI/PPP Project Loan Applications
Bankability Criteria for Evaluating Construction & Completion Risk in PFI/PPP Project Loan Applications Focus Groups
1 2 3 4 5

Construction Contractor’s Competence

1 Existence of Tried-and Test Technology for the construction of project. ü ü ü ü ü
2 Construction Contractor’s liability insurance cover ü ü ü
3 Construction contractor’s years of experience of successful completion of mega projects ü ü ü ü
4 Construction Contractor’s financial strength ü ü ü ü ü
5 Existing cost liability or debt commitments of the project to other creditors different from the lender ü ü ü
Robust Risk Mitigation Strategies
6 Pre-Completion Guarantee or Full Financial Guarantee from the sponsor at construction stage ü ü ü ü ü
7 Delay in start-up insurance to prevent cost and time-overrun ü ü ü
8 Existence of bank-financed construction cost overrun facilities ü ü ü
9 Contingent equity contribution from the project sponsors in case of cost overrun ü ü ü ü
10 Debt Buy Out arrangement ü ü ü ü
11 Full injection of equity funds by project sponsors at the start of the construction phase ü ü
Strong Construction Procurement Contract
12 Construction contractor to accept “Single -Point Responsibility” on other project subcontractors ü ü ü
13 Construction subcontract must represent very high value to the subcontractor ü ü ü ü
14 Construction contractor to accept Full Technology Wrap for the proper functioning of all project assets after construction ü ü ü ü ü
15 Availability of Independent Technical Consultant (ITC) ü ü ü ü
16 Fixed Price Turn Key (FPTK) contract ü ü ü ü
17 Project contract to introduce benchmarking arrangements ü ü ü
18 Contractor must accept exceedingly high liability caps ü ü
Performance-based Contract (Incentives and Penalties)
19 Construction contractor to must deliver exceedingly high performance and retention support ü ü ü
20 Contractor must handle the construction program and schedule in a conservative way ü ü ü ü
21 Contractual commitment to project’s output specifications and deliverables ü ü ü
22 Existence of clearly stated and objectively testable construction completion test requirements ü ü ü
23 Existence of liquidate damages for construction performance failures ü ü ü ü ü
376
15
377
378 Fig. 3 Framework for evaluating the bankability of construction and completion risk in PFI/PPP mega projects

16
379

380 1. Reliability analysis: is a statistical approach used in examining the consistency of the
381 measurement Likert scale used in the questionnaire, with the construct that is being
382 measured. In this study, we employed reliability analysis to confirm whether all the
383 criteria identified for evaluating completion risk truly measures the construct they
384 are expected to measure. The rule of thumb for reliability analysis is, since
385 Cronbach’s alpha coefficient is usually between 0-1, any value between 0.7 upward
386 is considered a good reliability of the data (Oyedele, 2013). Hence, we adopt the
387 following null and alternative hypotheses below.
388
389 H0: All identified bankability criteria for evaluating completion risk are
390 true measures of the construct.
391 H1: Not all the bankability criteria for evaluating completion risk are true
392 measures of the construct.
393 2. Descriptive statistics: the use of descriptive statistics in this study was focused on
394 identifying the top-ranked financiers’ criteria for evaluating construction and
395 completion risk in funding applications for PFI/PPP megaprojects. A mean ranking
396 approach was adopted in this case with top-ranked criteria arranged based on their
397 mean coefficient (between 0-5).
398
399 3. Comparison of groups: Comparison of ranking among respondent groups was
400 carried out using Kruskal-Wallis test of significance. Being, a non-parametric
401 statistical approach, Kruskal-Wallis test examines the statistical differences in
402 opinion among two or more independent groups in a study (Fowler et al. 2013). In
403 this study, we examined whether all the three categories of respondents (Senior
404 Lenders, Equity Investors, and Infrastructure Financiers) perceived the criteria
405 similarly or differently, based on their respective ranking in the questionnaire.
406 Hence, the following null and alternative hypotheses below were developed:
407
408 H0: There is no differences in research participants’ perception of all the
409 identified bankability criteria similarly.
410
411 H1: There is a difference in research participants’ perception of all the
412 identified bankability criteria similarly.
413

17
414 4. Principal Rank Agreement Factor (PRAF): using the PRAF, the study quantitatively
415 measures the general agreement pattern in the ranking of each criterion among all
416 the financier stakeholders that comprises senior lenders, equity financiers, and
417 infrastructure financiers. Hence, the null hypothesis suggests “any criterion on
418 which respondents have a strong agreement, will have a high PRAF score. But a
419 low PRAF score indicates disagreement among the respondent groups on the
420 criterion”.
421
422 5. Regression modelling: With regression analysis, relationship between a dependent variable
423 and independent variables (predictors) can be estimated. Hence, regressions analysis
424 facilitates understanding into how changes in predictors influence the dependent variable
425 (Field, 2005). The statistical hypothesis in this study’s regression analysis follows the
426 regression rule of thumb. That is, since R² (regression coefficient) usually ranges between 0
427 and 1, and a higher R² value indicates how well the model fits/predicts the observed data.
428 Any model with the highest R² value is selected as the right regression model for the study.

429
430 After thorough arrangement of data into SPSS, the study started by conducting reliability analysis on the
431 data set. According to Faravelli (1989), when analysing a survey data conducted with Likert-scaled
432 questionnaires, a reliability analysis is essential to ascertain the internal consistency of variables being
433 analysed. The formula for reliability analysis can be mathematically represented thus,
434
((((((
# $ %&'
435 != - $ … . (1)
∑+./ *+ + ∑- +./ %&'+
436 Reliability analysis helps discover whether the scales used in measuring the various bankability criteria
437 can consistently and truly reflect the construct it was intended to measure (Huang et al., 2006). As argued
438 by Field (2005), in a reliable data, the rule of thumb in Cronbach’s Alpha (α) coefficient is often between
439 0 and 1. However, George and Mallery (2003) argued that a coefficient value of 7 is much acceptable,
440 while a value of between 7 and 8 indicate strong internal consistency of the data set. Based on results from
441 the analysis, the overall Cronbach’s Alpha (α) coefficient for this study is 0.851 (see. below Table 5 for
442 results of the statistical test). This suggests a very strong internal consistency and overall reliability of the
443 bankability criteria identified in the study. Going further, to uncover whether all the bankability criteria in
444 the study are truly contributing to the internal consistency of the construct, “Cronbach's alpha if item
445 deleted” shown in column three of Table five was examined. According to Field (2005), any criterion no

18
 446 Table 5: Criteria for Evaluating the Bankability of Completion Risk and Associated Statistical Results
447
CR. Criteria Influencing the Bankability of Completion Risk in funding Non-Parametric Financier Stakeholders’ Descriptive Statistics
Applications for PFI/PPP Mega Projects Reliability ᵃ Test
Kruskal-Wallis
Cronbach’s Chi
1-Way Asymp.
ANOVA Senior Senior Equity Equity Infrastructure Infrastructure
α Square Sig. ᵇ Lenders Lenders Financiers’ Financiers’ Financiers’ Financiers’
If Item Mean Ranking Mean Ranking Mean Ranking
Deleted
CR1 Existence of Tried-and Test Technology for the construction of project. 0.737 1.693 0.429 4.45 3 4.28 3 4.2 9
CR2 Construction Contractor’s liability insurance cover 0.718 0.387 0.824 4.16 7 4.14 4 4.37 7
CR3 Construction contractor’s years of experience of successful completion of mega projects. 0.827 1.686 0.43 4.65 1 4.86 1 4.47 4

