Project bonds have been a critical source of debt financing in the project finance space for many years. Most commonly, a project sponsor will seek to issue project bonds to refinance existing bank debt to provide long-term funding for projects that have already reached a relatively stable and predictable stage of operation.
Unsurprisingly, these refinancings are far from the only example of project bond issuances, as the details of each issuance, sponsor and transaction will dictate variations. For example, investors in project bonds do not shirk away entirely from purchasing bonds that fund projects that are still under construction. Additionally, project bonds can be issued by the project company (i.e., the entity that directly owns the project being financed) or by a holding company that indirectly owns one or more projects to finance a portfolio of operating assets or to take advantage of a structurally subordinated repayment obligation akin to back-leveraged bank debt. And, though they may seem slight to an outside observer, there are substantial differences between the types of bond issuances that predominate in the market and the rules that govern them, which in turn have procedural and legal implications for the pool of investors, the disclosures that precede the issuance, and the covenants and governance structure of the transaction itself after the bonds are issued and effective.
This chapter provides an overview of key features of project bond issuances, encompassing those features that are common to any issuance (including the important financial terms, the general approach to covenants and monitoring, and the role of diligence and credit support) and how the nature of a project bond transaction affects those key features, with a focus on the procedural and substantive differences that distinguish offerings made under Section 4(a)(2) and Rule 144A of the Securities Act – the two principal safe harbours for issuing project bonds that are exempt from registration under the US securities laws.2 Tangential to that discussion will be a focus on intra-bondholder and inter-creditor issues, along with potential workarounds and compromises, in the context of administering a given transaction. The chapter will also discuss project bonds in comparative terms to bank debt financings.
II MARKET OVERVIEW3
Over the past decade, the project bond market both in the US and globally has been an increasingly active space, with year-to-year variations up from a recent low in global project bond volumes of approximately US$9 billion in 20094 following the financial crisis to at least US$47.393 billion worldwide in 2019. The overall amount of global project bond issuances represents a slight increase over 2018, when US$46.846 billion in project bonds were issued worldwide. (The 2019 figure, by contrast, is down from a recent high of approximately US$64 billion in worldwide project bond issuances in 2017.)5 In 2019, the US project bond market was by far the most active (based on the aggregate dollar amount of issuances), accounting for just shy of US$21.8 billion of project bond issuances. This is in keeping with recent end-of-year results, which have generally shown the US project market to be the largest – one exception was 2018, when European project bond issuances totalled US$21.981 billion as compared with US$10.67 billion in US issuances. For instance, according to materials published on Practical Law, the aggregate dollar amount of project bonds placed by US issuers has risen from approximately US$7.1 billion in 2012 to approximately US$18.8 billion in 2017. Aside from North America (the US and Canada), which saw a pronounced increase in year-over-year project bond volumes from 2018 to 2019, driven by the significant boost in US project bond issuances, and Europe, which saw a pronounced decrease in issuances (to US$9.589 billion), other regions saw roughly consistent activity (or lack thereof, in the Middle East and Africa) in terms of project bond issuances in 2019 as compared to 2018. While overall volumes have decreased from the recent 2017 high, project sponsors are still seeking access to the international bond markets, illustrated by the following example transactions, each of which closed in 2019:
- the US$1.5 billion Rule 144A/Reg S offering of notes due 2029 by Cheniere Corpus Christi Holdings, in support of trains 1 through 3 of the Corpus Christi LNG export terminal and associated gas pipeline in Texas;6
- the US$3 billion issuance of senior secured notes due 2039 to refinance existing debt of Cameron LNG, LLC in connection with the Cameron LNG export facility in Louisiana;
- an aggregate US$2.5 billion notes offering to refinance project debt for the Dakota Access Pipeline between North Dakota to Illinois, consisting of a US$650 million tranche due 2022, a US$1 billion tranche due 2024 and a US$850 million tranche due 2029;
- a US$1.5 billion dual-tranche green bond facility (consisting of 10-year and 30-year notes) to finance a portfolio of MidAmercian Energy's wind projects;
- a €2.326 multi-tranche PPP refinancing of the 62 km Brebemi Toll Road in Italy; and
- the €905 million multi-tranche issuance in connection with the LGV Tours-Bourdeaux high-speed PPP rail project.
