In this article I am going to address a common complaint that we’ve seen ABS investors have: that when they’re putting together systems, too much automation creates a “black box” which then doesn’t permit the user to adjust the data in the manner in which they see fit.

Let’s face it, traders are on the front lines evaluating complex securities such as ABS bonds and the more you can permit users to take the data and create useful models that don’t “lock them into a particular view” of what’s being traded, the better it will be. Most often, traders build their own spreadsheets and, in general, do a great job. However, the lack of ability to dynamically communicate with a database of securities information can cause a great deal of trouble in the ABS market, if only when the next month’s data set comes out from trustees and they find themselves scrambling to manually update their spreadsheets.

Additionally, IT departments blanche at the thought of those overly flexible, manipulable spreadsheets that defy “systemization”. In this article we will discuss a specific example and how to satisfy the needs of both areas: IT and the Trading Desk.

Let’s take up the subject of “Credit IO’s”.

Definition: A Credit IO is an ABS bond which is sufficiently far down in the Capital Structure of an ABS deal that, based on the level of collateral defaults and loss severities that the market is currently experiencing, cause an investor to NOT expect any payment of principal.

Assumption: the bond’s principal WILL be written down to zero at some point. The investor expects NOT to get any principal back. However, until that point, the bond can earn interest cash flows therefore it’s an “Interest Only” bond.

Key Factor: Loss timing. Between now and precisely WHEN the bond is fully written down, the bond will be earning interest. Those monthly cash flows are worth something. The faster the bond will be written down, the less interest cash will be received. The longer the bond exists, the longer the bond will receive cash flows. The trick is to figure out when the losses will hit the bond. The timing of the losses will therefore have a dramatic effect on the price that an investor should be willing to pay for the bond. Less time until the fully-written down point = lower price.

So let’s take a look at some of the elements relating to the data side of this. Here are some of the relevant points:

1. Delinquencies

2. Foreclosure and REO timelines

3. Loss Severities to be used in determining how much of each loan will be lost due to defaults.

4. Credit Enhancements levels – primarily overcollateralization (OC) and each tranche’s current level of credit support (how much of the capital structure is supporting the particular tranche(s) we are evaluating).

On a Bloomberg you can bring up a simplistic method of evaluating this by typing an ABS cusip followed by the Mortgage key (F3) and then typing “MTCS” . This gives you the ability to take the deal’s current level of 60 day and 90 day delinquencies and apply a particular percentage of each that you expect to go through to default. The amounts of loans in Foreclosure (FC) and Real Estate Owned (REO) are assumed to be 100% in default. So we have as an example:

Table % % that will default default amt

% of Deal 60+ Day Delinq 8% 60% 4.8%

% of Deal 90+ Day Delinq 5% 70% 3.5%

% of Deal in FC 3.5% 100% 3.5%

% of Deal in REO 2.5% 100% 2.5%

For a total of 14.3% that we expect to end up in full default and thereby experience a loss.

Sum those figures up (14.3%) and multiply by a single loss severity input and you will have the approx amount of the deal that you will experience as a loss. Let’s say we use 50% Loss Severity. That will give us 7.15% of the outstanding collateral balance in the deal that we expect to impact the deal’s capital structure in the form of losses. Compare that amount versus the particular bond’s credit support that you’re evaluating and if you have a ratio (called the “Coverage Ratio” on Bloomberg), that is less than 1.00, then that bond is likely to disappear completely because there is simply not enough support for the bond to survive. Anyone with access to a Bloomberg can do the above. The above doesn’t actually try to predict WHEN the losses will occur – only that they are expected to occur at some point in the future. It also does not let you consider future loans that are current on their mortgage payments or are 30 days delinquent that will come down the “pipeline” into the more severely delinquent states and finally into realized losses. It also doesn’t try to tell you what it all means in terms of a “price” that you might be willing to pay for the bond.

So let’s kick this up a notch.

Loan-Level Delinquency information

First of all, let’s assume that we have access to loan-level information and that we know, not only the current delinquency status of each loan but exactly when the loan entered that status. Intex provides good loan level data for deals from about 2006 and onwards. Loan Performance provides loan-level information for all deals – loan level information is generally what Loan Performance is known for (but they don’t have very good data about the capital structures nor can they do really good cash flows on the bonds as Intex does). The point is that loan-level delinquency information is available.