CR4 Construction Contractor’s financial strength 0.721 1.61 0.447 4.63 2 3.99 7 4.81 1
CR5 Existing cost liability or debt commitments of the project to other creditors different from the 0.772 2.962 0.027*** 3.06 22 2.53 22 2.78 20
lender
CR6 Pre-Completion Guarantee or Full Financial Guarantee from the sponsor at construction stage 0.632 0.565 0.754 3.91 12 3.45 16 3.56 17
CR7 Delay in start-up insurance to prevent cost and time-overrun 0.738 1.363 0.506 3.67 18 3.05 20 3.7 15
CR8 Existence of bank-financed construction cost overrun facilities 0.819 2.523 0.283 3.92 11 3.66 12 4.55 2

CR9 Contingent equity contribution from the project sponsors in case of cost over run 0.829 3.336 0.281 4.27 4 3.79 10 4.03 11
CR10 Debt Buy Out arrangement 0.711 1.724 0.422 3.81 13 3.58 15 1.85 23
CR11 Full injection of equity funds by project sponsors at the start of the construction phase 0.842 0.122 0.941 3.94 10 3.87 9 4.15 10
CR12 Construction contractor to accept “Single -Point Responsibility” on other project subcontractors 0.852* 0.03 0.99 3.55 20 3.66 12 3.59 16

CR13 Construction subcontract must represent very high value to the subcontractor 0.835 2.944 0.229 3.72 16 1.54 23 3.99 12
CR14 Construction contractor to accept Full Technology Wrap for the proper functioning of all project 0.815 2.541 0.001*** 3.69 17 3.76 11 3.5 18
assets after construction
CR15 Availability of Independent Technical Consultant (ITC) 0.843 2.392 0.189 4.22 5 4 6 4.51 3
CR16 Fixed Price Turn Key (FPTK) contract 0.849 1.978 0.372 4.2 6 4.37 2 4.22 8
CR17 Project contract to introduce benchmarking arrangements 0.839 1.017 0.601 2.53 23 3.42 17 2.84 19

CR18 Contractor must accept exceedingly high liability caps 0.857* 5.473 0.065 3.53 21 3.41 18 2.46 22

CR19 Construction contractor to accept exceedingly high performance and retention support 0.791 0.362 0.835 3.77 14 3.14 19 3.87 14

19
CR20 Contractor must handle the construction program and schedule in a conservative way 0.802 14.373 0.001*** 3.56 19 3.62 14 2.56 21

CR21 Contractual commitment to project’s output specifications and deliverables 0.636 6.08 0.048 4.02 9 3.9 8 4.46 5

CR22 Existence of clearly stated and objectively testable construction completion test requirements 0.801 2.967 0.227 3.75 15 3.03 21 3.88 13

CR23 Existence of liquidate damages for construction performance failures 0.783 1.96 0.375 4.09 8 4.07 5 4.42 6

448 Cronbach’s Alpha (α) Reliability Coefficient for the study is 0.851; CR = Criteria;

449 Significance at 95% Confidence Level=0.05%; Reject the null hypothesis where a criterion is below 0.05

450

20
451 contributing to reliability of the data will have a higher reliability coefficient compared to the overall
452 reliability of the data (0.851). This suggests that such criterion with higher value if deleted, would increase
453 the overall reliability of the entire data set (Santos, 1999). Using this rule as a yardstick, the null hypothesis
454 was confirmed on all the criteria except only two criteria, CR 12 and CR18, which were identified to have
455 values higher (0.852 and 0.857) than the overall reliability coefficient of the study. The two criteria are
456 CR12=Single -Point Responsibility from the main contractor to be responsible for other subcontractors
457 and CR18= Construction contractor to accept exceedingly high liability caps. These criteria were
458 identified not to be contributing to internal consistency of the data and so were considered unreliable and
459 subsequently deleted. On this regard, we were left with 21 reliable criteria influencing the bankability of
460 completion risk in PFI project deals.
461
462 Non-parametric Test (Kruskal-Wallis One-Way ANOVA)

463 After establishing the reliability of all the criteria included in the questionnaire survey through Cronbach’s
464 Alpha Reliability Analysis, the study proceeded to examine whether the three major financier stakeholders
465 (Senior Lenders, Equity Investors, Infrastructure Financiers) surveyed viewed all the criteria in the same
466 way or differently. Given that the data is considered not to be normally distributed, a non-parametric
467 statistical analysis known as "Kruskal-Wallis one-way analysis of variance" was employed. This tests the
468 null hypothesis that is, no statistically significant differences exist in the perception of the three
469 stakeholders on the 21 remaining criteria. Based on this hypothetical assumption, where a criterion has a
470 significance level less than 0.05, the null hypothesis is rejected. As shown in the fifth column of Table 5.
471 Three out of the 21 criteria, representing 14.28% of the entire criteria, were perceived differently by the
472 three stakeholders, with their significant level falling below the decision rule (0.05). These include CR14=
473 Contractor's acceptance of Full Technology Wrap for proper functioning of all project assets after
474 construction, CR20= Contractor must handle the construction program and schedule in a conservative
475 way and CR5= Existing cost liability or debt commitments of the project to creditors different from the
476 lenders. The implication of this result is that the stakeholders demonstrate general agreement in their
477 perception of 85.71% of the criteria (3 out of 21 reliable criteria). This therefore means that, though there
478 are differences in perception of the various criteria among the stakeholders, as explained by the pattern in
479 which they have ranked them, these differences seem to be unusually low across the entire criteria. As
480 such, the entire data from the surveyed respondents remain very useful in helping to understand patterns
481 of agreement among the stakeholders. To investigate this, the study adopted Principal Rank Agreement
482 Factor (PRAF) represented in Section 4.2.2 below. Additionally, the data was later used to develop a

21
483 regression model to identify the main drivers of bankability of completion risk in funding applications for
484 PFI/PPP megaprojects, based on the views of all the three stakeholders.
485
486 Financier Stakeholders’ Descriptive Analysis