In recent years, the general trend has seen pricing on investment grade project bonds grow more competitive with commercial bank or term loan A financings (which are traditionally lower-priced than financings in the term loan B market). Observers attribute this convergence to rising interest rates and to the gradual implementation of the Basel III international banking regulations, which have been implemented to varying extents globally but have broadly tightened banks' minimum capital requirements to protect against cyclical macroeconomic changes and led to minimum leverage ratios that protect against over-borrowing.7 As a result, the ability to lock in a long-term fixed rate coupon on a project bond, even if prevailing rates in the capital markets may have been higher than the bank market until recently, has attracted project sponsors to the bond markets in an environment where banks' cost of borrowing has increased. Moreover, while transaction costs associated with bond issuances, including the payment of ratings agency fees and the potential for negative carry in funding construction projects have not diminished, those transaction costs are less of a deterrent in a Basel III universe. Further, the pool of institutional investors has been growing as of late, introducing another means of making bond pricing more competitive.
Recently, certain sponsors have begun to use bilateral private placements (i.e., a Section 4(a)(2)-governed issuance to a series of funds advised by a single investment manager, rather than a group of unaffiliated investors). This trend has been especially pronounced in the liquefied natural gas (LNG) space, where bank debt is commonly refinanced in increments because of the high construction costs involved, so bilateral private placements represent a relatively convenient way of closing a partial refinancing. Ideally, bilateral private placements are characterised by a simplified negotiation period, because only the investment manager is across the table from the issuer, and the investment manager's funds can join any inter-creditor arrangements that already exist, as discussed in more detail in Section VI.
III COMMON FEATURES OF PROJECT BOND ISSUANCES
Regardless of how any project bond is issued or the pool of investors the issuance attracts, certain key features are common to or strongly associated with any project bond offering, which distinguish project bonds in particular from bank financings. Six commonalities among project bond offerings are described in Section III.
i Fixed-rate issuances
Project bonds typically bear interest at a fixed rate, commonly referred to as the coupon, which is the most obvious factor distinguishing project bonds from bank debt in project finance circles. In an ideal project bond placement, the fixed-rate coupon benefits both project sponsors and bond investors. On one hand, the sponsors lock in a fixed price of debt proceeds, generally over a longer term than banks offer and without susceptibility to fluctuations in interest rates or any step-up in margin that is a common component of longer-term bank financings. The fixed-rate nature of the bonds also obviates the need to enter into interest rate hedges and monitor breakage costs. As mentioned above, project bonds are most commonly issued in the refinancing context; that is, sponsors often avail themselves of bond markets precisely to avoid pricing fluctuations once the markets have determined the sponsors' cash flows are sufficiently stable. On the other hand, the types of institutional investors who are repeat participants in project bond issuances – the insurance companies, pension funds and other asset managers who will be referenced throughout this chapter – prioritise the ability to receive a reliable source of income, exemplified by the bond coupon.
ii Longer tenors and amortisation profiles
From an investor's perspective, the relatively lengthy tenors of project bonds complement their fixed pricing. Project bonds frequently amortise over long periods ranging from 10 to 30 years, and offerings often consist of tranches with different maturities and different coupons (as with the Dakota Access Pipeline, MidAmerican Energy, Brebemi and LGV projects discussed in Section II). By comparison, term loans financed with bank debt will usually have tenors of just a handful of years, and almost always less than 10 years. Insurance companies, pension funds and asset managers all retain long-term liabilities by the nature of their respective businesses, and so the long tenor of fixed-rate debt provides these investors with a predictable investment that offsets these liabilities. Another consequence of the longer terms of project bonds is an amortisation profile that features a smaller balloon payment at maturity than would be found in a shorter-term bank financing of a similar quantum. Bank financings are designed to be refinanced prior to or at the end of their terms, whereas bond financings are often fully amortising. With project bonds, this reduces refinancing pressure on sponsors, and, as indicated above, a sponsor seeking access to capital markets is signalling to investors that it is prepared to accept long-term financing of its assets. The maturity and amortisation profiles of project bonds thus line up neatly with what is considered a key tenet of project finance (though it is not without exceptions): for a project to be financeable, it requires a long-term, predictable source of steady revenues, typically by means of one or more offtake contracts that pay the project for its products or services. To wit, bonds have been a common source of financing in the LNG and certain power sectors, where offtake contracts can have terms of 15, 20 or even 25 years. Bond investors, as the typical maturity profile indicates, are comfortable with a minimal cushion between a bond's maturity date and the scheduled expiry of the relevant project's revenue contracts. In rare cases, bond investors will even accept a merchant tail at the end of a bond's maturity (i.e., a scenario where a project's revenue contract expires before the bonds mature), depending on the projected strength of the market for the project's products or services, the likelihood of extension of the offtake contracts, and the expected life and performance of the project assets.
