An Insider’s View of an Actual RMBS Securitization at Mandelman U.
So, do you remember the article I posted the other day about accounting for a pool of loans and how values are based on assumptions about the performance of the pool into the future?¬† It was really Part 1 … meaning you should read it first…¬† and it was called…
“If We owned a pool of loans, would WE allow principal reductions?“
Well, think of this as Securitization Accounting, Part 2… at Mandelman U.
It’s complexity we eschew, and everyone’s welcome at Mandelman U.
I thought it might be exciting if I showed you something very few people have ever seen… an honest to goodness peek behind the curtain, if you will.
What follows below are slides from an actual presentation of a Residential Mortgage-backed Securities – RMBS/REMIC deal… but NOT the slides from a “road show” presentation to potential investors… what you’re about to see are slides from an INTERNAL meeting that was actually held back in February of 2006 when WMC Mortgage’s management presented the company’s second RMBS securitization deal to the management from parent company, GE.
And it’s not a proposal being presented to GE’s management… it’s an explanation of a deal WMC had already closed back in late 2005, the mortgage subsidiary’s second such deal… “GE-WMC 2005-2.”
By the way… WMC’s history is fascinating squared.¬† Want to know it?¬† Okay, follow me…
Weyerheuser Lumber Company had a finance company they called Weyerheuser Mortgage Company, or WMC… and they sold it in 1997 or 1998 for $192 million to a company called Apollo Global Management, which was founded by Leon Black in 1990, and today manages an estimated $100 billion in assets.¬† And although you’ll never find it in print, much less prove it in court… the rumor on The Street has always been that Leon’s “silent partner” in Apollo… none other than Michael Milken.
If you’re old enough to remember the first bubble that wiped out your retirement savings, then you’re old enough to remember Leon… he was the Managing Director – of Mergers & Acquisitions at Drexel, Burnham & Lambert… although he didn’t make a stop in federal prison as did Mr. Milken, before getting into the mortgage business and going on to become a billionaire several times over.¬† By the way, it might interest you to know that Mike Milken’s son, Lance Milken, still works at Apollo.¬† Perhaps it was decided that young Lance needed some mentoring, and who better than Leon to make sure his career went swimmingly?
(And I’m not beating up on Mr. Milken, I’m just jealous of Lance.¬† What?¬† Yeah, like you’re not.)
Okay, here’s my favorite part of the story… Leon’s net worth reportedly fell to just $1 billion after the 2008 market meltdown, but before you shed any tears… in the BOOM years since then… he’s done quite well and by 2011, was reportedly right back up to a net worth of $3.5 billion… proving that you just cannot keep a good man down.
So… after Apollo Management acquired WMC, they started adding value primarily by mandating the liberal use of GE-inspired-jargon and redecorating their offices with Six-Sigma-drapes and other window dressings, purchased after a three-hour presentation using a 450 slide deck of Power Point slides… with a corroborating opinion from McKinsey, of course.
And wouldn’t you know it, in 2004, Apollo had no trouble selling WMC to GE for $650 million, thus giving the mortgage company access to the virtually unlimited capital of Wall Street’s darling-of-those-days, the triple A rated… GE, then under the leadership of Jeffery Immelt.¬† You’ll notice when you look at the cover slide of the presentation below, that the name of this securitization deal is “GE-WMC 2005-2,” and WMC Mortgage called itself, “A GE Money Company.”
Back then, if you didn’t already know, GE was a BIG deal… one of only 7 companies in the country with an AAA credit rating.¬† It’s interesting, because if you go back to the early 1970s, there were 60 US companies rated AAA… fast forward a decade to 1982, and that number had been cut in half to 30.¬† By the early 1990s, we were down to just 20 AAAs, and at the dawn of our new millennium you could count America’s AAA-rated companies on two hands even if you’d lost a finger… only 9 remained.