487 To quantitatively designate the top-rated criteria among the three stakeholders, the study adopted mean
488 ranking approach using SPSS, as represented in columns 6 to 11 of Table 5. Based on the descriptive
489 statistics results, the top-five rated criteria from senior lenders’ perspectives are as follows: CR3=
490 Construction contractor with years of experience of successful completion of mega projects, CR4=
491 Construction Contractor with financial strength, CR1= Existence of Tried-and Test Technology for the
492 construction of project, CR11= Contingent equity contribution from the project sponsors in case of cost
493 over run, CR15 =Availability of Independent Technical Consultant (ITC).
494
495 The top five criteria from the perspectives of Equity financiers, as represented in Table 5 include, CR3=
496 Construction contractor’s years of experience of successful completion of mega projects, CR16=
497 Existence of Fixed Price Turn Key (FPTK) construction contract, CR1= Existence of Tried-and Test
498 Technology for the construction of project, CR2= Construction Contractor’s liability insurance cover, and
499 CR23= Existence of liquidate damages for construction performance failures. Going further, the top five
500 rated criteria for evaluating the bankability of completion risk from the perspective of the infrastructure
501 financiers include CR4= Construction Contractor with financial strength, CR8= Existence of bank-
502 financed construction cost overrun facilities, CR15= Availability of Independent Technical Consultant
503 (ITC), CR3= Construction contractor with years of experience of successful completion of mega projects,
504 and CR21= Contractual commitment to project’s output specifications and deliverables (See Table 5
505 above).
506
507 However, it is important to note that, out of all the criteria, CR3= Construction contractor with years of
508 experience of successful completion of mega projects; CR1= Existence of Tried-and Test Technology for
509 the construction of project and CR5=Existing cost liability or debt commitments of the project to other
510 creditors different from the lender were identified to be common and rated similarly by both the senior
511 lenders and the equity financiers. This result (CR3) suggest that engaging an experienced construction
512 contractor with good record of successful projects execution was critical to mitigating completion risk in
513 mega projects, and therefore a key criterion for financiers’ consideration. In the same view, the implication
514 of stakeholders’ agreement on CR1 confirms studies such as He et al. (2015) and Xu et al. (2015) who
515 argued that experimenting with state-of-the-art construction technology on large-scale projects is a
516 requisite for failure as such technology may be difficult to repair in the event of machinery breakdown. In

22
517 addition, stakeholders' agreement on criterion CR5 is perfectly in line with Delmon (2015) who
518 highlighted excessive financial burden as one of the many causes of insolvency in construction firms.
519 From the stakeholders' view, the possibility that such construction contractor will liquidate while project
520 is ongoing portends enormous risk to project completion and financiers' investment.
521
522
523 Principal Agreement Rank Factor (PRAF)
524
525 As part of the objective of this study, it was important to examine the degree to which the three financier
526 stakeholders agree on the significance of each criterion, based on their rankings of the 21 remaining
527 criteria. In order to achieve this objective, a Principal Agreement Rank Factor (PRAF) and Rank
528 Agreement Factor (RAF) were adopted. This is in line with previous studies such as Chan and
529 Kumaraswamy (2002), Usman et al. (2012), Ubani and Ononuju, (2013), Oyedele et al. (2015) who have
530 quantitatively examined pattern of agreement in ranking of factors among diverse stakeholders. RAF and
531 PRAF can be mathematically computed as:
∑ &'(
532 !"# = (2)
)
+,-./0 1+,-2
533 *!"# = × 100% (3)
+,-./0
534 The PRAF for all the completion risk bankability criteria were computed using Equation (2) and (3).
535
536 Based on the equation, !"# 89: is the maximum RAF of all the criteria !"# ; is the RAF for criteria
537 <, N is the number of criteria being ranked, which are 21 and ∑ =>? is the sum order of ranking for
538 Senior Lenders, Equity Financiers, and Infrastructure Financiers. By principle, a higher PRAF value
539 indicates more agreement among the stakeholders with respect to a criterion, as against when the PRAF
540 is low. Hence, a PRAF of 100 suggest strong agreement while zero indicates complete disagreement
541 among the financier stakeholders. On the other hand, the Rank Agreement Factor (RAF) could be > 1,
542 with a higher value indicating more disagreement in ranking. In this regard, a RAF of zero suggests
543 excellent agreement, more than a RAF of 1 or 2. Results from this statistical analysis can be seen in
544 Table 6 below, which presents the pattern of agreement in ranking of the 21 criteria among the three
545 financier stakeholders (Senior Lenders, Equity Financiers and Infrastructure Financiers) that were
546 surveyed.
547

23
548 In line with the null hypothesis on PRAF, result of the analysis as shown in Table 6 above revealed, seven
549 key criteria influencing the bankability of construction and completion risk in PFI/PPP mega projects, all
550 with high PRAF score. These criteria were identified as:
551 § CR3 = Construction contractor’s years of experience of successful completion of mega projects.
552 § CR4 = Construction Contractor’s financial strength
553 § CR15 = Availability of Independent Technical Consultant (ITC)
554 § CR1= Existence of Tried-and Test Technology for the construction of project.

24
555 Table 6: Principal Agreement Rank Factor (PRAF) among Senior Lenders, Equity Financiers and Infrastructure Financiers
No Criteria Influencing the Bankability of Completion Risk in funding Applications for Senior Equity Infrastructure Sum of RAF PRAF Ranking
PFI/PPP Mega Projects Lenders Financiers Financiers Ranking Order
CR3 Construction contractor’s years of experience of successful completion of mega 1 1 4 6 0.29 89.29 1
CR4 projects.
Construction Contractor’s financial strength 2 7 1 10 0.48 82.14 2
CR15 Availability of Independent Technical Consultant (ITC) 5 6 3 14 0.67 75.00 3
CR1 Existence of Tried-and Test Technology for the construction of project. 3 3 9 15 0.71 73.21 4
CR16 Existence of Fixed Price Turn Key (FPTK) construction contract 6 2 8 16 0.76 71.43 5
CR2 Construction Contractor’s liability insurance cover 7 4 7 18 0.86 67.86 6
CR23 Existence of liquidate damages for construction performance failures 8 5 6 19 0.90 66.07 7
CR21 Contractual commitment to project’s output specifications and deliverables 9 8 5 22 1.05 60.71 8
CR8 Existence of bank-financed construction cost overrun facilities 11 12 2 25 1.19 55.36 9
CR11 Full injection of equity funds by project sponsors at the start of the construction 10 9 10 29 1.38 48.21 10
CR9 phase
Contingent equity contribution from the project sponsors in case of cost overrun 4 10 17 31 1.48 44.64 11
CR10 Debt Buy Out arrangement 13 15 18 46 2.19 17.86 12
CR19 Construction contractor to accept exceedingly high performance and retention 14 19 14 47 2.24 16.07 13
support
CR6 Pre-Completion Guarantee or Full Financial Guarantee from the sponsor at 12 16 20 48 2.29 14.29 14
CR13 construction
Constructionstage
subcontract must represent very high value to the subcontractor 20 13 16 49 2.33 12.50 15
CR5 Existing cost liability or debt commitments of the project to other creditors different 17 22 11 50 2.38 10.71 16
CR17 from thecontract
Project lender to introduce benchmarking arrangements 23 9 19 51 2.43 8.93 17

CR22 Existence of clearly stated and objectively testable construction completion test 15 21 16 52 2.48 7.14 18
CR7 requirements
Delay in start-up insurance to prevent cost and time-overrun 18 20 15 53 2.52 5.36 19
CR14 Construction contractor to accept Full Technology Wrap for the proper functioning 22 11 21 54 2.57 3.57 20
CR20 of all projectmust
Contractor assets afterthe
handle construction
construction program and schedule in a conservative 19 14 23 56 2.67 0.00 21
556 way
557

