naked capitalism

The structured credit analytics/research firm R&R Consulting released a bombshell today, and it strongly suggests that prevailing prices on non-GSE (non Freddie and Fannie) residential mortgage backed securities, which are typically referred to as “private label” are considerably overvalued.

R&R Consulting described how the reports presented to RMBS investors show losses at the loan level (which is super eye-numbing detail in the investor reports) that have NOT been allocated to the bonds (boldface ours):

On the securities performing at December 2011, a universe of approximately $1.42 trillion, R&R estimate the amount of additional losses likely to materialize is $300 billion, with one-third concentrated in ten arranger names, including Countrywide, Morgan Stanley and JP Morgan. About 17,000 tranches, or 34% of the universe analyzed by R&R, may lose up to 83% of their remaining principal.

In addition, R&R estimates that approximately $175 billion of losses already incurred on the loans have not yet been allocated to the bonds in the related transactions. Failure to allocate realized loan losses could distort the valuation of related RMBS tranches.

In the course of conducting valuations on RMBS, the R&R analytics team discovered widespread, serious, repeated data discrepancies. Ann Rutledge, a founding principal, asked the team to measure the magnitude of the discrepancy on the RMBS universe. To do this, R&R subtracted cumulative losses allocated to the tranches from unallocated, expected losses, calculated as the sum of defaults, bankruptcies, foreclosures and REOs minus recoveries. “The results were very disturbing: $175 billion of unallocated current losses and $300 billion of imminent losses,” Rutledge said.

Now you might say, how can investors NOT know this is happening? Have you ever looked at an investor report on MBS? They are really really nerdy. Summary stuff up front, tons of pages of detail. Now bond fund managers are presumably paid to care about nerdy stuff like this, but I have spoken to some MBS lifers who have gone to the buy side, and they tell me that the level of expertise among MBS investors is not high.

But, but, but….some of you are protesting….surely these errors are just innocent mistakes? That’s a nice theory, but the numbers are huge, and the “mistakes” happen to line up with more profit for servicers:

The implication for bond holders in RMBS is significant with respect to both estimates. Subordinated securities in the RMBS with probable future losses ought to be written down by such losses but instead may be continuing to receive interest owed to more senior tranches. It could also mean that servicers are earning fees against loans that have already been liquidated, which also reduces the amount of cash to pay senior bond holders. For example, in one month, servicers could generate $75 million or more in inappropriate fees against the $175 billion in unallocated losses.

Translation: when the servicers don’t write down the bonds in a securitization to allow for ACTUAL and pretty certain losses, the effect is that junior tranches show artificially high balances (remember, as losses occur, the effect is to wipe out tranches from the bottom of the securitization up. The riskiest tranche fails first, then the next riskiest, and so on).

Servicers ALSO advance principal and interest to bondholders when borrowers quit paying, in theory up to the mortgage balance (we’ve seen cases where advances exceeded the mortgage balance). Then when they foreclose and liquidate the loan, the servicer reimburses himself for the advances and his fees and foreclosure costs first.

So, if they report artificially high balances in junior tranches, they are paying interest to investors who don’t deserve it. The result, when the foreclosure occurs and the real estate is sold, is that the interest overpayment to the junior bondholders reduces the monies that should have gone to the senior bondholders. Oh, and those junior bondholders are more likely to be hedgies, and those senior bondholders are more likely to be pension funds, bond fund (the sort that you might hold in your 401 (k) and insurers. The costs to the insurance industry alone means that this is not a fat cat investor issue but affects all of us (losses to insurers eventually lead to higher insurance premiums to compensate for the shortfall in investment income).

Reports like this are why I am cynical about talk of mortgage “investigations”. The evidence of servicer misconduct is rife. I’ve gotten numerous reports about various types of servicer scams, not just ones that hurt borrowers, but tons that impact investors (for instance, it is common, and it may be pervasive, that servicers delay reporting the liquidation of a loan to the investors. Why? The longer a loan appears to be in the pool, the longer they can collect servicing fees).

To my knowledge, the R&R report is the first effort to place a dollar figure on one type of mortgage-investor-related abuses. I’m not surprised it is so large. What I am surprised at is that no investor seems to have noticed this type of pilfering.