A precious 7 of our AAAs managed to make it to 2009… when the list bade farewell to two companies that no one ever thought would go… Warren Buffet’s Berkshire Hathaway and the venerable GE itself… were both downgraded to a relatively disgraceful… AA.
Yes… Tommy Edison’s electric candle company, that had been the first to bring good things to light… one of the original 12 companies that made up the Dow Jones Industrial Average, and the only one of those 12 still part of the Dow today… along with the Oracle of Omaha’s private mutual fund they call Berkshire Hathaway… yes, they both fell from grace at the hands of irresponsible sub-prime borrowers during the housing bubble.
(In case you’re wondering, the remaining five are Microsoft, Pfizer, Exxon-Mobil, Johnson & Johnson and ADP.)
When GE purchased WMC in 2004, all that WMC did was “whole loan sales,” meaning that it would loan out money for mortgages, and then sell the loans to Wall Street investment banks, who would package and securitize those loans.
So, basically, with GE’s essentially unlimited and AAA-rated access to cheap cash, each month WMC would get, let’s just say, $100,000 from GE, which it would loan out on a mortgage, and then sell that loan to Wall Street.¬† In the beginning anyway, that $100,000 loan would sell for $106,000… then later for $105,000… and then for $104,000… $103,000… $102,000… $101,000.¬† Near the end… before such loans couldn’t be sold to Wall Street at any price and WMC shut its doors during the Spring of 2007… it went all the way down to 25¬Ę on a hundred dollars, which I think would make the sale of a $100,000 loan… $100,025.
Okay, so you want to see how this sort of thing happens?¬† Follow me and I’ll show you.
Leon Black received his undergraduate degree from Dartmouth College in 1973.¬† Jeffrey Immelt graduated from Dartmouth College too… but five years later in 1978.¬† No reason to think they knew each other back then, five years is a big difference at that age.¬† Leon then got his MBA at Harvard Business School in 1975.¬† Jeff got his MBA from HBS too… but seven years later in 1982.¬† No reason to think they knew each other at that time either… seven years difference is a long time.¬† I have no idea, for example, who graduated from one of my alma maters seven years before or after I did.
But then, in 2000, Jeff Immelt became GE’s CEO, after the legendary Jack Welch stepped down.¬† That’s a BIG JOB to be named CEO of GE… especially to follow the larger-than-life, Jack Welch… awfully large shoes to be filled there.¬† In 2000, to be named CEO of GE… well, you might as well have been named King of American Conglomerate-land.
Interesting though… that in 2002… Leon Black became a member of the Board of Trustees for… wait for it… oh yes… Dartmouth College.
Now, in case you have never considered the job description for members of the Board of Trustees at a college or university… you can be sure there’s some fund raising involved… not let’s hold a raffle fund raising… I’m talking BIG TIME fundraising… the kind of fundraising one does by calling on alums who have made it to the top.¬† And nothing says, “the top,” like the C-suite at GE.
And it’s just two years after Leon joins Dartmouth’s Board… and Leon’s company, Apollo starts dressing in Six Sigma garb,¬† that WMC is acquired by GE for $650 million.¬† And you don’t have to be Inspector Clouseau to figure out what went on there, do you?
(Leon… you totally rock… I am not worthy.)
GE – Imagination at Work…
If you imagine WMC selling just one of those $100,000 loans to Wall Street every month for one year… for let’s just say $106,000… you’ll quickly see why the mortgage origination business became Wall Street’s version of a Cash Call Payday Loan.¬† Do the math… $6,000 a month… equals $72,000 a year… by loaning the same $100,000 twelve times and selling all twelve loans to Wall Street bankers.
I know, I’m over-simplifying, but I’m also correct when I describe that as making 72 percent on your money with no appreciable risk.¬† Of course, WMC also had to keep the lights on, no pun intended… but we haven’t even talked about loan origination fees and other “closing costs,” that always find their way into a mortgage transaction.
