Leveraging Private Capital for Infrastructure Renewal (2019)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

16 C H A P T E R 3 P3 developers in the United States and internationally generally implement P3 projects through the establishment of an SPV. As previously described, an SPV is a private company formed with the special purpose of developing, operating, and maintaining the infrastructure asset in question under a long-term concession or other agreement. Financing for the P3 project is typically done through limited or nonrecourse debt (loans or bonds), referred to as project financing, and private-sector equity. In project financing, lenders/bond investors (referred to generally herein as lenders or debt providers) look to the SPV, its performance and resulting cash flows, and security over the SPV’s assets for repayment of their loans/investment. They have no or limited recourse (depending on the precise structure) to the public sponsor or the equity investors themselves in the event that the SPV is unable to make the required principal and interest payments on its debt. Properly structured, the P3 contractual arrangements, including the limited or nonrecourse structure, transfer certain project risks to the SPV, and performance requirements for the SPV can enable the public sponsor to achieve P3 procurement objectives in infrastructure development. This chapter discusses the capital/financial structure and the application of project finance in P3 transactions as well as general characteristics of equity and equity investors in U.S. trans- portation P3 projects. It also includes a discussion on the evolution of the global infrastructure investment market. 3.1 Equity in the Project Capital Structure The P3 procurement model effectively overlays a corporate capital structure on a public works project. By contrast, under traditional procurement, the equity owners for that project are effectively the taxpayers, represented by their elected officials and public agencies managing the procurement and construction of the project. Under a traditional procurement, the sponsoring government agency generally issues debt to pay for the capital costs of the project. In a P3 transaction, debt through a project financing structure and private-sector equity generally provide some or all of the up-front capital required to fund the project’s costs. As mentioned, lenders/investors in the project rely on the project’s cash flows and security over the SPV’s assets as the means of recovery of capital invested and return on that capital. As a result, if the cash flow generated by a given project is insufficient to repay debt and investors, lenders and equity investors generally have no guarantee or only a limited guarantee of some form from the government to make them whole. Lenders to P3 projects typically require private-equity investment in the SPV as a portion of the required total financing precisely because of the nonrecourse or limited-recourse nature of the project’s debt. Since there is no (or in some instances a very limited) government guarantee The Role of Private Equity in Public–Private Partnerships

The Role of Private Equity in Public–Private Partnerships 17 with respect to debt repayment, lenders are generally unwilling to provide all the capital required to fund the projects. Lenders want to see that the P3 developer has an economic interest in ensuring project success through its equity that is at risk. That is, equity at risk ensures that the P3 developer has incentive to work diligently to achieve strong project performance so that the SPV generates sufficient cash flow to repay debt. Typically, it is only after debt principal and interest payments are made that the equity investor receives a return on its capital since it is lower in order of priority in terms of payments from the SPV’s available cash flow. In a P3 trans- action, if project revenues are too low or costs are too high to allow the SPV to pay debt principal and interest (together referred to as debt service), the SPV may eventually default on its debt and enter the bankruptcy process. Equity investors may lose their equity investment in the SPV, and lenders generally exercise step-in rights, which are rights to step-in to manage the project. On the other hand, debt providers for traditional public projects funded with public or municipal debt do not require equity at risk. In this case, the lenders look to the public-sector entity issuing the debt for repayment. There is an implicit or, in many cases, explicit public- sector guarantee of the debt, and the government often has certain means to ensure sufficient cash flow for debt repayment through its power of taxation or control of user fees (tolls, etc.). Governments often pledge to bondholders that they will raise taxes or user fees on their citizens such that they will be able to pay debt service as necessary. In this way, the burden of debt repay- ment ultimately falls on the taxpayers of the borrowing government. 