25
558
559 § CR16 = Fixed Price Turn Key (FPTK) contract
560 § CR2 = Construction Contractor’s liability insurance cover
561 § CR23 = Existence of liquidate damages for construction performance failures
562
563 Multiple Linear Regression Model
564 After identifying the reliable and top-rated criteria based on the perceptions of respondents across the three
565 stakeholder groups surveyed, the study proceeded to unravel the key drivers of bankability for completion
566 risk in funding applications for Private Finance Initiatives and Public Private Partnerships (PFI/PPP) mega
567 projects. To realise this objective, the study constructed a linear regression model. This approach became
568 necessary based on the proposition that one or more criteria (independent or explanatory variables) will
569 hugely correlate with the response variable (dependent variable), which is "bankable completion risk". The
570 response variable was therefore measured in the questionnaire by asking respondents to indicate the extent
571 to which they believe each criterion contributes towards achieving a bankable completion risk in funding
572 applications for PPP megaprojects. The mathematical formula for a regression model is:
573 ! = #$ + #& '& + #( '& + #) ') + ⋯ + #+ '+ + , ………………… (4)
574
575 However, with the 21 bankability criteria for evaluating completion risk representing independent variables,
576 the regression model for the study is thus expressed as:
577 ./0 = #$ + #& /0& + #( /0( + #) /0) + ⋯ . . +#+ /0+ + , … … … (5)
578
579 Where ./0+ = value of response dependent variable (Bankability of Completion risk), #$ = is the intercept
580 term and is constant, #& is the coefficient of the first criterion (CR1), #( is the coefficient of the second
581 criterion (CR2), #) is the coefficient third criterion (CR3), #+ is the coefficient of the 5 criterion /0, while
582 , is the mean-zero random error term (the difference between the predicted and actual value of .//0 for
583 the 5th respondents. Through the aid of SPSS, a step-wise model was performed on the data. Table 7 show
584 the summary of the model that contains five possible models and their associated predictors. The third
585 column shows R², which is often referred to as coefficient of determination and suggests the correlation
586 between the observed values of .//0 and the predicted values of .//0 in the regression. As a rule, R²
587 usually ranges between 0 and 1, and a higher value reflects how well the model predicts the observed data.
588 Considering that Model 5 shows the highest R² value (in line with the regression hypothesis), it is therefore
589 selected as the most suitable regression model for this study. With a R² value of 0.632, this indicated that the
590 model is capable of predicting 63.2% of the variability in the dependent variable. As such, the model is
591 appropriate for predicting the bankability of completion risk in funding application for PPP mega projects.

26
 592 Table 7: Regression Model Summary

Model R R² Adjusted Std. Error Change Statistics Durbin-Watson ANOVA

R² of the
R² Change F Change Sig. F F Sig.
Estimate
Change

1 .575a .331 .320 .513 .331 29.202 .000 1.830 29.202 .007b

2 .706b .498 .481 .449 .167 19.300 .000 28.780 .005c

3 .733c .537 .512 .435 .039 4.768 .033 22.022 .004d

4 .756d .571 .541 .422 .035 4.585 .037 18.701 .003e

5 .795e .632 .568 .409 .032 4.421 .040 16.759 .001f

Dependent Variable: Achieving bankable completion risk in funding proposal for PPP Mega Projects
a. Predictors: (Constant), CR1.
b. Predictors: (Constant), CR3, CR1, CR22
c. Predictors: (Constant), CR16, CR14, CR10.
d. Predictors: (Constant), CR4, CR23, CR3, CR2
e. Predictors: (Constant), CR3, CR4, CR1, CR15, CR16.
593
594 Q Table 8: Regression Model Results
Model Unstandardized Coefficients Standardized t Sig. Collinearity Statistics
Coefficients

B Std. Error β Tolerance VIF

Constant (Dependent variable) 3.09 0.52 4.17 0.013

CR3. Construction contractor with years of experience of successful completion of mega 0.43 0.08 0.57 5.404 .000 .839 2.191
projects
CR4. Construction Contractor with financial strength 0.36 0.09 0.41 2.620 .001 .952 2.124

CR1. Existence of Tried-and Test Technology for the construction of Project 0.28 0.11 0.34 2.070 .003 .877 1.177

CR15. Availability of Independent Technical Consultant (ITC) 0.25 0.07 0.27 2.141 .004 .845 1.050

CR16. Existence of Fixed Price Turn Key (FPTK) construction contract 0.21 0.04 0.23 3.897 .023 .734 1.000

595 Dependent Variable: Achieving bankable completion risk in funding proposal for PPP Mega Project

27
596 Going further, other criteria that confirm the model accuracy include the adjusted R², the Durbin-Watson
597 test, standard error of estimate and the significance level of the ! statistics. According to Field (2005), the
598 Adjusted R² is a measure of how well the model is capable of generalising beyond the available data, which
599 in ideal situations, should be equal or close to the R² values. This difference, which indicates a loss in
600 predictive power of the model, is small in this model showing a value of 0.064 (0.632 – 0.568). This
601 suggests a 6.4% less variance in the outcome and as such, indicates the model has a good cross-validity. The
602 standard error of estimate is the measurement of the accuracy of predictions that is made with a model or a
603 measurement of errors in predictions. In a good model, the relationship between the explanatory variables
604 and the outcome is expected to be perfect, thereby indicating less error by being closer to zero. Based on
605 analysis in this study, the model with the standard error value that is closest to zero is model 5 with a value
606 of 0.409. This confirms the predictive power of the model. In addition, as suggested by Engle and Yoo
607 (1987), any two predicted observations should show uncorrelated and independent errors. In this study,
608 Durbin-Watson statistics test was therefore used to examine these correlations. According to Hill and Flack
609 (1987), the recommended value for these correlations vary between 0 and 4, with a value of 2 indicating
610 uncorrelated residuals and are thus a good model. In this study, the Durbin-Watson test value, as shown in
611 Table 7 is 1.830, which can be approximated to two. This therefore indicates the absence of autocorrelation.
612 Lastly, ANOVA in this study also helps confirm whether the model perfectly fits the data examined and
613 should have a recommended value of less than 0.05 at 95% confidence interval. Table 7 confirms the fitness
614 of the model 5 with a value of 0.01.
615
616 After confirming the model fitness and predictive accuracy, the study proceeded to identify the key criteria
617 predicting bankability of completion risk in funding application for PPP megaprojects. In this regard, model
618 5 indicates that there are five best criteria that a necessary for ensuring bankability of completion risk from
619 financiers' perspective, out of the 21 criteria analysed. It is important to note that these 21 were the reliable
620 criteria identified after conducting reliability analysis on the 23 criteria that were put in the questionnaire to
621 project financiers. These five criteria are therefore referred to as the critical success factors for ensuring the
622 bankability of completion risk in funding application for PFI/PPP megaprojects. They comprise:
623 § CR3=Construction contractor with years of experience of successful completion of mega projects
624 § CR4=Construction Contractor with financial strength
625 § CR1=Existence of Tried-and Test Technology for the construction of project
626 § CR15=Availability of Independent Technical Consultant (ITC)
627 § CR16=Existence of Fixed Price Turn Key (FPTK) construction contract
628 Going further, the study proceeded to check for the significance of these five criteria using the t-test
629 significance value for each criterion, as well as the collinearity statistics, as demonstrated in Table 8 above.