iii Aversion to prepayments; make-whole
Bond investors are generally averse to prepayment, given their focus on long-term yield as borne out through the amortisation profile discussed above – another reason why project bonds are more frequently used to take out bank debt than vice-versa. As such, an issuer typically must accompany any optional prepayment of the bonds with a make-whole payment of the scheduled remaining payments of principal and interest on the amount prepaid, discounted to present value. Additionally, the list of mandatory prepayment events may be shorter in bond issuances than bank facilities, though prepayments following casualty events, material asset sales and suspended distributions are relatively standard. In many cases, and especially following a material asset sale, investors will require a premium on such mandatory prepayment if not a make-whole. Complicating matters, the procedure for distributing prepayments among investors now often requires, particularly where international purchasers are in the investor pool, that prepayment proceeds first be offered to the investors and then distributed among those who accept the offer, owing to recent changes in the regulatory landscape.
Investors will typically require that the bonds receive a rating from one or more ratings agencies. The ratings agencies also play a key diligence role in the investors' evaluation of the offering: as discussed elsewhere in this chapter, bond investors (and especially investors in Rule 144A offerings) are less involved in monitoring an issuer's compliance with debt terms than banks, and always rely heavily on ratings agencies' assessment of a given project. The number of ratings required will depend on the nature of the placement: ratings from multiple agencies are more likely to be required in Rule 144A offerings, where investors' role in diligence is even further removed than that of investors in Section 4(a)(2) placements, which frequently close with a rating from only one agency.
v Looser, incurrence-based covenant package
As mentioned above, relatively loose covenant packages that allow for more flexibility for the issuer (as compared to bank financings) are common features of bond terms. Broadly speaking, we can attribute this feature to an investor pool in project bond issuances (and particularly in issuances that qualify for the Rule 144A safe harbour) that is likely to be more diffuse and passive, and to place a greater emphasis on timely payment of a coupon as an investment strategy, than a banking syndicate. Though any project finance transaction, be it a bond or a bank financing, will include covenants that require the issuer to maintain the financed project as a going concern or are tailored to unique features of the financed project, the key covenants in project bond documents are more typically 'incurrence'-based, in that they depend on an event's occurrence or a specific action by the issuer. While incurrence-based covenants are of course prevalent in bank financings too, in the context of bond financings, these incurrence-based covenants and compliance therewith depend heavily on qualifiers, objective thresholds (i.e., thresholds that must be exceeded before the issuer is in breach) and rating agency reaffirmations. A more practical reason underlying the looser covenant package in bond financings can be found in the relative investment philosophies of deal teams at institutional investors (which are typically small in number and whose investments stretch across a breadth of unrelated products) and at banks (which may have a small handful of financiers and a greater degree of specialisation in project finance and sub-markets in the project finance space). There is an obvious downside to the looser covenant package and lesser oversight by bond investors: should any waiver or approval become a necessity, there will likely be high transaction costs involved in obtaining that waiver or approval. We will revisit the subject of bond document covenants and the greater administrative difficulties in Rule 144A transactions in Section V.