The first slide is to show you the overall structure of the deal… remember the one I was describing in the last article on pool accounting I said was an over-simplification?¬† So, now you’ll see what happens when some complexity is added…
GE-WMC 2005-2 Deal Summary Remic Chart
Okay, next we’ll take a look at how the deal was rated by Moody’s, Standard & Poor’s and Fitch, related to its “tranches,” which is French for “slices,” in case you’d forgotten.¬† This deal had 17 tranches, by the way, so on this slide they’ve been condensed… in a minute you’ll see them in an expanded view.
Follow my directions, and I’ll point out and explain in simple terms, everything of importance on each slide.
Start on the left side of the slide.¬† Look at the first line of the chart and you’ll see that between 75.4 and 77.5 percent of the deal is rated AAA (S&P) and Aaa (Moody’s).¬† I looked it up here and this means that roughly three-quarters of the RMBS certificates in this deal were supposed to have a .52 percent chance of defaulting.¬† That’s about a one-half of one percent chance of default, which admittedly would have to be considered a pretty darn safe investment, but it also highlights one of the many fallacies at work in these deals.¬† U.S. Treasuries are also rated AAA, but U.S. Treasuries are treated as “cash,” with essentially no chance for default.¬† So, how can anything less be rated the same?
From there, scan down and you’ll see AA+/Aa1. AA+/Aa2 and so on… and all the way down past the BB+/Ba2 you’ll find a line identified as being O/C… and that stands for “Over Collateralization,” and if you recall, at the end of my article on how we valued loans in a pool based on expectations, over collateralization consists of things like extra loans in the pool, and/or cash, and was done to make triple A investors feel that much more secure that they still wouldn’t incur losses even if loan defaults turned out to be higher, or happen sooner than expected.
And you know why, right?¬† Investors don’t receive the amounts designated as O/C, nor are they paying for the amounts of O/C when they buy their certificates.¬† The over collateralization amounts are purely there to absorb first losses, especially those that occur in the first two years that could easily be the result of fraudulent reps and warranties about underwriting standards.¬† If all loans pay as agreed, the O/C amounts will be returned to the issuer, in this case… GE-WMC.
Now look at the right-hand side of the slide and you’ll see the first bullet… “Higher O/C equals less up-front cash and higher back-end residual,” and that sentence is telling GE’s management that because of the over-collateralization in this deal, GE will receive less cash from selling certificates to investors, but in exchange can expect a larger back-end payout.
Just so everyone understands… and this is REALLY IMPORTANT to your education here at Mandelman U...
I realize that on the last slide, you might have noticed indications that millions in cash payments were going here and there, but you’d never be able to figure out much of anything from trying to follow the amounts and arrows on that diagram.
So, I can tell you that in this specific deal there was $46 million in O/C cash, which initially represented 3.25 percent of the deal.¬† But that doesn’t mean that WMC or GE wrote a check for $46 million.
The $46 million in O/C could be some cash… it could also be generated by some extra loans in the pool.¬† But, it could also be handled as a claim against excess interest that’s been built into the deal’s assumptions… remember the “assumptions about the future of the loans” from my last article on valuing loans in a pool?¬† Well, if you change a few assumptions, you’re financials would show some excess interest payments and you could consider those amounts when calculating the amount of O/C as well… get it?¬† Just lower an assumption about loan prepayments for the pool… or lower the number of expected defaults, or reduce the loss severity associated with defaults… reduce the amount of any of those forecasts and presto… extra cash in the deal that can be classified as O/C.
See… isn’t accounting fun?
This specific deal was structured so that during the first two years there would be a relatively higher amount of O/C than after the first two years had passed. ¬† Ostensibly, that was done to create the appearance that the triple A investors were being protected from the sort of early loan defaults that would have to be the result of shoddy underwriting or even fraud.¬† You know… like in case WMC was lying in their representations and warranties about the underwriting of the loans.
