A few readers have been kind enough to pass along Wall Street research reports on these beasties.
I won’t break any copyright laws or pass them to anyone else, so anyone else interested in joining the effort is welcome. What I will do is paraphrase and provide qualitative reaction to the pieces I read, providing I can add something to the discussion by commenting.
You can reach me via the blowback page, which provides a link to my e-mail. Your e-mail address will not be displayed to the world, and neither will your message.
Today I got a summary of the estimated change in book value at a half dozen mREITs over the course of the second quarter. No surprise that NLY was the leader, given the fact that fixed rate MBS outperformed (on a price basis) their ARM MBS cousins.
The best-performing coupon was the 4.5% fixed rate passthrough, no doubt in part because the Fed’s QE program bought more of these bonds than the mortgage banking world was able to create. That caused the largest fail-to-deliver situation in (the cash) bond market history.
I have to specify cash, because the first generation of Credit Default Swaps specified delivery of the defaulted bonds to receive payment. Of course, since those swaps were protected from prying regulators, counterparties and investors eyes by the Commodities and Futures Modernization Act, we can’t know how many bonds would have to be delivered against the CDS contracts.
That said, it was a shock to the market when the Delphi bankruptcy caused a massive short squeeze in those defaulted bonds. The music stopped, and there were more than ten times as many insurance holders as there were bonds being insured.
I personally would have liked to see that first big case resolved by having all the winners that had taken out insurance against bonds they didn’t own make good on their contracts. I really don’t think they deserved to be let off the hook with a cash settlement, but should have suffered the market effect of having to buy the bonds to deliver them if they wanted their settlement money.
Such an outcome might have slowed the impulse of the bears to keep adding to their positions, and, strangely enough, might have led to the subprime mortgage crisis and subsequent housing price decline being less than half as severe as it became. But that would be the result of a really free market, not the faux free market we have.
Imagine what a different economy we’d have today if the nationwide housing stock had suffered only a 10% decline, which would have, in turn, left Fannie and Freddie still solvent, and the taxpayer not footing the bill to maximize profits for a few speculators.
But that’s not what happened, and the Fed spent $1.25 trillion buying MBS after the horses bolted from the barn. I recall seeing the weekly and monthly stats on the program as it ramped up, and I was amazed that they were buying more current coupon (mostly 4.5% and 5%) MBS than existed or were being created.
In the mortgage market, the aggregate fails exceeded $1.5 trillion on Settlement Day each of the past two months. The “punishment” for not delivering bonds you’ve sold is that you pay the interest on the bonds to the next buyer in the chain without collecting it yourself.
Let’s stop a second and not go nuts by interpreting that massive total fail number into its potential liability of 40 basis points (4.80% is an easy-to-use approximate number for the coupon on the fails) times $1.5 trillion, or $6 billion, in money bleeding out of mortgage traders.
In fact, most of the fails are matched off against other purchases, so it’s a daisy chain of housekeeping/accounting hassles, with each player in the chain having to “claim” money from their failing counterparty and, in turn, paying essentially the same money to the trading counterparties they sold to.
Still, there’s no doubt that most participants in the market would rather be done with it, and deliver the bonds so they don’t fail any more. That pushes up the prices, especially the prices for “renting” the MBS for a month via “dollar rolls.” Bloomberg’s Jody Shenn has been writing about this for a couple of weeks, highlighting both the record high prices for MBS and the Fed’s efforts to accomodate traders that sold bonds they can’t seem to locate and deliver.
For any readers that have gotten through this far, I’ll share the information now on mREIT book value estimates. The research piece I’m looking at guesses that NLY jumped a full 12.3%, and ARMs-oriented ANH a solid, though less exciting 3.6%. They don’t cover HTS, AGNC or CYS, but it would be easy enough to take their respective mixes of fixed and floating MBS and interpolate from the NLY and ANH datapoints.
I can’t get too exercised about precise calculations in an estimate, anyway, because the variance due to hedging strategies or asset sales, or even timing of reinvestment can blow right past the modeling variables and make the numerical analysis look silly.
For the non-Agency mREITs, it wasn’t nearly so pretty. MFA has been reinvesting in private label paper, and didn’t participate in the first quarter book value sweepstakes as a result (estimated change in book value 0.2% for the quarter). IVR was also flat, and CIM gained 1.6%, purely on the back of a large secondary that was accretive to book value.
One mREIT that I haven’t mentioned, but appeared in the report, was Two Harbors (symbol TWO). They are estimated to have lost nearly 4% of their book value over the course of Q2 this year. Not good.
If you’re east of the Mississippi, just keep telling yourself that global warming is a political scam by Al Gore. Maybe it will make you pay less for the air conditioning.