vi Tendency towards operating projects
A final feature that typifies, but does not exclusively define, project bonds is the sponsors' tendency to seek bond financing for projects that are already operating, as opposed to those that are under construction. For starters, projects that can demonstrate at least a minimal operating history will be immediately more attractive to project bond investors because of the greater confidence in steady, long-term revenues that operating history can help predict. Said another way, commensurate with bond investors' looser covenant packages – operating projects are theoretically easier to monitor than construction projects, for which the ability to achieve a tight development timeline and full performance projects are less certain. From another angle, construction projects bring a risk of negative carry: bonds are typically funded all at once (or occasionally with limited delayed draw features), meaning that the bond coupon will exceed whatever low-yield investments the issuer makes with the bond proceeds until after the project reaches operation and generates sufficient revenues. Deal parties can ameliorate the risk of negative carry by introducing delay-draw mechanics into the bond documents, such that this risk can be spread out more evenly over the construction period, but even a well-oiled delay-draw mechanism is likely to be less flexible than the banks' ability to process construction requisitions and to fund draws on a purely as-needed basis. Construction projects with risks that are easily understood or that are enhanced by completion guarantees or other forms of credit support are likely to increase their appeal to bond investors. In addition, sponsors with expensive construction projects initially funded by commercial bank loans may seek to refinance the initial loans with bonds incrementally in partial refinancings (utilising both private placements and Rule 144A issuances) over the course of construction and early operations, as market appetite allows. This is particularly common in the capital-intensive LNG space, where bank debt is initially funded on terms that include pre-negotiated inter-creditor arrangements and certain permitted debt incurrences that envision future partial refinancings.
IV TYPES OF PLACEMENTS
In the US, project bonds are securities that are subject to the Securities Act of 1933, Exchange Act of 1934 and Trust Indenture Act of 1939, and the rules and regulations promulgated thereunder. Unless an exemption is available, issuers are required to register the project bonds to be sold and traded, and become subject to the same extensive disclosure requirements that govern any public securities sale. Though project bonds can, of course, be traded publicly, when preparing for an issuance project sponsors typically seek to take advantage of either the private placement exemption available under Section 4(a)(2) (for 'transactions by an issuer not involving any public offering') or the safe harbour available under Rule 144A. These two exemptions afford sponsors and issuers the ability to avoid some of the more stringent disclosure requirements of US securities laws (and, importantly, to provide certainty around timing of closing),8 but issuances under these two safe harbours do differ markedly on points of procedure and substance, as do issuers' and investors' roles in analysing the bond terms. This section highlights distinguishing features of Section 4(a)(2) and Rule 144A placements, which will be revisited in the discussions of the investor pool and approaches to covenant administration found in Section V.
i Section 4(a)(2) placements
To meet the requirements of the Section 4(a)(2) private placement exemption, the purchasers of project bonds must be sophisticated investors that can evaluate the risks and merits of investment in the project bonds and bear the economic risks of the investment. There are no limits to the amount of project bonds that can be offered under the Section 4(a)(2) exemption, and Rules 506(b) and (c) of Regulation D provide a safe harbour for qualifying issuers to sell the project bonds to an unlimited number of 'accredited investors' (defined in Rule 501 of Regulation), plus up to 35 non-accredited investors if an issuer is relying on Rule 506(b). The insurance companies, pension funds and other institutional investors that are repeat purchasers of project bonds will meet the accreditation threshold. Though the number of accredited investors is theoretically unlimited in a Section 4(a)(2) private placement, issuers may only engage in general solicitation or marketing of the project bonds in certain circumstances, and satisfaction of the Rule 506(b) or (c) safe harbour also requires issuers to take certain steps to confirm the accredited investor status of investors in the offering. Bonds purchased under the Section 4(a)(2) exemption are 'restricted securities' under US securities laws that may not be resold unless they are registered or the resale qualifies for an exemption therefrom. Section 4(a)(2) investors generally seek to buy the bonds with an eye towards long-term investment and will typically make a representation in the applicable bond documents that they are not purchasing the bonds with a view towards distribution.
As a result, Section 4(a)(2) investors generally have a more involved role in carrying out due diligence of projects that are the subject of the bond issuance and in commenting on the bond documents than Rule 144A investors would. In addition to the Section 4(a)(2) investors' long-term view of the investment and the traditionally smaller pool of investors in a Section 4(a)(2) offering – which manifests in a larger role in compliance monitoring compared to Rule 144A investors, as will be discussed in Section V – the Section 4(a)(2) investors' level of involvement is distinguished from that of Rule 144A investors by the absence of Rule 10b-5 liability (discussed in more detail below) of the investment banks that act as initial purchasers, which is a main driver of the Rule 144A diligence process.
