In my first article, Asset Securitization Perspectives Part I of “Until I’m Done”, I offered some musings about asset securitization in general and, I hope, provided some insight to those who haven’t been directly exposed to the process. In that article, I mentioned a couple of things that I am going to harken back to now. The first was the observation that CU Revest had submitted a comment letter on the NCUA’s proposed securitization rule and the second was my promise to discuss bond ratings, rating agencies methodologies, and when ratings were necessary in a later article. Well, welcome to the later article. So now you are asking, what does the comment letter have to do with bond ratings? I’m glad you asked.
In addition to submitting a letter to the NCUA, I read the other letters that were submitted and noticed a few themes. One of these themes intended to allow for CU’s to increase the size of their securitization offerings. The comment took two primary forms, the first was “let CU’s securitize assets that they purchase in addition to assets they originate” and the second was “allow CUs to pool assets with other credit unions to increase deal size”. I have thoughts about both of those and will discuss the latter in detail a later article when I discuss tranching. The reason I mention the two themes now, however is because each, while relating specifically to transaction size, does so toward the end of mitigating bond cost of issuance.
There are a number of things that contribute to cost of issuance related to asset backed securities (ABS), including rating agency fees, but what type of transaction monumentally affects cost of issuance and whether ratings are even necessary. I know I’m talking in circles a little (it’s a character flaw and I’m working on it) but hang with me and I promise there’s a point and we’ll get to it.
First, let’s quickly look at the primary things that contribute to the cost of issuance of asset backed securities for a first time issuer. The first time issuer caveat is a big one since there is a lot of work that has to be done on the first transaction which can be recycled for subsequent deals.
The first item is underwriting fees. Depending on what type of issuance is completed, this may or may not be a misnomer since the name implies that the investment banker will underwrite the offering (essentially buy the bonds and closing and then sell them). In some cases that is precisely what happens. In other cases the bonds are offered on a best-efforts basis meaning that the investment bank will buy the bonds as long as they have identified who is buying them and have standing orders in place (this role is generally referred to as the initial purchaser and as discussed below, depends on the offering type). The important thing is that underwriting fees are essentially the structuring fees you pay to the investment bankers. They are generally a percentage of the issuance ranging from 0.25% to 2.00% (or even higher as discussed below) depending, again, on the type and size of issuance.
The second component of cost of issuance is legal fees. I have seen legal fees on securitizations run between $50,000 and $400,000 and perhaps even higher. These again vary widely depending on the type of issuance. Next are rating agency fees which run from $75,000 to $125,000 depending on a number of factors. Ratings may or may not be necessary, but we’ll get to that. Then there are a bunch of smaller fees which can aggregate to $50,000 to $100,000. These include things like back-up servicer fees, owner trustee fees, accounting fees, and the like. These fees are to some of the more esoteric participants in a securitization and, while they are important, the specific roles each plays and why that role is necessary would certainly qualify as minutia.
I keep saying, it depends on what type of issuance is done, so let’s (finally) talk about the different types. The first and by far the most common by issuance volume is issuance of public securities. These are issued off a shelf, require full Reg AB compliance, and are the most liquid asset backed securities. As you may have guessed they are also the most expensive. Ongoing compliance aside (and that a big aside) doing a public issuance of securities requires far and away the most in upfront costs. Setting up the shelf runs at least $100,000 in legal on top of all of the other legal fees. In order to increase liquidity, which is the whole point of public securities (higher liquidity equates to lower coupons so cheaper borrowing), you are going to need at least two ratings, so double the ratings range above. Once an issuer establishes a track record, an investment bank may be willing to actually underwrite these bonds due to the liquidity. Long story short, if your credit union is in the position to offer a few billion a year in ABS, these fees and expenses may be worth it due to cost savings over time. Since realistically, that number of credit unions is very small, I’ll move on the next category.
Private securitizations are structured similarly to public deals in most ways. They use SPEs, have structural credit enhancement, and isolate the assets from the seller. The primary difference is the way the securities are marketed and placed. In a “true private” securitization, the investor may be intimately involved with the structure, credit enhancement, tranching, etc. In these offerings, they are essentially negotiated with the investor and may not need a rating based on the investor’s criteria. However, since most investors are regulated in some way or another, they often require a rating so as to comply with capital ratios and the like. I have seen true private securitizations completed for as little as $15 million. While this could be an option for CUs, usually the fixed upfront costs are not justified unless the assets are relatively high yielding to cover both the upfront expense and the higher interest rate of a “true private” transaction.
Now, at long last, the issuance type that many credit unions will care about, I’ll call them semi-private securitizations. Again, the legal structure is the same, but these securities are not offered to the public nor are they negotiated directly with the investor. They are rated by at least one agency and credit enhancement is dictated by that agency’s rating criteria (another future article). The securities are eligible for purchased only by qualified, institutional investors (insurance companies, pension funds, etc.) and the liquidity and coupon both fall somewhere between public transactions and true privates. As these transactions can be economically completed at $50 million in issuance, they are very much in line with the asset capacity of the larger tier credit unions. Since this product is relevant to the CU movement, I’ll discuss the costs and rating agency requirements in a little more detail.
First off, let’s assume we are talking about securitizing auto loans. Not only do CUs have a bunch of them, but also they are very well established in the securitization market. Assuming a $50 million dollar transaction, upfront fees on the first deal would run between 1.25% and 1.60% with ongoing fees less than 15 bpts. Of these fees, rating agency fees would run from $75,000 to $125,000. Now, finally, at long last, a little about rating agencies.
Rating agency methodology (they call them criteria and you can download them from their websites if you have an account) is asset class specific. While the methodology differs some, the primary purpose is to identify the historic static pool cumulative net losses for a portfolio. This number is then multiplied to dictate credit enhancement. For example, and we are talking in generalities, if a CU’s historic losses on its auto portfolio are 2% on a static basis, any bonds rated triple-A would have to pay out with interest at a 12% net loss rate on the portfolio, double-A bonds, at a 8% loss rate, etc. While these numbers can and do vary based on the amount of historic data available, the consistency of that data, any changes to underwriting during the period, and anything else the rating analyst decides to spring on you, that is basically how it works. The rating analyst generally does an onsite visit for a day or two and reviews policies and procedures, disaster recovery, and internal controls in addition to the historic analytics.
That about covers it from a 30,000 foot view. These articles are going to continue until I think I’ve provided a decent overview, but with each one I realize there is more and more to cover. In the interim, if you’d like to explore if securitization could be a good option for your credit union, if you have specific questions, or would like me to get to a specific aspect sooner rather than later, please feel free to send me an email at jmicheletto@CURevest.com.
Jason Micheletto serves as Executive Vice President of CU Revest as well as Servatus, and has 15 years’ experience in the capital markets, including structured finance, asset-backed securities, and lending. He also serves as President of DPM, the consumer finance company affiliate which originates refinance loans to eligible members to aid in their recovery. He began his corporate finance career at RBC Dain Rauscher as an analyst and subsequently served as Vice President at BMO Capital Markets and PBL Capital. While Mr. Micheletto has worked with several different asset classes, he has specialized in distressed, non-performing, and re-performing debt. Since joining Servatus in 2009, he has focused on cash flow modeling and projections, portfolio analysis and pricing, and oversees the trading desk related to portfolio resale activities.