28
630 By rule, any criteria showing a significance level of 0.05, is considered to be making significant contribution
631 to the model (Field, 2005). As such, the closer a value is to 0, the higher the significance of such criteria.
632 Based on evidences from our model, all the five criteria have values, which are less than 0.05. As shown in
633 Table 8, CR3=Construction contractor with years of experience of successful completion of mega projects
634 shows the highest significance value at 0.00, while CR14. Existence of Fixed Price Turn Key (FPTK)
635 construction contract shows the least significance at .023 respectively. The collinearity statistics estimates
636 the existence of any significant relationship among the criteria, which may weaken the model. This can be
637 confirmed via the variance inflation factor (VIF), which should not be more that 5 and the tolerance statistic
638 which works with VIF and should not be less than 0.2. Based on this model, all the VIF statistics are between
639 1.0 and 2.1, which is less than 5, while all the tolerance statistics are above 0.2, as shown in Table 8. The
640 results therefore confirm the absence of multicollinearity among the predictors/criteria.
641
642 With values from unstandardized coefficient as shown in Table 8 above, the optimum regression model,
643 which demonstrates mathematically, the statistical correlation between bankability of completion risk and
644 associated key success factors is therefore re-written as:
645
646 # = %. '( + '. *% (-.%) + '. %0 (-.*) + '. 12 (-.3) + '. 14 (-.34) + '. 13 (-.30) + 56 (0)
647
648 Model validation
649 As a part of the research, it was important to confirm the validity of this model on a real life PFI/PPP project
650 case study. As such, using snowball sampling method, a team of financier experts in a reputable financial
651 institution in the UK was approached. The team comprised three senior financial risk analysts, six credit risk
652 analysts, two infrastructure lending officers, three senior managers, and one head of structured finance. This
653 makes 15 financier experts with all having an average of 13 years’ experience in international project
654 financing. This team was approached to examine the relevance of the developed model to a specific PPP
655 mega project they have been involved. Using one-page questionnaire survey, the experts were asked to rank
656 the five critical success factors based on the extent to which they contributed to their due diligence appraisal
657 on completion risk in the chosen PPP mega project. The team chose a University Student Housing PPP
658 project valued at US$1.4 billion. This project, located in one of Europe’s capitals, was to provide 842
659 additional bed spaces for students and will operate under a 40-year concession plan. The project, whose
660 construction phase lasted a period of 36 months and was completed in 2011, is currently in operation.
661
662

29
663 14 out of the 15 distributed questionnaires were returned making 93.33% response rate. The respondents'
664 ratings of the five critical success factors in the questionnaire were extracted and inputted in the regression
665 model (see Eq. 6). The overall success in achieving bankable completion risk in funding applications for
666 PPP mega projects was then mathematically calculated. Using Spearman rank correlation non-parametric
667 statistics, the association between two datasets measured on ordinal scale was compared. Here, the model-
668 computed score was compared to the ratings given by the 14 respondents. The strength of association in
669 correlated items is usually indicated in values between -1 to +1 (MacFarland and Yates, 2016). With the aid
670 of SPSS, the correlation coefficient for the data showed 0.735, with a significance level of 0.0315 at 99%
671 confidence interval. This result suggests a positive relationship between the ratings of the financier experts
672 and the model-computed scores. Based on this evidence, the model is therefore considered a strong predictor
673 and the five criteria were important for ensuring a bankable completion risk in funding applications for
674 PFI/PPP mega projects.

675 Discussion of Findings

676 Based on evidences as reflected in Table 8 above, the Construction Contractor’ years of Experience of
677 Successful Completion of Mega Projects was considered the most important bankability criteria for lenders
678 in evaluating completion risk in PPP loan applications. As argued by Flyvbjerg (2014), during construction
679 stage of projects, two important risk factors to stakeholders, including lenders are cost and time overrun.
680 Many existing studies have identified various reasons why construction projects often overshoot budget and
681 timeline (Song 2017; Perera et al., 2016; Budayan, 2018). Some of the factors include but not limited to
682 inaccuracy of materials estimates, unpredictable weather, inadequate planning, inaccurate prediction of
683 equipment production rates, skill shortages, complexity of project, inflationary material cost etc. (Larsen et
684 al., 2015; Amoatey, 2015; Budayan, 2018; Owolabi et al., 2018). However, according to Kaming et al.
685 (1997), contractor's lack of project type experience is one of the most crucial factors that may hinder
686 successful delivery of projects within expected budget and timeline. This is so because, previous projects'
687 experience tends to result in contractor's better understanding and capability to deal with the inherent
688 dynamics and risk factors which may pose a danger to successful project delivery (Hakeem et al., 2018). As
689 a result, given that projects are usually front-loaded with regards to funds at construction stages, combined
690 with associated huge loan drawdowns; the risk to lenders investments at such stage can be enormous. As
691 such, project banks will require a proven and tested construction contractor with similar project experience
692 and capacity to deliver the project, if bankability is to be achieved.
693
694

30
695 Following construction contractors’ project type experience, project banks consider the Construction
696 Contract’s Financial Strength as the second important criterion for completion risk bankability (see Table
697 8). This result confirmed evidences from studies such as Hoffman (2008) and Mills (2010) who argued that
698 timely project completion at stipulated price requires construction contractor with strong financial resources
699 needed to support contractual obligations relating to workmanship guarantees, liquidated damage payments,
700 indemnities, etc. As highlighted by Bing et al. (2005) considering the complex and high-risk nature of
701 Private Finance Initiatives and Public Private Partnerships (PFI/PPP) projects, the risk that insufficient fund
702 may result in various counter-party challenges with the construction contractor is a threat to limited-recourse
703 financing. According to Akintoye et al. (2003), the domination of PFI/PPP market by big construction firms
704 is not unconnected to their huge financial and technical capabilities. With huge finance war-chest, big
705 construction firms could cope well with the high cost of bidding and tendering exercise in PFI/PPP
706 procurements (Robinson and Scott, 2009). This is quite important for project banks considering that only
707 financially robust contractors can stay the course of the prolonged PFI tendering cost, timeline as well as
708 have deep pockets to meet contractual obligations on the project.
709
710
711 Further evidences from the study also suggest that the third important criterion for evaluating the bankability
712 of completion risk in PFI loan applications is the use of Tried, Tested and Reliable Construction Project
713 Technology (See Table 8). According to Mills (2010), most project banks are often wary of investing in
714 projects that propose a revolutionary project technology for the construction stage. This is because, in most
715 cases, there is always a likelihood of inability to maintain or repair such technologies in case they break
716 down. In other instances, such state-of-art technology might require engaging experts to drive its operations,
717 which may further increase the cost of constructing the project (Hakeem et al., 2018). As argued by Meng
718 and McKevitt (2011), lenders are more interested in projects with tested and reliable construction technology
719 that has good record of long operating hours and low-down times, as against latest technology whose
720 operational capability is less known. Using tested construction technology thus gives more confidence to
721 financiers concerning ability to forecast potential cost and time overrun on projects. From the perspective of
722 Lim and Mohamed (1999), the fear that a project may not pass completion test is topical issue in construction
723 risk due diligence appraisal. Mills (2010) argued that the construction delivery stage has significant impacts
724 with respect to strategic issues on a project especially concerning profit margins and returns on investment
725 for investors. As such, bankability can only be achieved where tested and tried project technology is made
726 to drive the construction stage of PFI/PPP projects.
727
728