So, with the high O/C during the first two years, I’m sure investors felt safe as far as being protected from WMC just packing this deal with loans that would never make a payment or default in year one or two.¬† After two years of making payments, one could no longer blame a loan that defaults on its originator.¬† Loans that default after two years, result from life events and other factors that are beyond what a loan’s underwriting could reasonably foresee or prevent.
To illustrate all of this, this deal provided that GE would receive $28.5 million from the O/C at month 26, assuming no “triggers” had already required that money to be used to absorb early losses.
The next bullet states that WMC believes that they’ve made some progress convincing Moody’s that their loans were of such high quality, that in the future they wouldn’t need as much O/C to get the ratings they needed to make the deal attractive… which was obviously included to please GE management about the potential for future deals to require less cash.
The next heading is telling GE management that it’s the same process that they’ve used successfully in the past when they only sold whole loans, and again this is included as a warm and fuzzy for GE management.
And lastly, on that slide… there are a couple of points about split ratings.
A split rating occurs when the same bond is rated differently by rating agencies.¬† For example, a bond could be rated AAA by one agency and AA¬† or A, by another.¬† This can occur because one rating agency places a different amount of emphasis on certain variables, or because one agency views something about the issuer differently than another… perhaps a recent acquisition by the issuer is seen more or less favorably.
Split ratings also occur simply because the different ratings agencies each handle ratings differently… S&P ratings are said to measure the probability of default… Moody’s, on the other hand is said to be measuring the amount of the expected loss in the event of default.
(Think car insurance for a moment… S&P would rate me as a driver with emphasis placed on the probability of me having an accident… while Moody’s would base my rating primarily on the cost of the car I was driving.)
It’s a slightly different perspective on what’s most important in a rating, and you can see why investors want to trade based on Moody’s… they’re concerned with how much they might lose. The Basel Accord, by the way, is seeking to mandate that BOTH perspectives be provided to investors… leave it to the Europeans (who originated Basel), to introduce common sense to Wall Street.
In this deal, and remember this is referring only to the non-investment grade bottom tranches, rated BBB- and below… it says that the “tranches trade at Moody’s rating,” but the second bullet says that these tranches have already been priced as if the ratings were dropped, which appears to protect the investors.
The next slide shows you all of the different investors that bought pieces of this deal, and because this presentation was used by WMC to show its parent company, GE, just how fabulous everything was… the slide’s title reads: “Strong Investor Demand.”
The column on the left shows who bought the AAA/Aaa tranches, in the middle you can see who who bought the “Mezzanine tranches,” which those in “the biz” would just call “the mez.”¬† The column on the right shows investors in the various ‘B’ rated tranches, and below that is a list of those that bought the unrated “NIM,” or some would say, the “equity tranche.”
The “NIM” refers to the securitization of the excess or residual cash flows from one or more mortgage-backed or asset-backed security, which mostly come from the spread that exists between the interest rate on the mortgage-backed security… and the interest rate on the underlying pool of mortgages… the amounts that are not needed to absorb losses or increase the amount of credit enhancement in the underlying deal or deals.
Years ago, those issuing mortgage-backed securities earned this excess interest as the mortgages aged over time, but we should all know by now that Wall Street hates waiting for anything, least of all money that might otherwise be pocketed today.¬† So, once again, thanks to “financial innovation” and/or “market efficiencies,” issuers discovered that they could securitize their residual interests and sell them to investors… and the NIM… or “Net Interest Margin” bond was born.
NIMs became a fast growing sector in the home equity and mortgage market because they traded at much higher yields than bonds with similar ratings… back in 2002, we’re talking 8-9 percent.
One reason NIM bonds could be sold to investors was that companies like Radian Group Inc. started providing “credit enhancement” for NIM bonds.¬† In 2005, for example, Radian’s 10-K showed the company had written $99 million of default insurance risk on NIMs.¬† (Radian’s peers include PMI, MGIC, Ambac, et al.)
