Featured Product

    Examining Infrastructure as an Asset Class

    September 2021

    Examining Infrastructure as an Asset Class

    Summary

    This is an update to our article published on 5 May 2020. Project Finance is a well-managed asset class and we continue to observe stability in both the cumulative default and recovery rates in our consortium dataset. This not only indicates low volatility, but also demonstrates accuracy in the assessment of risk. Project finance loans have higher risk in the early years of construction and ramp-up and lower risk in the operational years. This is a stark contrast to corporate lending where risk tends to increase over time. This plays well for banks with expertise in assessing construction risk for short term lending and for asset managers and insurance companies looking for longer maturities for asset liability matching. Key findings from this analysis include:

    » The cumulative default rates for infrastructure loans are consistent with low investment-grade corporate loans. By year 11, infrastructure loans are comparable with Baa3-rated corporate loans (5.25% vs. 5.39%). By year 20, infrastructure loans are comparable with A-rated corporate loans (5.35% vs. 5.32%).
    » Project and default concentrations are mostly consistent between the Infrastructure subsectors except in Social Infrastructure, with 13.6% of projects and 4.5% of the defaults. 5.5% of the projects are in Media Distribution and Telecom with 9.7% of the defaults. Transportation holds 15.3% of the projects with 22.1% of the defaults. Transportation defaults peaked between 2009–2015. Media Distribution and Telecom defaults were concentrated from 1998-2003.
    » The cumulative default rates for infrastructure loans in high-income countries perform as well as Baa3-rated corporate loans by year 10 (4.96% vs. 4.88%) and as Baa3-rated in middle- and low-income countries by year 14 (7.07% vs 7.14%).
    » The cumulative default rates for infrastructure PPP Loans in High-Income countries are consistent with Baa3-rated corporate loans by year 8 (3.71% vs. 3.76%). Infrastructure non-PPP loans in high-income countries are consistent with Baa3-rated corporate loans by year 12 (5.84% vs. 5.87%). In middle- and low-income countries, infrastructure Public-Private Partnership (PPP) loans are consistent with Baa3-rated corporate loans by year 11 (5.05% vs. 5.39%). Infrastructure non-PPP loans are consistent with Baa3-rated corporate loans by year 15 (7.32% vs. 7.67%).
    Ultimate recovery rates for infrastructure loans averaged 83.8% compared to 79.1% for the consortium dataset.
    » Ultimate recovery rates for infrastructure PPP loans in high-income countries are 80.4% compared to 89.5% in middle- and low-income countries. Ultimate recovery rates for infrastructure non-PPP loans are 84.0% compared to 82.9% in middle- and low-income countries.

    Overview

    Moody’s Analytics Data Alliance Project Finance Consortium was established in 2010 with leading sector lenders to study the credit profile of project finance bank loans. In response to increasing regulatory pressure and funding costs, the consortium was designed to enhance the understanding of default and recovery behavior and to facilitate further evolution of the asset class. Furthermore, the participants—which now number 81 global institutions (banks, insurance companies, and asset managers)—view the consortium as a key to promote transparency, increase liquidity, and assist with capital relief under Basel capital requirements.

    Moody’s Analytics collected data from 9,332 project loans and losses (consortium dataset) from the consortium members. The project loans originated from 1983–2019. The dataset contains 611 project loan defaults based on the Basel definition of default and 306 ultimate recoveries. The definitions of project finance, the Basel definition of default, and ultimate recovery can be found in the glossary.

    Working with Moody’s Investors Service, a study has been published each year. This research highlights current trends and gives an overview of the performance of the 9,332 projects and covered the period from 1983–2019. The link to the Moody’s Investors Service March 9, 2021 published report can be found in the glossary.

    For this analysis, 7,357 of the 9,332 projects—adopting a wider definition of infrastructure than the social, transportation and environmental sectors than the annual study uses—were selected from subindustries within the social, transportation, water and waste, media distribution and telecom, oil and gas, and power sectors. The subindustries include 1,004 social projects with 20 defaults, 1,128 transportation projects with 98 defaults, 312 water and waste projects with 19 defaults, 74 other infrastructure projects with 4 defaults, 406 media distribution and telecom projects with 43 defaults, 287 oil and gas distribution and refining projects with 17 defaults, and 4,146 power generation and transmission projects with 242 defaults. Within infrastructure, there are 6,669 projects in high-income countries with 363 defaults, and 1,001 projects in middle- and low-income countries with 95 defaults. In addition to the 7,357 infrastructure loans, there are also 1,601 non-infrastructure loans that were not included in the calculations. The definitions of infrastructure and non-infrastructure are in the appendix.