31
729 Going further, results shown in Table 8 reveal that the fourth important criterion for assessing the bankability
730 of completion risk in PFI/PPP loan applications is the Availability of a Competent and Independent
731 Technical Consultant. This evidence confirms findings from existing studies like Robinson and Scott (2009)
732 and Hakeem et al. (2018) who argued that providing technical due diligence on potential PPP project is
733 crucial towards the preparation of projects’ business cases. According to Hoffman (2008), given the huge
734 risk associated with construction stages of projects, more rigour is usually applied towards technical due
735 diligence especially from lenders point of view. In most scenarios in PFI/PPP procurements, the project
736 consortium often comprised a construction firm who handles the project’s technical development. This
737 construction contractor plays crucial role in providing technical details and analysis needed in projects'
738 business cases. However, in some circumstances, project banks often require an independent technical
739 consultant hired by the sponsors’ team. The objective here is to have an independent consultant, who is
740 dispassionate about the project, to provide technical insights and recommendations on the technical
741 development plans of the project. Financiers will require the technical consultant to simulate various
742 scenarios, which may threaten the technical feasibility of the project (Mills, 2010). This approach often gives
743 many assurances to project banks concerning assessing the possibility of project completion.
744
745 Finally, the fifth important bankability criterion for assessing completion risk in project loan applications is
746 Existence of Fixed Priced Turnkey Contract (See Table 8 for results). Fixed Price Turnkey in PFI/PPP
747 project finance describes a procurement approach in which the construction constructor assumes the
748 responsibility of constructing a project in line with contractually stated output specifications, at a fixed cost
749 and within a determined timeline (Yescombe, 2013). Under a fixed price turnkey method, the construction
750 contractor cannot change the agreed price of the project. As such, the risks of cost and time overrun are
751 passed down to the contractor, who has the mandate to deliver the keys to the constructed facilities, to the
752 clients at the end of a stipulated construction period. As argued by Mills (2010), although, turnkey contracts
753 are very common in PFI/PPP procurements, not all projects are delivered using turnkey approach. A huge
754 number of PFI projects are still be constructed under a “Cost Plus Approach” in which the contractor charges
755 a construction cost with the addition of a profit margin or mark-up (Hoffman, 2008). One of the major put
756 off for most project banks in the cost-plus approach is that responsibility for managing cost and time overrun
757 are borne by the project sponsors as against the construction contractor. From financiers' perspective, this
758 method creates a moral hazard situation in that; the contractor has no incentive to ensure optimum
759 performance, which should forestall time and cost overruns and could as well as act indecently. As such,
760 most project banks favours fixed price turnkey method which allows the construction contractor take
761 responsibility for construction risks (cost, time overruns and technology risks), and thus ensure greater
762 commitment from the contract towards successful completion of the project.

32
763 Implication for Practice
764
765 This study has huge strategic implication for most construction firms especially at the management level.
766 The enormous amount of time and cost overrun associated with mega-projects is such that, many
767 construction firms have gone burst under its weight, particularly in the absence of adequate parent company
768 support or risk guarantee. As a result, this study suggests contractors intensify their pre-contract efforts by
769 putting together bankable completion risk in funding proposals, as against trying to simply accept the
770 transfer of completion risk to them, which may prove more challenging to deal with considering the
771 complexities in PPP arrangements. In addition, going by a thorough analysis of findings from this study, the
772 various criteria influencing lender’s decision on the bankability of completion risk may be put into two broad
773 categories namely: contractor competency and a robust construction contract. These two factors are crucial
774 towards successful delivery of Private Finance Initiatives and Public Private Partnerships (PFI/PPP)
775 megaprojects in the UK construction industry. The UK construction sector is said to comprise big
776 construction firms and micro-businesses, often referred to as Small and Medium Scale (SME) construction
777 firms. While the big construction firms have dominated the construction sector by accounting for 55% of
778 UK’s built environments, the SME construction firms, which represents 96% of the industry have continued
779 to play the second fiddle roles. This scenario has also translated in many PFI/PPP projects being executed
780 by big construction contractors who play significant roles in setting up many Special Purpose Vehicles
781 (SPVs), given their huge experience, expertise, and financial wherewithal. SME construction firms on the
782 hand have been acting as sub-contractors on various projects and in many cases, restricted to small value
783 projects. However, considering the government’s sustained ambition to drive the procurement of critical
784 infrastructures in the UK through private sector routes such as PPP, a good understanding of how SME
785 construction firms can deepen their competencies will further position them for penetration into the project
786 finance market. This can be achieved by collaborating with project sponsors who have experience in
787 PFI/PPP megaprojects, to create a win-win relationship that will benefit each party. This mutual relationship
788 will rub off on the construction contractor, as he benefits by being involved in strong mega projects that are
789 implemented under robust construction contracts. The fixed price turnkey method, which is the popular
790 procurement approach in PPP mega projects, is usually comprehensive in nature in terms of output
791 specifications, availability requirements and various contractual details. As such, strong experience in the
792 execution of such type of construction contracts will improve the profile of the construction contractors in
793 terms of bankability. The implication of this study for construction contractors is also in terms of contract
794 negotiations in PFI/PPP megaprojects. Evidences from the study show that, there is a trade-off relationship
795 among some of the criteria influencing senior lenders’ bankability decision on completion risk. Where a

33
796 contractor has “project type experience” with strong financial capacity and tested construction technology,
797 the existence of pre-completion guarantee can be negotiated as unnecessary, given the strong contractor
798 profile. In the overall, only a competent construction contractor working under robust construction contract
799 will be competent to serve the interest of project financiers and other stakeholders in the delivery PFI/PPP
800 mega projects.

801 Conclusion
802
803 This study adopted mixed methodological approach towards investigating the bankability of completion
804 risk in Private Finance Initiatives and Public Private Partnerships (PFI/PPP) mega project appraisal. Based
805 on evidences from the study constructed, five key criteria representing critical success factors (CFSs) were
806 identified to have significant influence on achieving bankable completion risk. These are (1) Construction
807 contractor’s years of experience of successful completion of mega projects, (2) Construction Contractor’s
808 financial strength, (3) Existence of Tried-and Test Technology for the construction of project, (4)
809 Availability of Independent Technical Consultant (ITC) and (5) Existence of Fixed Price Turn Key (FPTK)
810 construction contract. From the opinion of project financiers, these five criteria would be crucial for project
811 contractors and sponsors, if PFI/PPP mega projects’ funding applications will be successful.
812
813 It is important to note that, most project banks have little knowledge of top-level technical details of complex
814 projects, which is typical with PPPs. As such, financiers’ risk aversion is often very high, especially when
815 bankability of completion risk element in funding proposals cannot be sufficiently justified. This has led
816 many PPP funding applications being turned completely down by financiers. In PFI/PPP mega projects,
817 which is also the case in other types of project procurements, competency of the construction contractor and
818 robust construction contracts are crucial to the roles played by construction contractors. Construction
819 contractors’ negotiations must also take cognizance of bankability requirements, which may need to be
820 traded-off with other risk mitigation strategies in the contracts. These requirements must be adequately
821 negotiated to relieve the construction contractor of cumbersome contractual obligations, which may become
822 a source of challenge in the near future.
823
824 This study contributes to knowledge with the identification of key bankability criteria that can help
825 construction contractors and PFI project sponsors to fulfil the bankability requirements for completion risk
826 in PFI/PPP megaprojects. Considering that most large-scale mega projects are usually non-investment grade
827 due to their high-risk profiles, which creates financing challenges, the findings of this study provides
828 valuable resource to stakeholders towards winning banks’ funding approval. Although this study

34
829 specifically centres on bankability criteria for evaluating completion risk in PFI/PPP megaprojects,
830 additional empirical studies are needed to examine what constitute bankability and the various criteria for
831 other project risks in PFI/PPP such as operations, legal, concession, political, currency, counter party risks,
832 etc. It will also be very pertinent to examine the perspectives of contractors and project sponsors on factors
833 militating against the bankability of PFI/PPP projects within the UK construction industry. Evidences from
834 this study were limited to the UK PFI/PPP and construction industry. As such, the findings should be
835 interpreted within this context. Studies focusing on country-specific factors that influence bankability of PPP
836 projects in other geographical locations will also be crucial for future research. This will help to contextualise
837 bankability of projects based on the public procurement climate in such nations.

