Just as one robin (or, depending on your geographic location, swallow) doesn’t make a spring, two project finance CLOs (PF CLOs)—Bayfront Infrastructure and RIN—may not confidently mark the re-emergence of PF CLOs. They are, nevertheless, a positive and potentially significant development, writes Paul Forrester, Chicago-based partner at Mayer Brown.

Collateralized debt obligations (CDOs) are an application of securitization technology that is able to take low-rated portfolios, with double-B or single-B ratings, for example, and, by tranching and internal subordination, create an optimized funding that includes a substantial portion of highly-rated, e.g., triple-A or triple-B, securities. By including an experienced asset manager with appropriate incentive compensation and permitting trading, an adept manager of a CDO can monitor and manage the portfolio to minimize loss and obtain possible gain. In other applications (for example, trust-preferred CDOs), CDOs have also provided effective access to capital markets pricing and other efficiencies to borrowers who could not otherwise avail themselves thereof.

Although the ability to include construction funding and related risks in a CDO is usually limited by both rating agency and investor concerns, the ability to “take out” interim bank or other construction financing with term funding with capital markets pricing and other terms will offer attractive additional funding flexibility to smaller borrowers or projects. It will also relieve the asset/liability mismatch for most bank lenders in providing long-term, fixed-rate financing, reflected in part in Basel III’s net stable funding ratio.

However, the recent PF CLOs differ from prior PF CDOs in one potentially significant way—they appear to be primarily driven by asset managers rather than bank sponsors and, apparently, are motivated by more traditional absolute or total return investment strategies rather than portfolio or other balance sheet management by sponsoring banks and other financial institutions.

If so, this is a welcome market development, signaling a new direction and substantial opportunity for PF CLOs, since it would move them closer to the highly successful broadly syndicated loan (BSL) CLO market. The latter has over $500 billion in aggregate outstanding and represents more than one-half of the buyers in the over $1 trillion U.S. BSL market.

BSL CLOs have brought significant changes to the U.S. BSL market. It is often observed that the “CLO bid” sets BSL pricing and that BSL CLOs bring substantial liquidity, transparency and discipline to the BSL market. Since BSL CLOs are not typically required to mark the related credit exposure to market and, as practical matter, cannot easily hedge that credit exposure (and are, as a result, “long” that credit exposure), they are generally seen as attractive “buy and hold” investors and not “hot money,” like some other BSL market participants, such as loan mutual funds.




Bayfront Infrastructure

Date of Offering

October 2017

July 2018


RREEF America, a wholly-owned subsidiary of Deutsche Asset Management

Clifford Capital

Offered Classes/


$273.3 million Class A/AAA(sf);

$51.8 million Class B/Aa3 (sf);

$57.4 million Class C/Baa3 (sf)

$67.5 million Preferred Shares/Not rated

$320 million Class A/AAA(sf);

$72.6 million Class B/Aa3 (sf);

$19 million Class C/Baa3 (sf)

$45.8 million Subordinated Notes/Not rated



$431.3 million comprising both project finance loans and broadly syndicated loans

$458 million comprising, initially, 32 loans to 30 different projects



At least 60% must be project finance loans and eligible investments, with various concentration limitations, including up to 45% of such loans in electricity (thermal) contracted and merchant sectors together (with up to 25% for electricity (thermal) merchant sector), up to 30% in large infrastructure and up to 15% in regulated assets/utilities sectors

Projects are in eight different industry sectors and are also diversified by being located in 16 different countries (although the primary project obligors are based in 22 countries) and approximately 38.2% of the portfolio is supported by multilateral financial institutions and export credit agencies

If nothing else, these two recent PF CLO transactions indicate that there is significant investor interest in PF CLOs and that these investors may be pursuing a more specialized investment strategy that is not available in typical BSL CLOs. This is not surprising, as pre-financial crisis PF CDOs generally performed well. Indeed, a report from Moody’s Investor Service in March, regarding default and recovery rates for a dataset of over 7,000 project finance loans during the period 1983-2016, confirmed that they potentially offer exposures to credits with lower default and better recovery rates than comparably rated corporate loans. Promisingly, the Moody’s research also showed that a more recent subset—namely, 1995-2016—had a slightly lower cumulative 10-year default rate (6.4% versus 6.7%), but essentially the same average recovery rate (79.3%), rates that are generally consistent with single A-rated corporate exposures seven years after closing/issuance. Project finance ratings are usually constrained to below investment grade, and especially so for smaller projects and less creditworthy sponsors.

Like pre-crisis, bank-sponsored PF CDOs the two recent PF CLOs (see table) offer tranched risk/reward choices, potential portfolio diversification, experienced and incentivized investment management and other benefits to investors.


A CDO is a securitization structure that repackages the credit risk associated with an underlying portfolio of bonds or loans. CDOs with loans as the primary type of underlying collateral are usually known as collateralized loan obligations (CLOs). The CDO structuring process is time-tested and conceptually sound. The notion that the risk of an asset portfolio can be allocated across different securities based on their depth of subordination is, after all, a basic tenet of corporate finance.