In a Section 4(a)(2) private placement, project bonds are sold directly to the purchasers, rather than purchased by an initial purchaser who then makes a secondary sale to the ultimate investor in a Rule 144A offering. A placement agent will often be retained to market the project bonds to the prospective investor pool and will assist the project sponsor in developing the financial model and other technical reporting, distributing diligence materials and preparing an investor presentation and private placement memorandum to be shared with the prospective investor class. The placement agent will not purchase the project bonds for its own account and so does not act as an underwriter or initial purchaser, which is how the placement agent may avoid liability under Rule 10b-5 of the Exchange Act, which prohibits any person from making an 'untrue statement of material fact,' in the language of the rule, in connection with the purchase or sale of securities. (By contrast, the investment manager negotiating a bilateral private placement on behalf of its affiliated purchasers will not act as an initial underwriter and also avoid Rule 10b-5 liability, but its responsibility to create and distribute diligence materials is significantly limited because of the structure of the bilateral transaction, as discussed in the next paragraph.) Predictably, the diligence materials (including the private placement memorandum) that a placement agent has a hand in preparing will be less fulsome than an offering memorandum drafted for a Rule 144A offering, where 10b-5 liability is a factor.
The trade-off for the placement agent's lack of 10b-5 liability in marketing a Section 4(a)(2) placement is a lengthier period for the investors to conduct their own diligence and to review the draft bond documentation. In addition to the draft bond documents, the investor materials will contain financial projections, technical information, a short-form term sheet for the bond documents and material terms of project documents. For a bilateral private placement, unsurprisingly, investment managers do not need to create their own marketing materials, because there is no broader group to which they need to market the deal beyond their own funds. These bilateral investors will perform their own diligence and look to the reports and summaries created by their outside advisers. In the case of any Section 4(a)(2) placement, deal teams at Section 4(a)(2) investors (including bilateral investors) will typically drill down beyond the elemental information regarding a project that is usually found in Section 4(a)(2) marketing materials, involving their own in-house counsel to review the financing documentation and engage in a prolonged question and answer diligence exercise that involves the sponsor, placement agent, their respective counsel and other independent advisers. In the Section 4(a)(2) context, review and comment periods between indications of interest from investors through submission of bids and pricing to execution of bond documents can extend for upwards of six weeks to two months, depending on the nature of the transaction. Investors and issuers can also expect lengthy diligence periods where bonds are issued by a holding company to finance a portfolio of projects, which may involve different offtakers and different revenue contracts. When project bonds are issued in the renewable space (where tax incentives and credits drive a significant amount of investment), Section 4(a)(2) investors will also need to grapple with any tax equity arrangements that are also used to finance the subject projects, because tax equity investors may have preferred rights in certain cash flows if the bondholders are structurally subordinated.
ii Rule 144A placements
Rule 144A provides for a safe harbour from the Securities Act registration requirement for resales of securities (including project bonds) to 'qualified institutional buyers'. The qualified institutional buyer cannot be a natural person – it must be an institutional investor – and it must be an 'accredited investor' under Rule 501 of Regulation D that, in the aggregate, owns and invests on a discretionary basis at least US$100 million of securities of unaffiliated issuers. Similarly, the pool of Rule 144A investors is theoretically unlimited, though the qualified institutional buyer rule limits the pool from the offering's outset and the rule places certain confines around the number of securities that can be sold to investors affiliated with the issuer.9
Rule 144A only applies to resales of securities, so to take advantage of the Rule 144A safe harbour, an issuer sells the project bonds to one or more financial institutions acting as initial purchasers, who then resell the bonds to the qualified institutional buyers described above, and it is these resales that qualify the transaction for the Rule 144A exemption. The project bonds sold in this manner are also restricted securities, but can be freely traded among qualified institutional buyers in another Rule 144A resale. This established pool of qualified institutional buyers often distinguishes Rule 144A placements from their Section 4(a)(2) counterparts in that the bonds sold are inherently more liquid. (Purchasers in Rule 144A placements can also resell bonds to non-qualified institutional buyers through another available exemption from registration under the Securities Act in what has become known as the 4(1-½) exemption, so long as the resale is sufficiently private and the new purchasers intend to hold the bonds.) This increased liquidity among qualified institutional buyers that is part-and-parcel with Rule 144A resales can result in a pool of investors that does not have a view towards long-term investment and instead takes a mark-to-market approach with the bonds. However, for the most part, investors in Rule 144A project bonds do generally have a view towards long-term investment, and the trading in Rule 144A project bonds is relatively limited (particularly if the project performs in a stable manner over the life of the bonds).
