Rise Of Re-Remics

Rise of the Re-Remics

Over the past few several years we have all seen the impact of the economic crisis ostensibly cause by bad loan underwriting standards, rating agency goofs, greed and the rest of it. This has resulted in many, many RMBS bonds formerly rated AAA being downgraded to CCC or worse. What to do with all the downgraded bonds? I know, let’s put them into a new securitization and then tranche the good loans and the bad loans, get the “good tranche” newly rated AAA and so forth. Read on to get a bit more about this process.

You can get a decent basic education on re-remics at this link (it includes an excellent graphic illustrating the basic concepts of why Re-Remics are done).

http://www.theatlantic.com/business/archive/2009/10/re-remics-really/27639/

and a related link to an article from the WSJ here (which does require a subscription to read the entire article):

http://online.wsj.com/article/SB125434502953253695.html

At first glance, these securities look quite a bit like the old “CDO” structure in which other securities were packaged up as assets in a new deal and then tranched by issuing new bonds that were backed by the cash flows on those underlying securities. Of course a CDO had a “collateral manager” whose job it was to monitor the assets and that person also had a “reinvestment period” of several years during which the assets could be traded in and out of to maximize profits etc.

However, Re-Remics are NOT CDO’s – even though they smell like them quite a bit (take that comment any way you want).

These are technically called “ Resecuritization of Real Estate Mortgage Investment Conduits”. In this instance a “REMIC” can be thought of as a plain old RMBS security. And “Resecuritizing” those RMBS securities means the process of packaging them up into a NEW REMIC, ie. A “Re-REMIC”.

From the links earlier you can see that the primary reason for doing a resecuritization is to separate the bad loans from the good loans that are “locked up” inside a formerly AAA RMBS security. In their original form, and because these bonds are backed by a pool of assets which have deteriorated considerably in most cases, the entire RMBS bond is priced very low. However, if we can separate the good from the bad, then we create a new bond backed by the “good” loans and a lower, more risky tranche backed by the “bad loans”. Sort of a mini “good bank / bad bank” sort of situation. The Re-Remic tranche backed by the good loans all of a sudden gets a much higher price; is rated by the rating agencies as AAA (yup, that’s right); and the original investor now does not have to set aside nearly as much regulatory capital as previously. Magic!! What a concept! Of course, the Re-remic tranche backed by the bad loans is priced even lower and is usually not rated by a rating agency. So the amount of regulatory capital set aside for this piece can be quite a bit higher than before. But hey, if we stuff 70% of the new Re-Remic into the AAA and only 30% in to the unrated piece, voila! Net effect is that less money is having to be locked up in set aside regulatory capital and that money becomes freed up to use more productively elsewhere. Financial alchemy, indeed!

A few years, back, much was made of various financial institution’s inability to properly price those complex beasts known as CDO’s. Well the same thing goes for Re-Remics. In any single one of these Re-Remics, we may have up to 50 or more formerly AAA assets, now rated as low as CCC or worse. What is the proper way to value these. If you have a portfolio of these Re-Remic tranches, how do you go about valuing them. The answer is: It’s not simple – just like valuing CDO’s was fraught with complexity – so are these Re-Remics (although we think less so than CDOs).

If you use a predictive modeling company to generate forecast prepayment, default, severity and delinquency vectors (or if you produce these yourself), then you should ideally want to be able to apply these vectors to each and every asset that backs the particular Re-Remic tranche that you’re trying to price. If you are like most firms, you probably produce a wide range of “vector sets” for each economic scenario that you’d like to stress your bonds at. Some firms produce eight economic vectors sets (scenarios), some produce 10, some produce 60 or 100’s (using Monte Carlo Simulation). Once produced, that’s a LOT of vectors and manually loading them into Intex Desktop can be a painful, mistake-prone process and then the process of running the price/yield and cash flow tables can be time consuming, and can also eat up all the processing power on your computer.

You should be able to price these securities easily. Can you do it? Do you have clients that have a large portfolio (or even a fairly small one)? I’d be interested to hear how people are handling the pricing of Re-Remics currently. Are you doing “just enough” to get the job done? What are the stumbling blocks you’re running into? How often do you have to price them? Is that an adequate amount of frequency of pricing and so forth. If you want to take the conversation off-line from this blog, please email me at Jack@thetica.com and I will respond.

In addition to that, we probably all can remember the concept of a CDO “Squared”. This was where a CDO started to invest in the tranches of other CDO deals. Thereby “telescoping” the risk outwards even further. To analyze the underlying bonds required plumbing the depths layers downwards – and this become a particularly difficult task to adequately stress test these bonds.

This past week I have actually heard of a new security called a “Re-Re-Remic”… Similar to a CDO Squared, this is a situation where the Re-Remic tranches have been put into yet another Re-Remic structure and those assets have been “tranched” in order to separate the wheat from the chaff once more. Impossible you state. Nope. It’s for real. Will you be able to prices these securities. I hesitate to personally call these securities “Re-Remic Squared” because of all the horrible connotations associated with the term “CDO Squared”, but that’s essentially what they are.

Is this just another Wall Street concoction guaranteed to give us waves of further financial nastiness down the road? Well, whether they are or not, my question still stands. How are you planning on evaluating and/or pricing these bonds. Depending on your position in the market (trader, risk manager, financial controller, mortgage research analyst, IT professional, analytics provider, and so forth), this may be heading your way or you may already be looking at it square on.

I would love to hear your comments on the overall subject of Re-Remics and also the new-fangled “re-re-remics”.

Best,
Jack S. Broad
Co-Founder
Thetica Systems, LLC
Jack@thetica.com