3.2 Differentiating Debt and Equity Returns The cost of capital is the blended rate of return required by lenders and investors such that they are willing to provide debt (in the case of lenders) or equity (in the case of equity investors) to the SPV to fund the project in question. With respect to debt, the cost is the interest rate. The interest rate on the debt and the schedule of principal payments necessary to repay the debt are generally known from the outset. [This differs for variable-rate debt, which is a very small percentage of the overall municipal bond market, averaging less than 2% of total municipal issuances annually over the previous 5 years (SIFMA data).] Projects with variable interest rate debt are often required by lenders to fix the interest rate with respect to at least a substantial portion of the debt through hedging. Equity returns are neither fixed nor predetermined. They vary depending on project perfor- mance. Private-sector equity investors could lose their entire investment in an SPV or receive significant profits from their investments. The investment returns can fluctuate widely depend- ing on the project, its phase in the development cycle (e.g., preconstruction, construction, ramp-up, stabilization), and a multitude of other factors. While equity investors typically cal- culate a projected, or pro-forma, return on equity when preparing a proposal for a project, the actual rate of return, referred to also as return on equity, is uncertain until all of the project’s actual cash flows are known at the end of the project’s life many years after the initial invest- ment or, if earlier, when the equity investor sells its equity interest in the project. This makes any public discussion or review of the cost of equity much more difficult than that of the cost of different debt options. To appropriately compare the cost of delivering a project through a P3 versus a traditional public procurement model, governments considering the use of a P3 must compare the risk- adjusted cost of delivering the project via a P3 with the risk-adjusted cost of delivering the project via traditional procurement. Comparing the cost of capital between the two alone does not give the full picture because risks in a P3 are transferred from the public sector to the P3 developer, and this benefit must be incorporated into the analysis. Specifically, in most cases, core line items in this comparison include the value of the risks transferred to the P3 developer

18 Leveraging Private Capital for Infrastructure Renewal versus financing costs, which for a P3 are generally higher because of the higher cost of capital that P3s carry. This comparison is subject to policy debate and strong public discourse regarding the use of P3s for infrastructure projects. Public-sector infrastructure officials and practitioners of traditional public finance (public finance includes the issuance of municipal or other public-sector debt to government entities) are familiar with debt instruments. They often frame equity investments as a more expensive form of capital when compared to municipal bonds or other forms of public debt issuance. As described previously, this does not account for the full picture since equity takes more risk and allows for the transfer of risk from the public sector to the private sector in P3 transactions. This can benefit the public sector by lowering overall project cost and providing additional certainty with respect to project cost to the public sector. 3.3 Project Finance Without Equity: The 63–20 Model There is some precedent in the United States for the use of project finance for public infra- structure projects without an equity investment by the private partner. This demonstrates some of the benefit that private-sector equity at risk brings. A no-equity example is the 63–20 model, named after the IRS revenue ruling that originally authorized its use. Under a 63–20, a nonprofit corporation is created to issue nonrecourse, tax-exempt municipal bonds to finance a project. To use 63–20, the nonprofit must be for an essentially public project, and the sponsoring gov- ernment must retain ownership of the asset after the project’s debt has matured. The model has commonly been used to develop social infrastructure to be leased by the sponsoring govern- ment, but it has also been used to finance some notable transportation projects (Kinnander 1999), including the Pocahontas Parkway in Virginia and the Southern Connector project in South Carolina (see case example that follows)—both toll roads. Governments choose to use the 63–20 model for a variety of reasons, including facilitating the use of P3s for infrastructure projects while still being able to access the lower cost of capital of the municipal bond market. Another considerable driver of their use appears to be the avoidance of statutory limits on general obligation public borrowing or the need to pass ballot measures to approve additional public borrowing. A significant number of 63–20 projects also involved lease arrangements, in which the project debt, while technically nonrecourse to the sponsoring government, was secured by a lease of the facility by the government (Robbins and Meulen 2009). For a period, these leases were accounted for as current liabilities as opposed to long- term obligations, although U.S. government accounting standards have more recently adjusted to require governments to report their leases as public obligations (Government Accounting Standards Board 2017), which would appear to be more appropriate. This will presumably limit the incentives for governments to use the model to maneuver around statutory debt limits and liability accounting standards. 