NIMS in the 1990s v. after 2000…
Because the cash flows into NIMs are subordinated to the needs of the deal… meaning that NIMs take the first losses, or that the excess cash generated may be needed to increase the amount of credit enhancement in the deal… the volatility of prepayments that could occur when interest rates fell,¬† and/or the timing and severity of losses resulting from defaulting loans, often had a huge impact on a NIM’s performance.¬† (Now do you see why prepayment penalties were invented?¬† To protect bond holders, but especially the NIM bond holders because they take the first loss.)
Certain “triggers” were written into these deals that would change how cash flows¬† would be allocated and therefore impact the amount of excess spread that would be allocated to the residual holder.¬† The two most common cash flow triggers used were a “delinquency trigger,” and a “cumulative loss trigger.”
Such triggers might¬† require an increase in the targeted amount of O/C (over collateralization, remember?)¬† These were known as ‚Äústep-up‚ÄĚ events.¬† Or, a trigger could prevent a release of O/C… in which case it would be called a ‚Äústep down‚ÄĚ event.
Some triggers are considered to be ‚ÄúNIM friendly‚ÄĚ triggers because they don’t allow for an increase… or “step-up” in O/C, above whatever initial amount of O/C was in the deal, and/or because they didn’t come into play until the step-down date.¬† But other deals had more onerous triggers that could cut off cash flows to the NIM by prohibiting any decrease… or “step-down” in the amount of O/C, or by requiring an increase in the amount of O/C once triggers were hit.
Now, don’t get overwhelmed by this stuff… it isn’t hard to get… read it again slowly if you’re feeling overwhelmed.¬† The concepts are simple, it’s the terminology that takes some getting used to, I understand.
The financial services industry, and especially the bond market in my opinion, goes out of its way to make things sound like you need one of their experts to decipher a secret code in order to participate in the race to riches.¬† But the truth is… it’s all about debt.¬† Someone is loaning money, someone is borrowing money… maybe someone else is insuring that something will happen, or betting that it won’t… and several different someones are setting up the deal, managing the deal, or selling the deal.
It’s like my favorite line from the movie “Bull Durham.”
“This is a very simple game.¬† You hit the ball, you catch the ball, you throw the ball.¬† Sometimes you win, somethings you lose, and sometimes it rains.¬† Think about that for a while.”
The Investors in this Deal…
This slide is going to be fun, I promise…
Start top left… under the heading “Investor.”¬† See it… FHLMC… Federal Home Loan Mortgage Company… otherwise known as good old Freddie Mac.¬† Out of roughly $1.4 billion… Freddie bought $319.3 million of the Aaa-2yr, and by the way… the FHFA is currently suing GE, and all the GE companies, Morgan Stanley (the deal’s ‘adviser’), Credit Suisse (co-manager and underwriter), et al.
Just so you know… as a result of the FHFA’s suit… the U.S. Attorney for the Northern District of California convened a Grand Jury, and the FBI is currently investigating to determine whether a crime has been committed.¬† So, now… with these slides… you can follow along and play the Mortgage Securitization Fraud Game at home.
Then we’ve got other investors you’ve heard of… Chase, PIMCO, BGI, the Agricultural Bank of China… HSBC London… Alaska Perm Fund… and just keep reading down the list and you’ll find BofA, JPMIM is a JPMorganChase… Societe General (don’t try to pronounce it, you’ll only butcher it… just say “Sock Gen.”) and then Fortis, Wharton… and it’s not “Robobank,” it’s “Rabobank,” although calling it “Robobank” is much funnier.¬† And remember… the amounts shown are in millions.
Then go down the middle column and see who’s there… Munich RE is an re-insurance company… Teachers, is the Teachers pension plan… Hyperion, which is a hedge fund… Hartford, insurance again… Highland Capital is another hedge fund… towards the bottom of the middle column, see “FIDAC- Annaly?”¬† Yeah, well they have a bunch of brand names, but they packaged and sold Collateralized Debt Obligations or CDOs, so they were probably buying the “Mez” in this deal to make it into a CDO, or CDO-squared… as were many of the others as well.¬† (After we’re done with the MBS portion of our class, we’ll move to CDOs.)