    The purpose of this research is to highlight the credit behavior of infrastructure project loans and compare the performance of high-income economies with middle- and low-income economies. Cumulative default rates (CDRs) from subsets of the 7,357 infrastructure loans will be compared to CDRs taken from Moody’s Investors Service-published research. The research focuses on default and recovery rates for corporate bond, loan and deposit issuers rated in the Aaa, A, Baa, Ba, and B rating categories for the period 1983–2019. A link to that report can be found in the glossary.

    Database composition

    Figure 1 presents the concentration of projects and defaults for the infrastructure project loans in each sector. Most of the industry sectors have similar concentrations of projects and defaults. Social infrastructure projects show a lower default concentration than the remaining industries, while media distribution and telecom and transportation show higher default concentrations.

    Figure 2 presents the concentration of projects and defaults for the infrastructure project loans in each region. Most of the regions have similar concentrations of projects and defaults. Latin America and Asia have larger concentrations of defaults; however, they also have lower project counts.

    Figure 3 presents the concentration of projects and defaults for the infrastructure project loans by country income classification. The overwhelming majority of the data is in high-income countries.
    Figure 4 presents the construction status of the project. More than twice as many projects are financed in construction compared to in operations. Most projects transitioned to operations prior to the default and the default count in operations is substantially larger than in construction. With an average time to default of 4.4 years, this is sometime towards the end of construction/beginning of operations. Recoveries on the defaults in construction are lower than recoveries in operations where almost two-thirds of the projects received a 100% recovery.
    Figure 5 presents the comparison of project and default counts for PPP projects and Non-PPP projects. PPP projects have a lower incidence of default compared to non-PPP projects.

    Infrastructure debt performance

    Figure 6 compares the 20-year CDR for the infrastructure project loans of 5.4%. It is consistent with the 20-year CDRs for corporate issuers of low investment-grade credit quality. By comparison, corresponding CDRs for Baa1, 2, and 3-rated; Ba1, 2, and 3-rated; and B-rated corporates are 5.7%, 7.5%, 11.0%, 17.2%, 20.2%, 38.8%, and 48.2% respectively.

    The CDRs from the infrastructure project loans flatten out and perform as investment grade by year 10. This indicates a decreasing amount of risk as the loans mature—unlike corporate debt, which displays increasing CDRs over time. The marginal default rates— the likelihood of defaulting during a given year—for infrastructure project loans are less than 0.01% by year 17 and 0% as the loans approach 20 years. This suggests that projects, following construction and rampup, are less likely to default as they mature.

    Figure 7 compares the 20-year CDRs for the infrastructure project loans from each region. infrastructure loans from 1983–2019, to CDR data. The data was taken from Moody’s Investors Service published research on default and recovery rates for corporate bond, loan, and deposit issuers rated in the Aaa, A, Baa, Ba, and B rating categories for the period 1983–2019. The regional analysis also shows curves flattening beginning in years 4 through 6.

    The 20-year CDR of 5.4% is consistent with the 20-year CDRs for corporate issuers of low investment-grade credit quality. By comparison, corresponding CDRs for Baa1, 2, and 3-rated; Ba1, 2, and 3-rated; and B-rated corporates are 5.7%, 7.5%, 11.0%, 17.2%, 20.2%, 38.8%, and 48.2% respectively.

    Figure 8 compares the 20-year CDRs for the non-infrastructure project loans from each region. infrastructure loans from 1983–2019, to CDR data. The data was taken from Moody’s Investors Service published research on default and recovery rates for corporate bond, loan, and deposit issuers rated in the Aaa, A, Baa, Ba, and B rating categories for the period 1983–2019. The regional analysis also shows curves flattening beginning in years 4 through 6.

    Figure 9 compares CDRs from high-income and middle- and low-income countries’ infrastructure loans from 1983–2019, to CDR data. The data was taken from Moody’s Investors Service published research on default and recovery rates for corporate bond, loan, and deposit issuers rated in the Aaa, A, Baa, Ba, and B rating categories for the period 1983–2019.

    The 20-year CDR of 5.2% for high-income countries and 7.1% for middle- and low-income countries are consistent with the 20-year CDRs for corporate issuers of low investment-grade credit quality. By comparison, corresponding CDRs for Baa1, 2, and 3-rated; Ba1, 2, and 3-rated; and B-rated corporates are 5.7%, 7.5%, 11.0%, 17.2%, 20.2%, 38.8%, and 48.2% respectively.