Unlike Section 4(a)(2) placements, the financial institutions acting as initial purchasers will have potential 10b-5 liability in a Rule 144A-eligible transaction, because they act as intermediaries between issuer and the subsequent purchasers in the resale. This results in a significantly more comprehensive offering memorandum, usually hundreds of pages long for project bonds, describing, among other things, the risk factors involved in the issuance itself and the subject project's operations and contracts, and the terms and conditions of the bond documents in greater detail than would be found in a Section 4(a)(2) private placement memorandum. Disclosure materials will generally be accompanied by a comfort letter from the issuer's auditors and a '10b-5 letter' from counsel, which are issued in favour of the initial purchasers; the 10b-5 letter provides a negative assurance that, based on the work undertaken by counsel, counsel has no reason to believe the disclosures contain an untrue statement or omission of material fact. The trade-off calculus in the diligence process is therefore much different in a Rule 144A transaction as compared to a Section 4(a)(2) issuance: in exchange for the more extensive disclosure and consultants' assurances, potential investors are given a much shorter period to review diligence materials, will often not involve in-house counsel and will be expected to make very few, if any, comments to the financing documentation.
V PROJECT BOND INVESTORS: APPROACH AND ADMINISTRATION
Certain types of institutional investors are more likely to be repeat players in the project bond markets. Insurance companies are an obvious example and are more likely to purchase project bonds for long-term investment. Pension funds and other large asset managers also help fill the pool and are more flexible with the increased liquidity features of a Rule 144A resale. Increasingly, infrastructure debt funds have begun investing in project bonds. Depending on the jurisdiction in which the issuer is organised, local investors may be called upon, or even required, to invest in cross-border issuances of project bonds, which may have currency restrictions.
In the project finance market, the investor pool is almost certain to be composed of a mix of the foregoing types of investors, and many of the institutions participate in both Section 4(a)(2) or Rule 144A-eligible transactions. However, the deal teams at those institutions who participate in both of these types of project bond transactions will come from different desks at the applicable institution, depending on whether the transaction is structured to satisfy Section 4(a)(2) or Rule 144A. As mentioned above in Section IV, an investor's Section 4(a)(2) deal team is much more likely to involve internal counsel, and the analysts will be expected to have a greater degree of familiarity with the types of assets that are typically financed in project bond markets, given the longer review period and larger diligence effort required of the investor. An investor's Rule 144A deal team may have a more generalist approach, focused on a wider pool of investments.
When compared to banks that lend in the project finance markets, institutional investors generally take a more passive role in monitoring a project's performance and compliance with the bond documents, a truism that applies especially to purchasers of bonds issued under a Rule 144A placement. Institutional investors, as can be expected given the nature of their businesses, will broadly remain hands-off unless and until their yield is imperiled. This more passive approach provides issuers with a carrot (less desire on the part of the investors to oversee compliance) and a stick (investors' relative inflexibility in changing market-standard covenants and granting waivers when oversight becomes necessary, plus difficulty in managing and keeping track of current holders of the project bonds, especially in the context of a Rule 144A transaction). As intimated above, this carrot-and-stick feature is significantly more prominent in placements under Rule 144A, because the holders of Section 4(a)(2) bonds are likely to be relatively manageable in number and typically also have a deeper knowledge both of the financed project and the assets in general, and so will be more flexible in granting waivers and negotiating defaults.