And notice the one labeled, “U/W Purchase?”¬† That’s Credit Suisse First Boston, or CSFB… this deal’s underwriter showing the other investors how much confidence they have in the deal by buying in at $12.6 million.
Then in the column on the right… the investors in the “Sub Tranches,” including Ellington, a hedge fund in San Francisco… JPMIM again, or JPMorganChase… Eurohypo, a German bank… Princeton, the university’s investment fund… Citi London… Fischer Francis is a fixed income investment management firm… Deerfield is another hedge fund… BUT STOP on C-Bass for a moment, because C-Bass was this deal’s loan servicer.
C-Bass, which stands for “Credit Based Asset Servicing and Securitization,” was basically a securitization advisor until they bought Litton Loan Servicing, which they sold to Goldman Sachs in 2007, and who sold it last year to Ocwen.¬† MGIC, a monoline insurer, owned 45.5% of C-Bass… and MGIC’s competitor, Radian ALSO owned 45.5% of C-Bass.¬† It’s so clubby.
C-Bass filed bankruptcy in 2010, and why not?¬† The company’s filing listed assets of $10 to $50 million… and debt… OMG at over $1 BILLION.
You’ve got to love these guys…
I mean, if you’ve only got $10 million… and you can go bankrupt for a BILLION… I say you win.¬† By the way… the CEO of C-Bass was a Goldman Sachs alum… and his partner came from Citigroup.
Just above the NIM you’ll see that GE-WMC bought in as well, which is another way to make investors feel like they’re not being viewed as a fatted pig that’s about to be served with an apple in its mouth.¬† As in… “Look, Tom… they’re investing in it too… it must be safe, right?”
Okay, finally there’s the investors in the NIM.¬† There’s Amaranth, a hedge fund… MKP, alternative investment advisers… and on down the list of mostly hedge funds that were all greedily swinging for the fences by investing in the deal’s un-rated, but crazy-high-yield, NIM bonds.
A COUPLE MORE KEY POINTS…
Now let’s talk about the question, “Who owns my loan?”¬†
The answer you often hear is, “the investors,” but you should be at least starting to see why that’s not a particularly helpful answer, right?¬† I mean, which investors?¬† This deal has 17 tranches, and 70 investors, although some are duplicates because the same company bought into more than one tranche.¬† And some of the investors are hedge funds, while others are buying in order to repackage what they bought into a CDO that they will be reselling.
So, who is the investor?¬† Well, not only do you heave to think in terms of which one?¬† But, depending on the tranche they bought, they may be receiving payments… or they may have already been wiped out by losses and that depends on the timing and the triggers and the over collateralization, or O/C, etc. etc.
And, let’s just say that we’re talking about one of the investors in the deal’s NIM bonds, and let’s say that investor has been wiped out due to losses resulting from loan defaults… but maybe that investor had purchased default insurance from Radian… then what?¬† Does that mean your loan was paid off when that investor’s NIM bonds defaulted?
Obviously, the answer is no.¬† Keep in mind, the deal we’re looking at was approximately a $1.4 billion dollar deal.¬† At the bottom right of the chart below, you’ll see that Citadel… a hedge fund… was an investor in the deal’s NIM bonds to the tune of $1.3 million.¬† Should those bonds default, assuming Radian had insured Citadel’s interest, Radian would be responsible to make the remaining bond payments with interest to Citadel.
However, even that simplified example assumes a lot.
The insurance companies that offer this type of insurance are known as the “monoline insurers,” because in 1989, the State of New York enacted Article 69, which amended the state’s insurance law to make “financial guaranty insurance” a separate line of coverage, and the new law prevented other types of insurers, such as property and casualty, life, and multiline insurers, from offering it. For years, the monolines claimed this exclusive focus made them stronger, but when the meltdown in the sub-prime bond market hit in July of 2007, their limited focus made it a certainty that they would be wiped out.