35
838 References
839 Aladağ, H., & Işik, Z. (2017). Role of Financial Risks in BOT Mega transportation Projects in
840 Developing Countries. Journal of Management in Engineering, 33(4), 04017007.
841
842 Akintoye, A., Hardcastle, C., Beck, M., Chinyio, E. and Asenova, D., (2003). Achieving best value in
843 private finance initiative project procurement. Construction Management and Economics.
844 21(5), pp.461-470.
845
846 Amoatey, C. T., Ameyaw, Y. A., Adaku, E., & Famiyeh, S. (2015). Analysing delay causes and
847 effects in Ghanaian state housing construction projects. International Journal of
848 Managing Projects in Business, 8(1), 198-214.
849
850 Amidu, A. R. (2017). A review of funding and its implications for construction clients. In
851 Clients and Users in Construction (pp. 113-128). Routledge.
852
853 Babatunde, S. O., & Perera, S. (2017). Barriers to bond financing for public-private partnership
854 infrastructure projects in emerging markets: A case of Nigeria. Journal of Financial
855 Management of Property and Construction, 22(1), 2-19.
856
857 Boateng, P., Chen, Z., & Ogunlana, S. O. (2015). An Analytical Network Process model for
858 risks prioritisation in megaprojects. International Journal of Project Management, 33(8),
859 1795-1811.
860
861 Budayan, C. (2018). Evaluation of Delay Causes for BOT Projects Based on Perceptions of
862 Different Stakeholders in Turkey. Journal of Management in Engineering, 35(1),
863 04018057.
864
865 Bing, L., Akintoye, A., Edwards, P. J., and Hardcastle, C. (2005). The allocation of risk in PPP/PFI
866 construction projects in the UK. International Journal of Project Management, 23(1), 25-35.
867
868 Chan, D. W., & Kumaraswamy, M. M. (2002). Compressing construction durations: lessons learned from
869 Hong Kong building projects. International Journal of Project Management, 20(1), 23-35.
870
871 Chowdhury, A. N., Chen, P. H., & Tiong, R. L. (2015). Credit enhancement factors for the
872 financing of independent power producer (IPP) projects in Asia. International journal of
873 project management, 33(7), 1576-1587.
874
875 Chou, J. S., & Pramudawardhani, D. (2015). Cross-country comparisons of key drivers, critical
876 success factors and risk allocation for public-private partnership projects. International
877 Journal of Project Management, 33(5), 1136-1150.
878
879 Creswell, J. W., & Clark, V. L. P. (2017). Designing and conducting mixed methods research.
880 Sage publications.
881
882 Delmon, J. (2015). Private sector investment in infrastructure. Wolters Kluwer Law &
883 Business.
884
885 Demirag, I. (2017) A framework for examining accountability and value for money in the UK’s
886 private finance initiative. In Corporate Social Responsibility, Accountability and
887 Governance (pp. 77-92). Routledge.
888

36
889 Engle, R. F., & Yoo, B. S. (1987). Forecasting and testing in co-integrated systems. Journal of
890 econometrics, 35(1), 143-159.
891
892 European Court of Auditors (2018, Special Report). Public Private Partnerships in the EU:
893 Widespread shortcomings and limited benefits. Accessed Online
894 [https://www.eca.europa.eu/Lists/ECADocuments/SR18_09/SR_PPP_EN.pdf].
895
896 Faravelli, L., (1989). Response-surface approach for reliability analysis. Journal of Engineering
897 Mechanics. 115(12), pp.2763-2781.
898
899 Field, A. (2005). Discovering statistics using SPSS, 2nd Ed., Sage Publications, London.
900
901 Fowler, J., Cohen, L., & Jarvis, P. (2013). Practical statistics for field biology. John Wiley & Sons.
902
903 Fathali, E., & Ibrahim, H. (2015). Private Partner Selection and Bankability Assessment of PPP
904 in Infrastructure Projects (Doctoral dissertation, Concordia University).
905
906 Giezen, M., Salet, W., & Bertolini, L. (2015). Adding value to the decision-making process of
907 mega projects: Fostering strategic ambiguity, redundancy, and resilience. Transport
908 Policy, 44, 169-178.
909
910 Gatzert, N., & Kosub, T. (2016). Risks and risk management of renewable energy projects: The
911 case of onshore and offshore wind parks. Renewable and Sustainable Energy Reviews,
912 60, 982-998.
913
914 Hill, R. J., & Flack, H. D. (1987). The use of the Durbin–Watson d statistic in Rietveld analysis. Journal
915 of Applied Crystallography, 20(5), 356-361.
916
917 Huang, H.Z., Zuo, M.J. and Sun, Z.Q., (2006). Bayesian reliability analysis for fuzzy lifetime data. Fuzzy
918 Sets and Systems. 157(12), pp.1674-1686.
919
920 Hakeem Adedayo Owolabi, Lukumon Oyedele, Hafiz Alaka, Obas John Ebohon, Saheed Ajayi,
921 Olugbenga Akinade, Muhammad Bilal, Oladimeji Olawale, (2018) "Public private
922 partnerships (PPP) in the developing world: mitigating financiers’ risks", World Journal
923 of Science, Technology and Sustainable Development.
924
925 He, Q., Luo, L., Hu, Y., & Chan, A. P. (2015). Measuring the complexity of mega construction
926 projects in China—A fuzzy analytic network process analysis. International Journal of
927 Project Management, 33(3), 549-563.
928
929 Hodge, G., & Greve, C. (2017). Private Finance. The Oxford Handbook of Megaproject
930 Management, 362.
931
932 Kennedy, L. (2015). The politics and changing paradigm of megaproject development in
933 metropolitan cities. Habitat International, 45, 163-168.
934
935 Kaming, P.F., Olomolaiye, P.O., Holt, G.D. and Harris, F.C., (1997). Factors influencing construction
936 time and cost overruns on high-rise projects in Indonesia. Construction Management and
937 Economics. 15(1), pp.83-94.