As is to be expected, at least with respect to Rule 144A placements, covenants in an indenture or other governing bond documents are more likely to be looser and issuer-friendly than those in bank debt transactions.10 As noted above, Rule 144A investors can accept looser covenants than bank lenders or even Section 4(a)(2) investors would, because their ongoing role in project administration is limited, for reasons of practicality among others. One practical reason for the looser covenant package in Rule 144A transactions is the more formal and administratively burdensome process for seeking bondholder consent in those deals. This administrative difficulty is driven both by the greater changeover in bondholders, making those bondholders more difficult to locate and contact, and by the formal procedure for soliciting and granting consents, approvals and waivers that will be formalised in the bond documents unless the investor pool is very small. The process for obtaining approvals and waivers from holders in Section 4(a)(2) placements, though not without its challenges, is closer in process to obtaining approvals or waivers from a bank syndicate. Regardless of the transaction, the procedure for soliciting and obtaining consents is made more efficient by the general incurrence-based nature of the covenant package typically found in project bond offerings, which incentivises an issuer to seek a given consent or approval before they take action that would test a covenant's limits.
In any project bond transaction, and especially Rule 144A-eligible issuances, the deal teams on both the issuer and the purchaser sides are incentivised to create objective (but mutually acceptable) carve-outs from bond covenants, to avoid requirements to seek investor approval for immaterial matters and to consider creating certain agency or representative roles to increase administrative efficiency. The first and most obvious workaround to achieve these incentives focuses on drafting the covenant package: for example, many covenants in project bond documents will be qualified by 'material adverse effect,' the definition of which is often the subject of negotiations. Conceptually, however, 'material adverse effect' varies from project to project and sponsor to sponsor; and, because of its openness to interpretation and dispute, is perceived as being sponsor-favourable in affording issuers wide latitude in covenant compliance. As an alternative (and more objective) means of qualifying covenant compliance, the parties can stipulate that a specific monetary or other quantitative threshold be exceeded before the covenant is triggered. Even for Section 4(a)(2) investors that are well equipped to monitor, and that often require, narrowly tailored and project-specific covenants, these objective carve-outs and thresholds are also likely to be attractive. For instance, where a Section 4(a)(2) placement finances a portfolio of more than a dozen projects, investors can take advantage of the relatively long comment-and-diligence period prior to closing to help determine what projects are critical to operations and ultimately, repayment. This could involve negotiation of the 'material adverse effect' definition to mean a material adverse effect that affects a specific percentage of the project or portfolio being financed, or the negotiation of specific covenants and deliverable requirements that pertain only to individual projects, which in each case would lighten the investors' cost of monitoring compliance.
Unsurprisingly, the project bond covenant package in both Section 4(a)(2) and Rule 144A placements has become comparatively standardised with time, and bond investors will be wary of deviations from what they consider customary language. For example, in connection with a Section 4(a)(2) placement, purchasers' counsel will frequently be asked to prepare a memo that highlights differences in a proposed draft purchase agreement from a model form, such as the one published by the American College of Investment Counsel; in connection with a Rule 144A-eligible sale, the terms and conditions of the financing will be set forth in extensive detail in the offering circular.
As another workaround, because investors in either Section 4(a)(2) or Rule 144A placements of project bonds will almost certainly require that the issuer receive a rating for those bonds, the parties can agree in the documents to deem that investor approval will have been received if the bonds' rating is reaffirmed in connection with a desired action or outcome. Another solution is to establish roles that make administration more closely resemble those found in bank debt transactions. This could involve the appointment of an agent for roles of varying scope but that nonetheless improve efficiency: examples include a notes agent appointed by the issuer who maintains the bondholders' register on the issuer's behalf (thus improving communication inefficiencies) and makes payments on the bonds directly to the bondholders, a monitoring intermediary who assists the bondholders in making procedural decisions (and who may have had experience acting as a monoline insurer, which were commonly used in bond transactions to provide a credit wrap for bond issuances prior to the 2007 to 2008 financial crisis) or, in rare instances, a bondholder representative who is appointed by the bondholders to act on their behalf in granting waivers and amendments.11
VI INTER-CREDITOR ISSUES
Projects financed by a mix of bonds, bank debt or export credit agency (ECA) debt (or both) will require an inter-creditor agreement to govern the relationship among the different debt facilities. The effect of the inter-creditor arrangement on the typical project bond financing depends on the relative sizes of the multiple credit facilities, whether the financing parties want to establish common terms that will govern across all facilities, and where the parties land on voting. In a scenario where bondholders and other lenders agree to establish common terms, the bondholders (especially, again, bondholders in a transaction exempt under Rule 144A) should theoretically benefit from a tighter-than-normal covenant package, because the bank lenders can be expected to require tighter and more bespoke covenants than the bondholders would require independently. This can be a burden on the issuers, especially in a scenario where creditors' approval or waiver is needed but the bondholders are the majority creditor. A commonly seen modification to this approach is to have a set of common terms, but to limit the provisions in respect of which bondholders may vote. This leaves a greater degree of control in the hands of bank lenders, which reduces the burden on issuers seeking approvals or waivers.