3.3.1 Case Example: Southern Connector Project The Southern Connector project illustrates some of the problems with the 63–20 model. The state of South Carolina formed a nonprofit to develop, construct, and operate the Southern Con- nector toll road in 1998. The project was financed by selling approximately $200 million of non- recourse municipal bonds. While the bonds were technically not a general obligation of the state, the nonprofit corporation contracted the DOT to maintain the road, and the state agreed to adjust toll rates to ensure that sufficient revenue was generated by the project to service the debt. Once construction was completed and the asset was operational, actual traffic volumes on the road were significantly lower than forecast, leading to eventual bankruptcy litigation

The Role of Private Equity in Public–Private Partnerships 19 between the project’s bondholders and the state in 2010. The bondholders argued that the state did not live up to its agreement to increase toll rates sufficiently to service debt and was therefore responsible for some of the losses. The state argued that the debt was nonrecourse and further that the nonprofit corporation was not technically an instrument of the govern- ment. In the end the bankruptcy was settled with all parties sharing some of the burden, though most of the losses were borne by the creditors. The project’s debt was refinanced at a significant loss to bondholders. Subsequently, the state agreed to forego past unpaid and future license fees from the nonprofit corporation and to assume responsibility for maintain- ing the road, with repayment by the nonprofit out of future project revenues (Connector 2000 Association, Inc. 2010). It is not clear whether the Southern Connector project would have turned out differently had it been structured as a traditional P3 with an equity investor holding interest in the project company. In that scenario, the equity owners of the project company would certainly have been incentivized to manage or design the project to maximize demand and mitigate revenue risk and to conduct more diligence with respect to the traffic forecast. The Southern Connec- tor’s bankruptcy resolution largely prevented the liabilities of the project from becoming a burden on the state, but the project and a few other notable examples (such as the Pocahontas Parkway and Las Vegas Monorail projects) will likely render the use of the 63–20 model infea- sible for future transportation projects that entail significant revenue or other risks. Without aligned equity investors in the project company or some form of guaranteed or dedicated pub- lic revenue (such as a lease payment), lenders are unlikely to completely finance risky projects under the model. 3.4 Characteristics of Equity Investments in U.S. Transportation P3s To assess the trends in investment in the U.S. transportation sector, this synthesis focuses on a data set of 28 bridge and highway P3s that involved equity investments and that reached financial close no later than 2017. These projects were procured in 12 states and Puerto Rico. As of 2017, only Texas, Virginia, California, Florida, Indiana, and Colorado have used the P3 procurement model for more than one project. A complete list of the projects included in the data set, organized by financial close date, is shown in Table 1. 3.4.1 Project Type, Funding Sources, and Leverage While all the projects in the synthesis data set are road or bridge projects, the sample is highly idiosyncratic by project type. Many of the projects are new toll roads, while some are managed or HOT-lane projects designed to manage congestion on crowded corridors. Others are more unusual, such as the Rapid Bridge replacement project in Pennsylvania, which is for the replace- ment and maintenance of hundreds of small, non-tolled bridges throughout the state. While most of the projects are for new capacity, several are for replacement projects, such as the Goethals Bridge Project in New York. A minority are brownfield concessions on existing toll roads or networks, such as the sale of a concession on the ITR and the Chicago Skyway. It is important to note that the project type column in Table 1 refers to the revenue source for the SPVs, but does not necessarily indicate whether the project facility generates revenue or user fees. In other words, the toll revenue risk P3s are those in which the private partner takes all or some demand/revenue risk for the project. Some of the AP P3s, such as the Pennsylvania Rapid Bridge replacement and Presidio Parkway, are non-tolled facilities. Others, such as the Goethals Bridge, are tolled, although the SPV itself receives its revenue via an AP over time. For that project, the public sponsor is retaining demand risk and collecting tolls, while the project

