Sunday, August 10, 2008

High Risk, High Yield Strategy Keeps Anthracite Under Pressure


Anthracite reported mixed results on Friday, well shy of the Company's .31/share dividend, noting that it was still receiving margin calls on the 18% portion of its portfolio that was not match funded. The Company's cash levels are low ($38.6MM unrestricted) relative to an amortization payment of $31MM required on September 30th under its newly amended Bank of America credit facility, one of two primary credit facilities.

Anthracite recently established a third credit facility with its parent Blackrock. The support from Blackrock is encouraging, and AHR appears to be using that facility to meet its margin calls. AHR drew down $52.5MM when market conditions worsened around the time of the Bear Stearns failure, then paid it back down to zero in April as the market improved. AHR drew on this facility again on July 28th in the amount of $30MM.

It's worth noting that the Markit Group's analagous BBB CMBX index shot through the roof (up is down) in June and July, at the same time that AHR drew on the Blackrock facility, and a graphical illustration of that index helps put Chris Milner's market commentary in context:

"After a period of relative stability in April and May, the markets suffered another major setback in June and July as continuing economic weakness combined with the challenges faced by the residential mortgage market put significant pressure on financial stocks and credit spreads."


Investors in Anthracite need to continually reconsider the the damage that is being inflicted by this index. The reason is that this damage is divided into two categories: permanent capital losses, which until now have been largely confined to the subprime mortgage sector and related structured products; and temporary mark-to-market losses, which are hitting all credit products, including the securities in which Anthracite invests.

REITwrecks readers say that six times fast know that these mark-to-market losses on CMBS have had little or nothing to do with the underlying performance of the collateral, which with very few exceptions continue to exhibit strong financial performance.

However, more and more observers are questioning whether these paper losses will not soon become permanent capital losses. Obviously, with Anthracite's portfolio concentrated on the lowest rung in the CMBS food chain (BB rated "Controlling Class" CMBS), pressure on the stock has been consistent.

AHR has a fairly well diversified BB CMBS portfolio by geography, but the portfolio summary in the earnings release shows that AHR could be in some trouble with respect to the more recent vintages (2005, 2006 and 2007). These three years comprise almost 75% of the portfolio. Those three vintages also comprise about 49% of the outstanding CMBS market, so it's not surprising that AHR would have a concentration there. In those three vintages, AHR currently estimates that its portfolio will experience collateral losses of almost 50%.

As everyone knows, these particular years were also the height of the bubble, and Fitch recently reported that it believes defaults on CMBS issued in those years could quadruple from their current levels, under a worst-case scenario for the U.S. economy.

According to the Fitch estimates, borrowers would default on an average of 17.2% of securitized commercial mortgages over 10 years if the US economy dips into a recession with 0.2 per cent contraction in growth (compared with current default rates of 4 per cent), which is a rise of 330 per cent. Fitch's London office produced similar dour commentary on the European CMBS market a few days earlier.

Such a scenario corresponds “to the negative predictions currently offered by commercial real estate experts”, analysts at Fitch wrote. This would happen if the economy suffered a similar downturn to 1991, and assumes that the value of properties covered by the deals falls by 25 per cent, and cash flow from rents by 15 per cent.

Under a more mild recession, which Fitch thinks is more likely (0.8 per cent economic growth), the default rate would still rise to 13.7%, roughly double the norm.

"Controlling" Class CMBS: controlling what?

AHR says it likes the controlling class CMBS because it is given control over the collateral in order to effect workouts. But what if there is nothing left to work out? Fitch says the more severe scenario would cause non-investment grade bonds – B and BB rated CMBS – to suffer loss rates of 100% and 95.9%, respectively. Meanwhile, only 30.6% of the lowest-rated investment grade bonds – BBB rated – would experience losses, while loss severities would rise to 37.9% from an historical average of 33.5% (but still a far cry from being completely wiped out).