Another modified common terms approach is possible, whereby the debt facilities each have separate agreements and no common terms agreement is executed, but the required lenders or holders in one facility can decide to amend a provision in that facility and 'drag along' the lenders or holders in another facility who have a corresponding provision. In this modified approach, the facilities should stipulate certain fundamental provisions that cannot be dragged along. For example, bondholders may be particularly mindful of any changes to certain critical negative covenants (e.g., the incurrence of additional debt, especially where bondholders are a majority senior creditor) and compliance obligations on issuers, and seek to prevent other creditors' ability to make those changes via the drag-along.
The presence of ECAs alongside bonds in a project finance transaction also introduces thorny inter-creditor issues, because ECA lenders may have policy-specific reasons for pursuing a certain course of action. ECA lenders commonly seek to negotiate certain 'golden' votes that allow them to veto the bondholders' or other creditors' decisions.
And, where projects are refinanced incrementally with new bond issuances that gradually replace the initial debt, investors should be aware that they are likely providing debt to a project that has pre-agreed inter-creditor arrangements related to voting and enforcement of remedies, which can be quite complicated depending on the number of debt facilities and issuances sitting with a given borrower, but which new investors may have little leverage to renegotiate. There may be certain matters that are left to individual classes of noteholders, but separately there may be matters for which all bondholders are similarly situated and thus risk getting dragged along by a majority of all bondholders, even if not all of them invested at the same time or on exactly the same terms.
VII ROLE OF CREDIT SUPPORT
Where bondholders are attracted to stable, predictable projects, the presence of credit support can make riskier projects more palatable to the bond markets. This is particularly true in infrastructure bond markets, where the establishment of public-private partnerships or the financial support of a governmental or quasi-governmental development bank or guarantee programme can lend credibility and financial assurance to investment in a project, even during the construction stage. In the US, an example is the US Department of Transportation's credit enhancements available under the Transportation Infrastructure Finance and Innovation Act (TIFIA), which allows project sponsors seeking access to capital markets to benefit from direct loans, loan guarantees and standby lines of credit from the federal government for qualifying projects. In Europe, a similar role is performed by the European Investment Bank, which was projecting to provide €16.8 billion to support infrastructure projects in 2019.12 Bondholders investing in projects that are recipients of credit from governmental or quasi-governmental entities should be aware that these entities may have very little flexibility in negotiating an inter-creditor arrangement.
Credit support is especially important in developing markets and often takes the form of governmental involvement in guaranteeing a portion of the investment or in sponsoring the relevant project itself (generally through a state-run entity). Cross-border projects seeking access to the bond markets can also benefit (or suffer) from the credit rating of their host countries, which will be seen in many cases as a proxy for the creditworthiness of the project itself. Sponsors in less creditworthy jurisdictions will thus need to develop a robust history of constructing and operating reliable projects and seek out strong counterparties, to have a chance of exceeding their host country's own credit rating and attracting cross-border investment.
In sum, in the current economic environment that continues to support issuances of project bonds, not just in refinancing operating projects but also sufficiently financeable construction projects, prospective issuers and institutional investors alike need to avail themselves both of the exemptions to registration that govern project bond issuances and of the expectations for diligence, documentation and administration that accompany them. As an added layer, in larger projects, bondholders will need to develop an understanding of how the bond terms compare to those of other debt facilities, and both bondholders and issuers will need to understand the consequences of (and how they fit within) the creditors' established voting system, which may be established for political as well as economic reasons. Overall, sponsors must be mindful of what features attract institutional investors to project bonds – be they a project's expected future performance that maximises bondholder return, a mutually acceptable covenant package that maximises administrative efficiency in a market not known historically for well-defined administrative roles, or the availability of other credit that a project can obtain to spread risk.
This article is an extract from The Project Finance Law Review, 2nd Edition. Click here for the full guide.