The core concept of CDOs is that a pool of defined debt instruments will perform in a predictable manner (i.e., with default rates, loss severity/recovery amounts, and recovery periods that can be forecast reliably). With appropriate levels of credit enhancement, CDOs can be financed in a cost-efficient manner that reveals and captures the “arbitrage” between the interest and yield return received on the CDO’s assets and the interest and yield expense of the CDO securities issued to finance them.

Typically, CDOs require the CDO assets to meet certain eligibility criteria—including diversity, weighted average rating, weighted average maturity, and weighted average spread/coupon—in accordance with established rating agency methodologies. This ensures the highest practicable rating for the related CDO securities. A CDO allocates the interest and principal proceeds of such assets on periodic distribution dates according to certain collateral quality tests—typically an overcollateralization ratio and an interest-coverage ratio.

CDO securities are usually issued in several tranches. Each tranche, other than the most junior tranche, has a seniority or priority over one or more of the others. Subordination of junior tranches constitutes the required credit enhancement for the more senior tranches, which receive a credit rating that reflects such seniority or priority. Some CDOs use financial guaranties or insurance for the same effect.

The underlying CDO assets also affect the capital structure of the CDO. For example, if the underlying debt obligations are floating-rate, the CDO securities should also be floating-rate or must be swapped to avoid or minimize the interest-rate mismatch. If the underlying CDO assets require additional advances, as is the case with construction or post-completion working-capital facilities, the CDO securities should allow subsequent borrowings so the CDO can make the required advances. For this reason, CDO securities are often held by commercial paper conduits that offer attractive pricing and flexible funding. The conduit, however, will likely require a minimum rating to hold CDO securities, and the CDO will require a minimum rating of the conduit (which if “lost” effectively requires the conduit to find a replacement or to post collateral to cover the obligation to make borrowings). Obviously, these complex structural mechanics can be avoided if the CDO holds only fully-funded debt obligations.


PF CDOs are particularly attractive as a means by which project lenders can refinance their PF portfolios. Because PF CDOs provide investors with access to a diversified portfolio of PF debt obligations, project lenders may find it easier to refinance by selling to such a CDO than if each specific credit exposure had to be individually sold or refinanced. As such, the sale of a PF portfolio expands the seller’s opportunity to undertake additional PF business with favored existing clients or new borrowers. In addition, PF CDOs enable financial institutions to better manage their exposures to particular countries, industries, technologies and/or credits, enabling banks to achieve better economic results than they could if they sold the loans in secondary transactions. In addition, the sale of the PF portfolio will usually release excess regulatory capital, since under current Basel requirements, PF loans are usually perceived as having an excess of related economic capital requirements.

From an investor’s perspective, a PF CDO provides diversification and other portfolio management benefits, including a relatively low correlation to typical corporate bond portfolios held by most institutional investors. The tranched structure of CDOs, moreover, enables investors to determine their preferred risk/return.

PF loans, leases, and other debt obligations are viewed as attractive assets for CDOs because they have higher assumed recovery rates and shorter recovery periods than comparably rated corporate debt obligations. This allows PF CDO securities to be issued at a correspondingly lower cost (because less credit enhancement is required to obtain the same credit ratings), which effectively “expands” the arbitrage opportunity.

The higher assumed recovery rates and shorter recovery periods of PF debt are primarily attributable to the tighter covenants and events of default under typical PF documentation. These assumptions are intuitively reasonable and, most importantly, the rating agencies concur with them, even though there appears to be no great weight of authoritative research to support them.


Most of the significant 2007 and 2008 losses occurred on structured credit products (including asset-backed CDOs) with material exposures to subprime mortgages or related mortgage-backed securities. Yet the entire CDO and CLO markets have suffered from “guilt by association.” New issuance of both CDOs and CLOs plummeted in 2008 as investors fled the CDO market and widening credit spreads made the opportunity for yield arbitrage impossible. However, while asset-backed CDO issuance remains muted, BSL CLOs have successfully re-emerged and are setting new issuance records. U.S. BSL CLO issuance so far this year has already exceeded total 2017 issuance, which was the highest level of BSL CLO activity since the crisis other than 2014.

As the RIN and Bayfront PF CLOs demonstrate, there is evident investor interest in the specific PF CDO sector and both RREEF America and Clifford Capital are reportedly discussing further PF CLO offerings. Time will tell if the robin (or swallow) foretells spring, but the recent signs for PF CLOs are promising.

The content you are trying to view is restricted for Power Finance & Risk subscribers.

To continue reading, please log in using the login box in the upper right corner of this page, subscribe or take a free trial.


Set up your account today for full access to Power Finance & Risk.

Join our readership!


Free Trial

Want unlimited access, but don't feel quite ready to subscribe?

Start your free trial today!

Free Trial