Clearly, the report suggests that recently issued CMBS were the subject of inflated underlying property values and weaker underwriting standards experienced at the height of the boom in 2006 and 2007. The report's survey covered all Fitch-rated bonds issued during those two years, which were comprised of 74 deals worth $217.3billion. That was about 60% of all CMBS issued during the period.

Is AHR misunderestimating?

Most reasonable people agree that conditions are worsening. Not surprisingly, the Company also reported that it increased the loss assumptions on its controlling class (across all vintages) CMBS from 1.31% of outstanding collateral at December 31, 2007 to 1.44% at March 31, 2008.

However, the Fitch report suggests that losses in the controlling class losses could be much more severe. Indeed, the report notes that under the 1991 scenario, 3.6 per cent of all 2006 and 2007 bonds (not just BB) will suffer losses. Given that 75% of AHR's outstanding BB CMBS collateral consists of deeply subordinated 2005, 2006 and 2007 bonds, AHR's loss estimates could be light.

Nevertheless, even this pessimistic Fitch scenario flies in the face of the losses implied by the CMBX index. That index continues to suggest that losses on some CMBS will exceed the worst levels experienced in the 1990s.

Indeed, Bloomberg reported on Friday that "yields on commercial real estate securities relative to benchmark rates rose to the highest since March on concern that retailers won’t be able to repay debt as consumers cut spending. Spreads on AAA rated commercial mortgage-backed bonds widened 10 bps during the week… to 250.5 bps more than 10-year swap rates… Demand for commercial real estate securities is waning as retailers are forced into bankruptcy during the economic slowdown."

Obviously, the question on everybody's mind is how much worse will the economy get and how long will it last? Amidst the uncertainty however, one thing is clear: as reality unfolds, whatever it may be, AHR's "controlling class" CMBS will be among the very first to receive it. Unfortunately, being a bellwether (bell-weth-er, n. 1. a castrated ram ) in this particular market is not an enviable place to be.


Disclosure: None at the time of this writing.

Labels: , , ,

2 Comments:

Anonymous Anonymous said...

Not to take exception with some of your other points later in the piece, but when you start out your blog by quite incorrectly identifying the controlling class as BB rated, I must then only deduce that the content which will follow must be subjected to a high degree of scrutiny. Typical of the CMBS market (RIP) the BB bonds are publicly offered, the B bonds are privately offered and the NR piece (aka first loss or Controlling Class Certificates) are also privately offered. Why? Because these two classifications (B and NR) of risk are so high, the investors require analytical access to non-public loan files in order to create cash flow models used to predict returns and therefore to price bonds.

First of all, the controlling class certificates (CCC) (otherwise known as the first loss piece) are UNRATED. That is clearly NOT to be confused with any bond that IS rated (including C, B, BB etc.). The CCC/first loss piece of a structured transaction are essentially the “Equity” piece with absolutely no financially secure guarantee of the return of any of the face amount of principal on the certificates themselves.

Therefore, because there is quite literally almost no hope of principal return, these assets trade at deep, deep discounts to the par face amount. Typically these ‘bonds’ trade in the new issue market (i.e. when the deal is initially structured and sold to investors) at prices nearing 17-25% of par. This practice of pricing is essentially driven by a Timing-And-Severity-Of-Losses Model to calculate a projected yield.

AHR assumes ON THE DAY THEY BUY THESE that $0.00 of principal will be recouped from these assets. So who cares about loss rates of 50% when the assumption from day one is 100% losses on principal?

Why would any ‘bond’ investor ever buy a bond and then immediately write the principal down to $0.00? Seems like financial suicide, right? Here’s why: The loans will pay the stated interest on the 100% face amount of the loans for some period of time prior to experiencing any credit deterioration. The Art & Science of the investment process in these CCC/first loss pieces is accurately estimating how long the loan will pay the full interest before it begins delinquency as well as the depth of the deterioration of the credit in that time frame.

