Tuesday, March 31, 2009
the hancock building: CMBS barometer
This example is consistent with the 50% recovery rate on foreclosure in most residential real estate markets. There are entire tracts that will be left for the junkies, to be sure, but for the most part, the California market is clearing quite nicely at levels sufficient to avoid impairment on the AAA-rated tranches of subprime and Alt-A loans. If valuations hold at around these levels, the position of the banks should be quite survivable.
In a nutshell, securitization “worked.” It created an awful mess in some respects, but it did help buttress the banks against losses, while hedge fund and private equity investors died like flies.
creditsights some time ago asserted that, given expected loss severities averaging around 35%, subprime defaults in a typical securitization pool would have to rise to 30% to begin impairing a standard triple-a subprime bond. with apologies to goldman, we have to be very close to or beyond that already in california, where average price declines are in the area of 40%, likely more for subprime properties. (in some cases, much more.) add in the costs of foreclosure, and loss severity must be well in excess of 50% even as subprime foreclosure rates exceed 25%.
nevertheless, indexes on view at markit indicate a fair value of less than 30 cents for triple-a subprime. at this point, cash flow impairment can't be anything like that bad. this is a good example of the hows and whys behind what may well be the refusal of the banks to unload what would be, in the eyes of warren buffett, their most profitable assets from current market marks into any government-sponsored liquidation.
of course this doesn't do much to defang tavakoli's assertion that securitizations were designed to lose money for unsophisticated buyers -- the originating banks kept the super-senior stuff that goldman thinks is unimpaired only to foist massive losses onto others.
more importantly, however, it doesn't address the primary concern -- price declines and foreclosure rates appear nowhere near peaked. so triple-a subprime bonds aren't as severely damaged as marks would imply at this moment. will they be in a year's time? two years? it's a wide open question, given manifest uncertainty regarding government fiscal willpower.
in any case, the lack of systemic balance sheet that is a major side effect of the collapse of shadow banking will keep this divergence between market prices and cash flow valuations (to the extent that a value can be ascertained) open. there is far more of this stuff for sale than capacity to buy it at virtually any price -- the money is gone. unless tim geithner's PPIP is a success on a scale stunning even to the most hopeful, that will remain true for the duration of the crisis -- indeed probably grow more and more true throughout.
UPDATE: zero hedge with a differing view on the hancock auction from morgan stanley, which analysis discounts the implied value of financing.
In a foreclosure auction today, the John Hancock Tower - a marquee building in Boston - traded at $660MM to Normandy Real Estate Partners. That same property was appraised for $1.3BN in 2006 and traded for $935MM in 2003. This is VERY negative for commercial real estate. At face, it looks like even top quality assets are down 50% from their peak, but that forgets the value of the financing that Normandy now gets to assume. There will still be a $640.5MM mortgage on the property at a rate of 5.6%.
What is the value of being able to get a 97% LTV loan at 5.6% these days? Let's say you can get a 60% LTV mortgage ($400MM) at 8%, and the other $240MM in mezz financing (which has no chance of getting done in this market) could hypothetically get done at 15%. That combination produces a weighted average financing cost of almost 11%. A 5.6% mortgage at 11% yield is about a 70 $px, which means the value of assuming the existing financing on the Hancock Tower is close to $190MM. The real clearing level for the top commercial property in Boston was only $470MM - down 65% from 2006 levels and down 50% from 2003 levels. If we assume 2008 NOI numbers are still accurate, this would be a 9.5% cap rate adjusted for the financing. Without adjusting for the value of financing, the purchase price of $660MM looks like a 6.7% cap rate and $383/sqft - rich, relative to 1540 Broadway (NY office vs Boston office) recently clearing at ~$400/sqft.
**The main takeaway: property values are down A LOT more than people think, especially when considering the implied value of financing. Caveat Emptor.**
so the truer parameter of goldman's above analysis should be a 35% recovery in foreclosure -- a loss severity of 65% -- not 50%. and this for the widely-acknowledged best property in boston.