20 Leveraging Private Capital for Infrastructure Renewal company is compensated by an AP. Toll revenue is the project company revenue source for 18 out of 28 projects in the data set. The remaining 10 concessions are funded via APs. Total equity investments in projects within the data set were approximately $8.3 billion in aggregate. The average leverage ratio of the projects (debt/total financing) was 76%, but this varied significantly by the revenue source for the SPV. The average leverage ratio for the AP transactions was significantly higher at 87%. The average leverage ratio for the toll-funded trans- actions was significantly lower at 70%. This is intuitive since the added demand risk trans- ferred under toll revenue transactions necessitates a lower-leveraged capital structure and thus increased equity investment to secure financing. More recent toll revenue projects, such as Transform 66, exhibit higher ratios of equity in projects, whereas a decade ago it was reasonable Project Name Location Capital Cost (Millions) Equity Investment (Millions) Project Type Financial Close Dulles Greenway VA $350 $138.8 Toll revenue risk 6/1/1989 91 Express Lanes CA $130 $33.8 Toll revenue risk 9/1/1989 SR-125 South CA $773 $130.0 Toll revenue risk 9/1/1989 I-495 HOT Lanes VA $1,998 $350.0 Toll revenue risk 7/1/2002 Chicago Skyway IL $1,830 $855.0 Toll revenue risk 3/1/2004 I-95 Express HOT Lanes VA $940 $292.0 Toll revenue risk 10/1/2004 I-635 LBJ Managed Lanes TX $2,800 $700.0 Toll revenue risk 5/1/2005 Port of Miami Tunnel FL $1,263 $80.3 AP 9/1/2005 Indiana Toll Road IN $3,850 $760.0 Toll revenue risk 9/1/2005 SH-130 Segments 5 and 6 TX $1,358 $209.8 Toll revenue risk 1/1/2006 Northwest Parkway CO $603 $266.9 Toll revenue risk 9/1/2006 North Tarrant Express Phase 1 TX $2,047 $426.0 Toll revenue risk 12/1/2006 I-595 Managed Lanes FL $1,814 $208.0 AP 10/1/2007 Midtown Tunnel Norfolk VA $2,100 $272.3 Toll revenue risk 3/1/2008 PR-22, PR-5 PR $1,080 $421.0 Toll revenue risk 4/1/2008 Presidio Parkway CA $365 $45.6 AP 2/1/2010 North Tarrant Express Phase 2 TX $1,400 $418.3 Toll revenue risk 5/1/2010 Goethals Bridge NY/NJ $1,500 $106.8 AP 11/1/2010 I-77 HOT Lanes NC $635 $248.0 Toll revenue risk 1/1/2012 U.S. 36 HOT Phase 2 CO $113 $20.6 Toll revenue risk 2/1/2012 Ohio River Bridge East End Crossing IN $1,180 $78.1 AP 3/1/2012 I-4 Ultimate Managed Lanes FL $2,323 $103.5 AP 3/1/2013 I-69 Upgrade IN $370 $115.8 AP 5/1/2013 SH-288 Toll Lanes TX $800 $375.3 Toll revenue risk 5/1/2013 Portsmouth Bypass OH $557 $49.0 AP 6/1/2013 Rapid Bridge Replacement PA $1,119 $59.4 AP 12/1/2013 Transform 66 VA $3,724 $1,525.0 Toll revenue risk 11/9/2017 Central 70 CO $1,271 $65.9 AP 12/21/2017 Source: Public Works Financing (2018). Note: Estimated capital costs included in this table are for illustrative purposes only since capital cost estimates for projects often differ slightly in their calculation. Table 1. Synthesis data set.

The Role of Private Equity in Public–Private Partnerships 21 to see a debt-to-equity ratio of 70:30. AP transactions have consistently shown approximately 10% to 15% equity compared to debt, as evidenced by the recent Central 70 project. In addition to equity investments, projects in the data set were financed through combina- tions of bank loans, PABs, and transportation lending from the federal TIFIA program, which provides credit assistance for a portion of project costs for eligible transportation projects (Mallett 2017). Most of the projects (18, or approximately 70%) used TIFIA financing to fund a portion of their costs. Fourteen projects used PABs to finance the remainder of the project debt, while 12 used bank loans alone or in combination with TIFIA. The use of PABs has been more prevalent after the 2008 financial crisis, but both PABs and bank debt were used for highway P3s before and after the crisis. Seventeen of the projects in the sample set also included some form of public sponsor contributions to offset total project costs. These came in the form of cash contributions, land grants, or milestone payments during construction. 3.4.2 Sources of Equity Investment As delineated in the previous section, there may be different investors in project company equity, with the majority of project equity owners in any one project making up a consortium of multiple investors. Within this report, these infrastructure investors are broadly classified as either strategic or financial investors. The strategic investors are the business units of domestic or international construction firms that invest in project companies or companies (predominantly international) that specialize in constructing and sometimes operating infra- structure assets and investing in infrastructure SPVs. Financial investors include institutional investors (providing pooled or direct capital) such as domestic and international pooled fund vehicles, pension funds, endowments, and sovereign wealth