Because the CCC investor is paying 17-25% of par and receiving the ‘bond’ coupon (interest) on the full 100% of the par (face) amount, the income from the bond will eventually return a fat profit….as long as it keeps paying. There is a breakeven point in time at which the initial investment (17-25% of par) is surpassed by the interest payments on the full par amount of the bond and it becomes a profitable investment.

So the key to a successful analysis and therefore a successful investment in these assets is BOTH a TIMING OF LOSSES and LOSS SEVERITY prediction (model). Professionals in the structured markets will recognize this as the SDA (Standard Default Assumption). This is a mechanical-standardized-time-ramp model, which describes the historically experienced credit default curve of loans of similar ilk. This model is coupled by investors with a loss severity model which incorporates various timing of workout/recovery amount/period assumptions in order to present a cash flow projection of a given loan/deal/structure.

Combining these models with the cash flow calculator of a specific loan file (which by the way is ONLY MADE AVAILABLE TO THE FIRST LOSS AND B INVESTORS) in order to predict the cash flow, while accurately adjusting the percentages of these models (up or down) to reflect market conditions is where the Math PhDs make their money.

If losses occur earlier than predicted but are of a mild severity, the investment can workout fine. If the losses occur earlier than anticipated and are of a severity equal to or greater than anticipated, the interest stream will obviously be cut off by the losses (losses take the bond face amount down toward $0 by the amount of the realized loss) and the investment can incur a negative yield.

Remember that AHR ASSUMES THAT ALL PRINCIPAL IS LOST ON DAY ONE AND THAT ONLY THE INTEREST CASH FLOW STREAM FROM THE BONDS WILL CONTRIBUTE TO THE YIELD. The TIMING of the losses IS THE KEY and earlier is clearly worse.

The ‘controlling’ classes are so named because they provide to the investor in the most disadvantageous position with respect to losses (caused by poor performing loans) with the authority to direct the actions of the special servicer. The special servicer is essentially a loss mitigation function provider, which is activated at the command of and for the benefit of the controlling class investor upon the recognition of a loan performance snag.

The CCC holder has the right to control the special servicer to take action on their behalf to mitigate losses. They get to dictate the actions of the special servicer as their agent and for their economic benefit. Thus the class is tagged “Controlling Class”. This bondholder gets to call the shots; they are IN CONTROL. And they are in control because they are taking the most risk…AND THAT IS WHY THESE BONDS ARE NOT RATED. They are typically not registered (i.e. Private securities) and are refered to as NR (non-rated) classes in the prospectus.

The annual and the quarterly reports both take a few paragraphs to explain the loss assumptions on the controlling class interests. I suggest that you take some time and read the explanation of these securities, how they work, what factors affect the return and how they are booked on the balance sheet.

The scenarios you shared could be destructive or they could be irrelevant. You have not provided the depth of analysis necessary to provide much other than speculation. Your points might be accurate and have absolutely no effect OR they might have a big effect. Your post fails in its lack of depth of analysis or even an understanding of what these bonds are that you are primarily discussing. What it seems may have become a little obscure to you is that the return of principal is IRRELEVANT.

The Fitch report could be right and losses could mount to 50% over the life a of a deal AND IT COULD PROVE TO BE A COMPLETELY IRRELEVANT ASPECT OF THE RETURN CALCULATIONS FOR THESE BONDS DEPENDING ON THE TIME FRAME FOR THE LOSSES TO BE REALIZED. It is the timing and severity of losses, which are all important.

Your work is widely read and you are to be commended for the valuable service you provide to the market at large. No offense intended because I do respect your work…but, you are just a little out of your league with your last post. It is therefore likely that the net result is that your readers, whom I dare say have not been exposed to the specific nature of these arcane classes within the structured securities realm, will be frightened not enlightened.

42

August 13, 2008 10:52 PM  
Blogger REIT Wrecks said...

Please see Anthracite Gets Hot Again!" for the response to these comments.

September 3, 2008 3:30 PM  

Post a Comment

Links to this post:

Create a Link

<< Home