In early 2008, to give you an idea of how fast this last meltdown actaully melted down… when Fitch downgraded monoline insurer Ambac from AAA to AA, it triggered downgrades on more than 100,000 Ambac-insured bonds worth roughly half a trillion dollars.¬† Plus, the monolines got into the market of insuring CDOs, which today, for the most, part we call “toxic assets.”
Once insuring CDOs, it was only logical that the monolines, like Ambac and MBIA, get involved in placing large bets on CDOs through the use of credit default swaps, but at the same time, others in the market were using the same sort of contracts to bet against MBIA and Ambac.
Ambac filed for bankruptcy in November of 2010, subsequently filing lawsuits against just about everybody on Wall Street, from JPMorgan to Bank of America to Credit Suisse to Bear Stearns… alleging improper underwriting and more.¬† Last month, Ambac received permission to emerge from bankruptcy and just yesterday, Ambac sued JPMorgan Chase saying that it has incurred over $200 million in claims on seven RMBS deals that Bear Stearns fraudulently induced it to insure.¬† According to Reuters, the securities in question have loss over $1.8 billion.
MBIA filed for bankruptcy in 2009, and a whole line of banks showed up to challenge the filing, claiming that MBIA wouldn’t be able to make good on its credit default swaps if it went ahead with its plan to restructure into two units.¬† Since then, MBIA has settled its credit default swap liability with Morgan Stanley and others, and… well… the saga continues.
The point is this… maybe there was default insurance purchased by some investors in certain tranches of a given RMBS, certainly not all investors purchased such coverage, and even so… the insurance carrier may have filed bankruptcy and¬† investor claims may not have been paid, or paid at a reduced amount.
And even if a given claim was paid to a given investor in a given tranche, the default insurance we’re talking about here, WOULD NOT be paying off a MORTGAGE in the underlying pool of loans, as some people have apparently been led to believe.
And lastly… when you read the list of the 70 investors in this deal, does it really seem like homeowners would be in any way better off negotiating with any of those hedge funds and assets managers for a loan modification than they are with their servicer?¬† I’m guessing that we’ve got a much better shot with BofA at its worst, than we would trying to convince a hedge fund or asset manager that has just lost fifty or a hundred mil.¬† I’m just saying…
The next slide discusses the deal’s “Key Assumptions,” just like I talked about in my last article on valuing loans in a pool… you remember, right?
Now, this slide was titled “Residual Value – Refresher,” because it was prepared to refresh GE’s executives as to the investment that GE made in the deal.¬† And under “Key Assumptions” at the top, you’ll see all the same terminology we’ve been using since the valuing a pool article…
- Prepayment speeds (which includes refinancing of loans)
- Losses (includes loan defaults, which reduce the amount of over collateralization)
- Discount rate (that’s the rate used in the NPV calculation to arrive at the PV or present value)
- Triggers Pass/Fail (and we just discussed these above when talking about the NIM)
Next the “residual” is discussed (just like we talked about above when discussing the NIM)
The excess spread is the amount earned in excess of what’s paid out to investors, and the bullets describe how it’s treated, first covering CREDIT LOSSES, which are LOAN DEFAULTS, and then going to maintain the O/C before release beginning in month 17, with the $28.5 million expected to be released in month 26.
Once again… remember that the “Loss Assumptions” shown on the slide below, relate to GE’s investment in the deal.
Start at the chart at top left… DEFAULTS…
The loss assumptions data, which was provided by Merrill Lynch, separates out the first liens from the second liens and used 2003 as benchmark data.¬† Now remember… this presentation was created on a deal that closed during the fall of 2005.¬† Yet, the data being used as a benchmark is from 2003… roughly just 24 months later.
However… look at the graph on the right, which shows “Cumulative Defaults.”¬† Month 0 is when the deal begins its life and notice that 24 months from Month ‘0’ the cumulative defaults are between 1-2 percent.¬† So, looking at only 24 months of default data from 2003 clearly wasn’t enough to predict defaults in this deal going forward… not to mention the fact that no one lost homes to foreclosure in 2003 when home prices were going nowhere but up.