37
938 KPMG Global Construction Industry Survey 2015: Climbing the Curve. Accessed on 25th
939 September, 2018[https://assets.kpmg.com/content/dam/kpmg/pdf/2015/04/global-
940 construction-survey-2015.pdf].
941
942 Larsen, J. K., Shen, G. Q., Lindhard, S. M., & Brunoe, T. D. (2015). Factors affecting schedule
943 delay, cost overrun, and quality level in public construction projects. Journal of
944 Management in Engineering, 32(1), 04015032.
945
946 Lepage M., (May 24th, 2017). Types of Schedule Delays in Construction Projects. Plan
947 academy (Online). Accessed on 25th September, 2018
948 [https://www.planacademy.com/types-of-schedule-delays-in-construction/].
949
950 Lederman, L.C., (1990). Assessing educational effectiveness: The focus group interview as a technique
951 for data collection. Communication Education. 39(2), pp.117-127.
952 Le-Hoai, L., Dai Lee, Y., and Lee, J. Y. (2008). Delay and Cost Overruns in Vietnam Large Construction
953 Projects: A comparison with other selected countries. KSCE Journal of Civil Engineering, 12(6),
954 367-377.
955
956 Liu, J., Zhao, X., & Yan, P. (2016). Risk paths in international construction projects: Case study
957 from Chinese contractors. Journal of Construction Engineering and Management, 142(6),
958 05016002.
959
960 Lim, C.S. and Mohamed, M.Z., (1999). Criteria of project success: an exploratory re-examination.
961 International Journal of Project Management. 17(4), pp.243-248.
962
963 Liu, T., Wang, Y., & Wilkinson, S. (2016). Identifying critical factors affecting the
964 effectiveness and efficiency of tendering processes in Public–Private Partnerships (PPPs):
965 A comparative analysis of Australia and China. International Journal of Project
966 Management, 34(4), 701-716.
967
968 MacFarland, T. W., & Yates, J. M. (2016). Spearman’s Rank-Difference Coefficient of Correlation. In
969 Introduction to Nonparametric Statistics for the Biological Sciences Using R (pp. 249-297).
970 Springer International Publishing.
971
972 Mazher, K. M., Chan, A. P., Zahoor, H., Khan, M. I., & Ameyaw, E. E. (2018). Fuzzy Integral–
973 Based Risk-Assessment Approach for Public–Private Partnership Infrastructure Projects.
974 Journal of Construction Engineering and Management, 144(12), 04018111.
975
976 McKinsey & Company (2016). Imagining Construction’s Digital Future. Accessed Online
977 (https://www.mckinsey.com/industries/capital-projects-and-infrastructure/our-
978 insights/imagining-constructions-digital-future).
979
980 Morrison, R. (Ed.). (2016). The principles of project finance. Routledge.
981
982 Moser, H. (2016). Barriers to Bankable Infrastructure: Incentivizing Private Investment to Fill
983 the Global Infrastructure Gap. Rowman & Littlefield.
984
985 Osei-Kyei, R., and A. P. C. Chan. 2017. “Risk assessment in public-private partnership
986 infrastructure projects: Empirical comparison between Ghana and Hong Kong.” Constr.
987 Innovation 17 (2): 204–223. https://doi.org/10.1108/CI-08-2016-0043.

38
 988
989 Osei-Kyei, R., & Chan, A. P. (2015). Review of studies on the Critical Success Factors for
990 Public–Private Partnership (PPP) projects from 1990 to 2013. International Journal of
991 Project Management, 33(6), 1335-1346.
992
993 Owolabi, H. A., Bilal, M., Oyedele, L. O., Alaka, H. A., Ajayi, S. O., & Akinade, O. O. (2018).
994 Predicting completion risk in PPP projects using big data analytics. IEEE Transactions
995 on Engineering Management.
996
997 Özdemir, E. (2015). Experience in structuring public-private partnerships for airports. Journal
998 of Airport Management, 9(2), 154-161.
999
1000 Robinson, H.S. and Scott, J., (2009). Service delivery and performance monitoring in PFI/PPP projects.
1001 Construction Management and Economics. 27(2), pp.181-197.
1002
1003 Odeh, A.M. and Battaineh, H.T., (2002). Causes of Construction Delay: Traditional Contracts.
1004 International Journal of Project Management, 20(1), pp.67-73.
1005 Oyedele, L. O. (2012). Avoiding Performance Failure Payment Deductions in PFI/PPP Projects: Model
1006 of Critical Success Factors. Journal of Performance of Constructed Facilities, 27(3), 283-294.
1007
1008 Perera, N. A., Sutrisna, M., & Yiu, T. W. (2016). Decision-making model for selecting the
1009 optimum method of delay analysis in construction projects. Journal of Management in
1010 Engineering, 32(5), 04016009.
1011
1012 Osei-Kyei, R., & Chan, A. P. (2017). Implementing public–private partnership (PPP) policy for
1013 public construction projects in Ghana: critical success factors and policy implications.
1014 International Journal of Construction Management, 17(2), 113-123.
1015
1016 Patton, M. (1990). Purposeful sampling. Qualitative evaluation and research methods, 2, 169-
1017 186.
1018 Sainati, T., Brookes, N., & Locatelli, G. (2017). Special purpose entities in megaprojects: empty
1019 boxes or real companies? Project Management Journal, 48(2), 55-73.
1020
1021 Song, J., Hu, Y., & Feng, Z. (2017). Factors Influencing Early Termination of PPP Projects in
1022 China. Journal of Management in Engineering, 34(1), 05017008.
1023
1024 Suri, H. (2011). Purposeful sampling in qualitative research synthesis. Qualitative research
1025 journal, 11(2), 63-75.
1026
1027 Smith, N.J., Merna, T. and Jobling, P., (2009). Managing risk: in construction projects. John Wiley &
1028 Sons.
1029 Tam, V. W. Y., and Fung, I. W. H. (2008). A Study of Knowledge, Awareness, Practice and
1030 Recommendations among Hong Kong Construction Workers on using Personal Respiratory
1031 Protective Equipment at Risk. Open Construction and Building Technology Journal, 2, 69-81.
1032
1033 True, W.R., (1998). Weather, Construction Inflation could Squeeze North American Pipelines. Oil and
1034 Gas Journal, 96(35).
1035
1036 Ubani, E. C., & Ononuju, C. N. (2013). A Study of Failure and Abandonment of Public Sector Driven
1037 Civil Engineering Projects in Nigeria: An Empirical Review. American Journal of Scientific and
1038 Industrial Research, 75-82.

39
1039
1040 Usman, N. D., Inuwa, I. I., Iro, A. I., & Dantong, S. (2012). The Influence of Unethical Professional
1041 Practices on The Management of Construction Projects in North Eastern States of Nigeria.
1042 International Journal of Economic Development Research and Investment, 3(2), 124-129.
1043
1044 Van Staveren, M.T., (2006). Uncertainty and Ground Conditions. A Risk Management Approach.
1045
1046 Wibowo, A., & Alfen, H. W. (2015). Government-led critical success factors in PPP
1047 infrastructure development. Built Environment Project and Asset Management, 5(1), 121-
1048 134.
1049
1050 Xu, Y., Chan, A. P., Xia, B., Qian, Q. K., Liu, Y., & Peng, Y. (2015). Critical risk factors
1051 affecting the implementation of PPP waste-to-energy projects in China. Applied energy,
1052 158, 403-411.
1053
1054 Yang, J.B. and Wei, P.R., (2010). Causes of Delay in the Planning and Design Phases for Construction
1055 Projects. Journal of Architectural Engineering, 16(2), pp.80-83.
1056
1057 Yescombe, E.R., (2013). Principles of project finance. Academic Press.
1058
1059 Ye, S., & Tiong, R. L. (2003). The Effect of Concession Period Design on Completion Risk Management
1060 of BOT Projects. Construction Management and Economics, 21(5), 471-482.
1061
1062 Zhang, H., (2007). A redefinition of the project risk process: Using vulnerability to open up the event-
1063 consequence link. International Journal of Project Management. 25(7), pp.694-701.
1064

40