So let’s retrieve all the loans from a particular deal into a spreadsheet from our database of loan-level information. Ideally, this should be automated from within the spreadsheet so we can always refresh the data whenever we need to ensure that it is representative of the most current data in our database.

We now have our hands on which loans are in which delinquency condition. Now, if we simplistically project out maximum timelines that all the loans will experience in FC and REO before they hit their loss point, we can derive a table of months going forward and WHEN those losses will be experienced.

For example, we can state the following:

A. Let’s say that a loan has been in FC for two months already: Let’s permit 6 months for the total “normal” amount of time that a loan is going to be in FC so that there are expected to be 4 months more of FC time for this particular loan. Then permit 6 months more for the full REO process. This means that month 10 is WHEN we expect the loss to hit.

B. Let’s say that a loan is currently in REO and has been so for 4 months. Permitting 6 months of complete REO time suggests that we have 2 more months to go. So 2 months from now is when we think we will realize a loss on this loan.

C. Let’s say that a loan has just become 90 days delinquent for the first time. They’re probably going to be in FC real soon, but maybe we feel that we should allow an additional month of being 90 days delinq. So we would have 1 more month of 90 days delinquency. A full 6 months of FC and 6 months of REO so that we expect the loss to hit in month 13.

We can continue to do the above for 60 days delinq loans and 30 day delinq loans. And possibly take some current loans based on the idea that some of these will also hit the skids.

Let’s assume an overall “Loss Severity” of 60%. According to some market participants 60% is getting more and more real. This means that, given a loan amount of $100,000 you are expecting to lose $60,000. Apply the loss severity input to each of the loan balances and sum those loss amounts up into each of the months you have projected into the future.

The result is that you end up with a table of months into the future within which losses can be summed up – month by month. At that point we have a relatively simplistic table giving us WHEN we expect the losses to hit. These losses will be applied to the bond’s outstanding balance and will eventually “amortize” the bond’s principal, via write-downs, down to zero. At each month, you calculate what amount of interest the bond should receive. Then we apply the loss amount for that month and decrease the bond’s outstanding principal balance so that in the next month there will, of course, be less interest earned. We keep doing this until the bond’s balance has been written down to zero, at which point you’re not earning any more interest on the bond. At that point, the bond has disappeared. Then sum up the interest payments that you received during the time when the bond was still “alive” and you have the amount of cash you’re going to receive on this bond. Divide that by the principal currently outstanding on the bond and you have the price that might be indicative of what you would be willing to pay. Notice that this last sentence is disregarding the time value of money. It can be an enhancement to “present value” (PV) those interest cash flows and then sum up the PV-ed cash flows to get a more accurate price.

It should be noted that if there is any “OC” remaining at the bottom of the capital structure in the deal, you have to allocate the loss amounts to the OC first before they start to impact the bond you’re evaluating. Likewise if there are any bonds BELOW the one you’re evaluating, because of the fact that losses are allocated from the bottom of the capital structure upwards, then each of those bonds below your bond each have to be written down to zero before the loss amounts start to impact your particular bond. The point being that your spreadsheet application must retrieve all of the bonds and any OC BELOW your bond and apply the loss amounts to EACH of their principal outstanding amounts BEFORE the losses start to impact your particular bond. Of course, this means that ALL of the bonds below the one you’re evaluating are also, each one, a “Credit IO” bond.

A few other observations

I want to emphasize that decreasing the FC and REO timelines in the model will have the impact of decreasing the amount of time that the bonds will survive thereby decreasing the length of time that the bonds will earn interest resulting in a lower price that one would be willing to pay for the bond. Obviously, if you’re buying you want to pay as low as possible so underestimating time lines will help you. If you’re selling, you’ll probably want to consider that the time lines are longer so that you can sell it for a higher price. These are the normal competitive sort of interests in the market place.

The above represents a simplistic model but one which gives a much greater degree of flexibility than the Bloomberg MTCS function. Done correctly, it also permits the user to adjust the time lines and severities to ones which they feel comfortable with when evaluating “Credit IO’s”.

Also, by keeping all of the above factors in mind, the user/trader can still perform the analysis in the way that they see fits best for the environment they’re in. They’re not “locked” into a “black box” which they can’t see inside of. There are, of course, much more extensive features that can be built into such a model which are not within the scope of this article.