funds. Institutional investors in U.S. P3s include Aberdeen, Allianz, APG Infrastructure, Axium Infrastructure U.S., CalPERS, CDPQ, Dallas Police and Fire Pension System, DIF, Northleaf Capital Partners, OMERS, PSP Investments, Sun Life Financial, Teachers Insurance, TIAA-CREFF, and ULLICO (Public Works Financing 2018). Top investor developers for U.S. P3s include ACS Infrastructure, Cintra, John Laing, Plenary Group, and Transurban (Public Works Financing 2018). More detail on the types of financial investors is provided in the next section, including how these types of investors and methods can influence equity return expectations. In the data set, strategic investors were involved more frequently in U.S. P3 consortiums than purely financial investors were. Eighteen projects included both strategic and financial investors in the P3 consortium, and six projects included only strategic investors. Strategic inves- tors often have other contractual relationships with the project, such as the DB contract or providing engineering or maintenance services, and therefore may have more closely aligned investment objectives to ensure positive project outcomes. Purely financial investors play a critical role in aligning incentives for the project company for the projects they participate in, in part because they generally only have an interest in the project company itself and do not have additional contracts to construct or operate and main- tain the asset. Still, the trend of financial investors taking this critical role has not yet successfully emerged in the United States, at least for the highway and bridge P3s included in this report’s data set. Internationally, observations indicate that some projects have included equity owners that are only financial investors. There are two older projects within the data set (the Dulles Greenway in Virginia and the pre-restructuring SR-125 project in California) that could be char- acterized as having primarily financial investors. Both projects underwent restructuring that altered their equity investments significantly. This indicates that the presence of strategic or specialized investors in U.S. transportation projects is all but required given the complexity and high transaction costs required to pursue and manage the risks of those transactions.

22 Leveraging Private Capital for Infrastructure Renewal 3.5 The Nature and Evolution of Financial Investors in P3s As mentioned, the two main types of equity investors in transportation projects are strategic investors (tied to construction, development, or operations companies) and financial investors. Over the last three decades, institutional investors such as pension funds and sovereign wealth funds have started to invest in transportation infrastructure assets around the world. Their involvement in U.S.-based transportation infrastructure investments is expected to grow as more P3 opportunities come to market. Pension funds and sovereign wealth funds have been attracted to these assets because of associated favorable investment characteristics such as low competition and predictable and stable cash flows over the long term (10 years or more). They can use these long-term assets/cash flows to match their long-term liabilities and hedge against inflation. The global infrastructure investment market has centered around regions where gov- ernments have been able to offer a stable legal and regulatory framework under which the invest- ments are made. Accordingly, Western Europe, Australia, and Canada have been the regions among developed countries where most private institutional investor equity investments have been made in transport infrastructure. As is the case for other asset classes into which institutional investors invest, there are a number of different vehicles on offer for private investment in infrastructure. Debt and equity vehicles have both been used by investors to access core economic infrastructure. The infra- structure asset class is heterogeneous, and not all investments satisfy the same risk/return qualities. The vehicle selected for investment will therefore depend on the nature of the asset and how the investor has defined and allocated infrastructure in its portfolio. The various investment vehicles for infrastructure are summarized in Figure 6. There are a large number of financial products available to invest in infrastructure and a large amount of variability within each of the infrastructure products on offer (i.e., no two airports or roads are the same). As the market continues to grow and information about the asset class becomes more readily available, the existing vehicles will become refined and new offerings will emerge. Institutional investors have generally separated core brownfield economic infrastructure from riskier greenfield infrastructure development on the risk/return spectrum, as shown in Figure 7. Interview and survey data indicate that many institutional investors remain reluctant to invest in greenfield infrastructure projects for a variety of reasons, including uncertainty Figure 6. Infrastructure investment vehicles.