    The CDRs from infrastructure loans in high-income countries flatten out and perform as investment grade at year 10. Middle- and low-income country loans perform at investment grade by year 14.

    Figure 10 compares CDRs from high-income countries’ infrastructure loans by sector from 1983–2019, to CDR data. The data was taken from Moody’s Investors Service published research on default and recovery rates for corporate bond, loan, and deposit issuers rated in the Aaa, A, Baa, Ba, and B rating categories for the period 1983–2019.

    The 20-year CDRs for high-income countries in Energy, Media Distribution & Telecom, Social, Transport, Water and waste, and Other infra are 5.8%, 9.0%, 0.9%, 10.2%, 3.4%, and 5.4%respectively and are consistent with the 20-year CDR rates for corporate issuers of low investment-grade credit quality. By comparison, corresponding CDRs for Baa1, 2, and 3-rated; Ba1, 2, and 3-rated; and B-rated corporates are 5.7%, 7.5%, 11.0%, 17.2%, 20.2%, 38.8%, and 48.2% respectively. The 20-year CDR for high-income countries Social infrastructure loans was 0.9% crossing into investment grade by year 1 and as a A-rated debt by year-6.

    Within the high-income country infrastructure, there are 3,812 Energy loans with 205 defaults, 299 Media Distribution & Telecom loans with 27 defaullts, 975 Social with 17 defaults, 997 Transport loans with 87 defaults and 277 Water & waste loans with 14 defaults and 59 Other infrastructure loans with 3 defaults.

    Figure 11 compares CDRs from middle and low-income country infrastructure loans from 1983–2019 to CDR data taken from Moody’s Investors Service published research on default and recovery rates for corporate bond, loan, and deposit issuers rated in the Aaa, A, Baa, Ba, and B rating categories for the period 1983–2019.

    The 20-year CDRs for middle- and low-income countries in Energy, Media Distribution & telecom, Social, Transport, Water and waste, and Other infra are 6.0%, 14.4%, 9.0%, 6.5%, 9.6%, and 12.1%respectively and are consistent with the 20-year CDR rates for corporate issuers of low investment-grade credit quality. By comparison, corresponding CDRs for Baa1, 2, and 3-rated; Ba1, 2, and 3-rated; and B-rated corporates are 5.7%, 7.5%, 11.0%, 17.2%, 20.2%, 38.8%, and 48.2% respectively.

    The infrastructure loan CDRs for middle- and low-income country loans in the Transportation sector perform as investment grade by year 12. Energy loans perform as investment grade by year 14 and Other loans perform as investment grade by year 19.

    Within the middle- and low-income country infrastructure, there are 621 Energy loans with 54 defaults, 107 Media Distribution & telecom loans with 16 defaullts, 25 Social with 3 defaults, 131 Transport loans with 11 defaults and 35 Water & waste loans with 5 defaults and 15 Other infra loans with 1 defaults. Social and water infrastructure sectors perform better in high-income countries. The transport infrastructure sector performs better in middle- and low- income countries relative to high-income countries.

    Figure 12 compares Regional CDRs from the global consortium dataset to CDRs from infrastructure loans. Infrastructure loans perform as well as or better than the global consortium dataset in North America and Western Europe—the two largest markets—and Africa, Asia, and the Middle East. Infrastructure loans perform slightly worse in Eastern Europe, Latin America, and Oceania.
    Figure 13 displays the count of infrastructure project loans and defaults and consortium project loans and defaults as supporting information to Figure 12.
    Figure 14 compares Regional Ultimate Recovery Rates from the Consortium dataset to Ultimate Recovery Rates from infrastructure loans. The average recovery for infrastructure project loans is 83.3%, which is higher than the consortium dataset average recovery of 79.1%. Infrastructure loans perform as well as or better than the consortium dataset in North America and Western Europe—the two largest markets—as well as in Africa, Asia, the Middle East, and Oceania. Infrastructure loans received slightly lower recoveries in Eastern Europe, and Latin America.
    Figure 15 compares the recovery statistics from infrastructure project loans in high-income and middle- and low-income countries. High-income countries perform longer than middle- and low-income countries prior to default with 4.7 and 3.3 years respectively and remain in default shorter with 2.2 years and 3.0 years respectively.
    Figure 16 compares the Expected Loss (EL) PD*LGD, by country income classification for infrastructure loans to the EL from Moody’s Investors Service published research. The research focuses on default and recovery rates for corporate bond, loan, and deposit issuers rated in the Aaa, A, Baa, Ba, and B rating categories for the period 1983–2019. For simplicity, the Moody’s EL contains a single LGD value determined from Moody’s Investors Service Ultimate Recovery Database.

