Sorry!! The article you are trying to read is not available now.
Thank you very much;
you're only a step away from
downloading your reports.

Pimco: Viewing Covered Bonds Through a Structured Finance Lens


Demand for covered bonds has increased, however they exhibit several risks, including those related to maturity extension, liquidity, and bankruptcy procedures.

Covered bonds, which are senior asset-backed funding instruments with a long history in Europe, are expanding into new territory. More and more non-European issuers are looking into this product, and issuances are on track to debut in regions such as Singapore, Panama, Mexico and Morocco. Demand for covered bonds has also increased among certain global investors due to regulation such as the Basel III Liquidity Coverage Ratio for banks, which favors covered bonds as high quality, liquid instruments. For insurance companies, Solvency II assigns covered bonds preferential status in terms of capital charges. Furthermore, a June 2013 EU directive permanently excluded registered covered bonds from bail-in (i.e., covered bond holders would not suffer losses from a forced write-off of a bank's debt by regulators). The final implementation of the directive is foreseen for 2018.

Among some investors, the long history, favorable regulatory environment and commoditized market environment have created a perception that covered bonds are simple products. However, it's important for investors to know that covered bonds in fact exhibit several risks, including those related to maturity extension, liquidity and bankruptcy procedures, that may be unique compared with those of other credit-related securities. Investing in covered bonds requires more in-depth analysis because of these risks as well as the dynamics of the underlying cover assets.

Pimco has pioneered an approach to analyzing and modeling covered bonds using structured finance techniques. We have created a covered bond cash flow model whose primary aim is not to price the bonds, but rather to establish under what set of scenarios a covered bond could be expected to make all its interest and principal payments. This gives us some foresight into the possible future performance path of a bond, which helps us identify relative value opportunities in the covered bond market.

Covered Bond Basics

The first covered bond was recorded more than 200 years ago in Germany. At the end of 2012, the covered bond market totaled €2.8 trillion, and it is the second-largest private debt market in Europe after bank senior unsecured and before corporate bonds, according to the European Central Bank. The covered bond spectrum is well diversified, comprising 306 issuers and 465 different pools of assets backing the bonds as of Q1 2013, according to HSBC.

Covered bonds are senior, on-balance-sheet, asset-backed funding instruments. They are backed by a dynamic pool of assets known as the cover pool. Most outstanding covered bonds (79%) are backed by mortgages, while 18% are backed by public sector loans. Mixed pools or shipping loans back the majority of the remainder, but the market also saw the launch of aircraft covered bonds in 2012 and the first covered bonds backed by euro-denominated debt of small to midsize enterprises (SMEs) in early 2013.

Covered bonds are issued by regulated financial institutions only and are non-tranched (i.e., individual issuances are not divided into tranches of securities). As long as the issuer is a going concern, the payments due on the covered bonds are paid out of the issuer's operational income rather than from the cover pool. Nevertheless, covered bonds are structured to withstand an event of default of the issuer. Thus, covered bonds do not automatically default when the issuer defaults.

Upon an issuer default the assets in the cover pool are pulled out of the issuer's bankruptcy estate and continue to be operated by a trustee or administrator. The preferential claim from an investor's perspective is on the cover assets, first and foremost. If the cover assets fall short, the covered bondholders have a claim on the bankruptcy estate for the deficiency, ranking pari passu with senior unsecured creditors (referred to as dual recourse). However, it is unclear how this would play out in practice since the deficiency may not become evident for many years, possibly long after the bankruptcy estate has been distributed.

No covered bond has seen a default since the market's inception, but some issuers of covered bonds have failed in their programs. In those cases, the programs are in wind-down mode but continue to repay all outstanding covered bonds accordingly. Because there has never been an event of default on a covered bond, however, no covered bond structure or legal framework has ever been truly tested.

How Pimco Values Covered Bonds

Pimco's approach to covered bond analysis and modeling uses structured finance techniques. A closer look into covered bond structures reveals that most programs have much in common with residential mortgage-backed securities (RMBS). The most significant similarity for our purposes is that after a covered bond issuer defaults, the cover pool ceases to be dynamic. The pool becomes static, holding a fixed set of mortgages like the pool backing an RMBS issuance. From then on the outstanding liabilities of a covered bond are repaid from the cash flows the pool generates. However, there is a significant difference between covered bonds and RMBS in this respect. The covered bonds, being designed to provide financing for a dynamic pool that persists indefinitely, are non-amortizing. RMBS, however, are generally backed by a static pool of mortgages that typically amortize as the pool amortizes. The non-amortizing nature of the covered bonds can create a mismatch between the pool of assets and the outstanding liability structure of a covered bond program, known as refinancing risk.

To account for this difference, we take into consideration: overcollateralization (i.e., total assets versus total liabilities), any maturity extension and swap arrangements, plus any other structural features related to cash flow. Furthermore, the model allows for bridge loan funding for upcoming redemptions. We do not take recourse to the issuer into consideration, due to our concerns about the tangibility of the recourse. Once the structure, pool, liabilities and internal instruments are established, we make projections for the behavior of the pool assets, in particular the prepayment rate, default rate and loss severity.

Once the pool of assets no longer covers the outstanding liabilities or a bond payment is missed, it will trigger a default on the covered bonds. At this stage, we simulate the sale of the cover pool assets. Clearly an estimate of the fire sale price for the pool assets is required, which we derive from historical house price information and Pimco house price expectations based on our secular and cyclical macroeconomic outlooks. In addition, we assume a discounted sale price for the assets due to the distressed nature of their sale.

Having established all of the above, we can model the cash flows of a covered bond. We examine how the liability structure of the covered bond program performs in different macroeconomic scenarios. By running a large number of scenarios, it is possible to assess the robustness (or otherwise) of a bond – that is, when a write-down of principal is possible.

The result of our cash flow modeling and in-depth collateral analysis provides visibility as to the possible future performance of a bond. It allows us to find relative value opportunities within the entire covered bond spectrum and the most attractive value bonds in the program's liability structure, and at the same time identify additional investor protections.

Looking Ahead

The valuation comparison of covered bonds to other related credits is a bigger challenge than what some investors may normally have assumed. It requires investors to conduct deeper analysis of the cover pool and have a stronger understanding of the underlying risks.

Pimco is currently developing an approach that combines our capital structure model, which provides estimates of the probability of issuer default, with the covered bond model. Taken together, the likelihood of an issuer default and the consequences of such for a covered bond will allow us to make pricing estimates and deepen our relative value approach in the covered bond spectrum.

This article originally appeared on Pimco.
< Previous
  • 1
Next >
No positions in stocks mentioned.
Featured Videos