The Role of Private Equity in Public–Private Partnerships 23 around construction and political risk and because of the potentially long delay between the commitment of funds and the completion of the project and start of payments (Belt 2013; Della Croce 2012). In terms of fundraising, a 2013 survey found 37% of infrastructure investors to be interested in both greenfield and brownfield assets, and another 11% interested in only greenfield assets (PWC 2016). Of the investors surveyed, 27% were only targeting brownfield assets (Poole 2017). The decision of investors for transport infrastructure is related to how the investment is allo- cated within the investment portfolio (i.e., how investors define transportation infrastructure in their portfolio will influence the type of asset or vehicle invested in). The categories used to group the various investment vehicles for transport infrastructure have included fixed income, infrastructure, equities, alternatives, and private equity. Figure 8 illustrates how the allocation decision of investors will determine the type of vehicle and assets invested in. Most institutional investors have accessed the infrastructure market and U.S. P3 projects through an unlisted equity vehicle. Whether an institutional investor invests directly into a P3 SPV or through a separate intermediary will depend in large part on the size and internal resource capability of the investor. The governance and internal resources of an institutional investor will determine to what extent the investor will partner with or invest through inter- mediaries. The two main structural forms of infrastructure investment (direct or indirect) are illustrated in Figure 9. The indirect investment method involves the institutional investor deploying its capital in an infrastructure fund and passing on all responsibility to the fund manager for the invest- ment. The manager’s role in the process is to source appropriate assets to invest in on behalf of institutional investors. As shown in Figure 9, the direct method refers to the situation where the institutional investor has the capability to forego using a manager for the investment process and can invest directly in infrastructure assets. This means that these investors have the in-house resources to be able to originate investment opportunities, assess and finance the investments, and manage and maintain them. Direct investors can invest alongside co-investment partners in a consortium consisting of other pension fund investors, infrastructure fund managers, and strategic investors. The equity arrangement in a consortium of direct investors varies from asset to asset. In a co-investment Note: OECD = Organization for Economic Cooperation and Development. Figure 7. Infrastructure risk/reward profile.

24 Leveraging Private Capital for Infrastructure Renewal arrangement, it is usually the fund manager or largest pension fund investor in the consortium that leads the transaction. A further development in the direct investing method for institutional investors has been the forming of a club or partnership model consisting of like-minded investors in an attempt to avoid conflicts of interest associated with the fund manager route. Such a model consists of a group of institutional investors collaborating to align interests when approaching infrastructure investments. The rationale for joining a club has been to spread risk, negotiate better terms by reducing competition, and ensure the availability of local knowledge. The club concept is newer in the market and difficult to set up unless interests are well aligned and the decision-making process is efficient. As the size of assets under management has grown for various pension funds Note: SWF = Sovereign Wealth Fund. Figure 8. Investor asset allocation for infrastructure. Figure 9. Methods of institutional infrastructure investment.

The Role of Private Equity in Public–Private Partnerships 25 and sovereign funds, the cost economics of using third-party managers has become higher and higher, which has led several funds to move toward the direct model of investing or the collab- orative/partnerships model. [For more discussion on the shift toward the collaborative model of equity infrastructure investing, see Monk, Sharma, and Sinclair (2017), Reframing Finance. SUP, Stanford.] Most DBFOM projects in the United States have involved consortiums consisting of fund managers and strategic investors. This has meant that the majority of institutional investor capi- tal has been channeled through fund managers into U.S. P3 projects. While one might expect the return expectations of a pension or sovereign fund direct investor and a fund manager to be similar for the asset class, this might not always be the case. The typical private-equity fund structure adopted by most infrastructure fund managers provides performance incentives that push these firms to maximize their rate of return as much as possible, in a shorter time frame than what a pension or sovereign fund investor might be looking for in an infrastructure invest- ment. Fund managers often use significant amounts of leverage to not only finance the large investments required in infrastructure, but to maximize the return on the equity investment made. Leverage has also been used by fund managers to pay dividends so that the internal rate of return (IRR; the measure of performance of the investment) is higher. [See Phalippou (2008). The Hazards of Using IRR to Measure Performance: The Case of Private Equity. Available at SSRN 1111796.] Because IRR is time dependent, to inflate this value, fund managers are not incentivized to hold stable investments for a long time. A good perceived IRR helps fund man- agers raise future funds and collect further fees. All of these strategies and incentives add risk to the investments that might distort the characteristics of the underlying infrastructure assets that investors are after. In this way, direct investors may have different return expectations from their infrastructure investments compared with investors using the indirect, fund manager route.