    Figures 17 and 18 compare the recovery statistics from infrastructure project loans to consortium project loans. Infrastructure project loans, with an average recovery of 82.6%, performed better than the consortium project loans, with an average recovery of 77.9%.

    Infrastructure project loans performed for a longer time before a default with an average of 4.4 years. This compares to the consortium project loans in Figure 9 with an average time to default of 4.2 years. The average time in default is 2.3 years for the infrastructure project loans, compared to 2.4 years in Figure 10 for consortium project loans.

    Figure 19 compares the Expected Loss (EL) PD*LGD, by region for infrastructure loans to the EL from Moody’s Investors Service published research. The research focuses on default and recovery rates for corporate bond, loan, and deposit issuers rated in the Aaa, A, Baa, Ba, and B rating categories for the period 1983–2019. For simplicity, the Moody’s EL contains a single LGD value determined from Moody’s Investors Service Ultimate Recovery Database.
    Figure 20 contains supporting information for Figure 19 and displays project and default counts with their associated LGD for infrastructure loans in middle- and low-income countries.
    Figure 21 compares the Expected Loss (EL) PD*LGD, by region in middle- and low-income countries for infrastructure loans to the EL from Moody’s Investors Service published research. The research focuses on default and recovery rates for corporate bond, loan, and deposit issuers rated in the Aaa, A, Baa, Ba, and B rating categories for the period 1983–2019. For simplicity, the Moody’s EL contains a single LGD value determined from Moody’s Investors Service Ultimate Recovery Database.
    Figure 22 contains supporting information for Figure 21 and displays project and default counts with their associated loss given default (LGD) for infrastructure loans in high-income countries.

    Figure 23 displays the concentration of PPP projects in high-income countries compared to middle- and low-income countries. There are proportionally more PPP projects in middle- and low-income countries; however, the overall count is substantially smaller.

    76.1% of PPP projects in high-income economies are social and transport whereas only 40.5% are social and transport in middle- and low-income economies. 10.6% of PPP projects in high-income economies are power, whereas 33.0% are social and transport in middle- and low-income economies.

    Figure 24 compares CDRs from middle- and low-income country infrastructure PPP and non-PPP loans from 1983–2019, to CDR data taken from Moody’s Investors Service published research. The research focuses on default and recovery rates for corporate bond, loan, and deposit issuers rated in the Aaa, A, Baa, Ba, and B rating categories for the period 1983–2019.

    Middle- and low-income PPP loans perform as well as Baa3-rated debt by year 7 and non-PPP loans by year 16.

    Figure 25 compares the CDRs from middle- and low-income PPP and non-PPP projects. Non-PPP projects perform as investment grade by year 15 and PPP loans perform as investment grade by year 7.
    Figure 26 presents the project and default counts and loss for high-income and middle- and low-income countries.
    Figure 27 presents the project and default counts and loss by country income classification for PPP projects.
    Figure 28 presents the project and default counts and loss by country income classification for non-PPP projects.
    Figure 29 presents the performance of loans by origination year. Earlier cohorts have higher CDRs than newer cohorts. Earlier cohorts also have more time in which the loan can default. Loans originated from 1990-2019, 2000-2019 and 2010-2019 have an average time to default of 4.7 years, 4.5 years, and 3.3 years respectively. 80% of the defaults occurred before 2010. Also, newer loans benefit from structural improvements in the project as well as the loan covenants.
    Figure 30 presents the performance of infrastructure loans by origination year in high-income countries. The cohort behavior is the same as figure 29.
    Figure 31 presents the performance of infrastructure loans by origination year in middle- and low-income countries. The cohort behavior is the same as figures 29 and 30.
    var EEWD = EEWD || {}; EEWD.advanced_automation = EEWD.advanced_automation || {}; EEWD.advanced_automation.marketo_rest = 1; (function(){ var s = document.createElement('script'); s.type = 'text/javascript'; s.async = true; var host = (document.location.protocol == 'http:') ? 'cdn.snapapp.com' : 'scdn.snapapp.com'; s.src = '//' + host + '/widget/widget.js'; s.id = 'eeload'; document.getElementsByTagName('head')[0].appendChild(s); })();