So… the top left chart shows that defaults on first mortgages are supposed to be 8.83 percent, 8 percent on seconds.¬† The middle chart on left shows “Severity,” or as we’ve been calling it, “loss severity,” and it’s the amount expressed as a percentage that we think we’ll lose on the 8-9 percent defaulting loans.¬† And the bottom chart shows expected losses.
To see how it all works together, take the 8.83 percent… multiply by the loss severity of 35 percent… and you’ll get the 3.09 percent in losses, that appears in the chart at bottom.
Now, obviously, WMC didn’t have a whole lot of data on loss severity, which makes sense because in 2003 there probably weren’t any losses resulting from loan defaults to speak of… in fact, in 2003, a foreclosed home would have more than likely resulted in a gain as prices continued to rise.¬† So, the 35 percent loss severity that was plugged in would appear to use the SWAG methodology (that’s a “Scientific Wild Ass Guess,” in case you weren’t aware).
So, obviously using 24 months of data going forward from 2003 wasn’t sufficient, but what data should the issuers of these types of securities in 2005 used to forecast future defaults, loss severity percentages and prepayments?¬† I mean, 2000 to 2002 was a pretty bad recession, in case you’d forgotten, and going back into the 1990s would have been comparing completely different loan products.¬† I’m not defending them, nor am I saying it couldn’t have been better, but I just want to show you the nature of the problem from different perspectives.
The overriding point being that no one saw what was to come, specifically… and lots of people lost a lot of money… and there will be lots of litigation going on for lots of years to come as a result.
What went so terribly wrong was how our government handled it, choosing to provide untold TRILLIONS to banks, while abandoning America’s homeowners financially, leaving them to fend for themselves, and sending them hat in hand to giant financial institutions asking that their loans be modified.
I’m not defending the way the banks have handled the situation, but I don’t pay taxes to my bank, nor do I elect my bank’s President, and I certainly have never depended on my bank to protect my rights as an American citizen.
To those who won’t like what I’m about to say… tough.¬†
The specific failure of which I speak rests on the Obama Administration ALONE.
And on the next slide below, we have the basis for the deal’s “Prepayment Assumptions.”
These graphs and data presented were to examine most or all of WMC’s loan products, in different combinations with prepayment penalty and without… and WMC’s actual prepayment experience between 2003 – 2005… over 31 months.
This slide, which is showing WMC’s actual prepayment experience over that 31 month period, not only shows how they arrived at the key assumptions for this deal, but it also shows how frequently people were refinancing their mortgages between 2003 and 2005… which is pretty darn amazing, by any standard.
Prepayment Assumptions GE DEAL
Well… that’s all I’ll cover for now… I don’t want to overwhelm everyone… assuming anyone is actually still following along… so, we’ll leave the rest for the next installment.¬† Stuff like the Collateral and Due Diligence and Securitization Deal Structure… and the Computational Materials, which are the detailed data points they used for things like average life for each tranches, trigger events, and then swaps and the hedging strategies… things that sound more complicated than they actually are.
Nothing about this is truly hard to understand… it’s all grocery store math, really.
We’ll also be drilling down into all the different types of insurance involved in RMBS securitizations, that way maybe people will stop telling me they think their loan has been paid off three times.
And finally we’ll look at the Clean-Up Call, which is what happens at the END of a RMBS REMIC deal… here’s a preview to keep you coming back for more… the specific language in this deal says…
“The servicer may repurchase all of the mortgage loans and properties acquired on behalf of the Trust when the loans remaining have been reduced to less than 10 percent of the original balance.”
Oh my… I can hear the questions popping like popcorn.¬† Don’t worry… get this stuff down and we’ll move on to the next part in a few days.¬† I’ll have you securitizing your own pools of loans in no time.