It is hard to know where to start when trying to analyze the current problems in the subprime mortgage markets, there are just so many points that beg to be made. So, not necessarily in order of importance, let's look at a few items.
First, as Dennis Gartman so frequently states, there is never just one c/ockroach. If you see one, you know there are more in the wall. Bear Stearns is just the first. They may be the canary in the coal mine which warns us of more problems to come.
Will the problem in the subprime market spread to other areas of the debt market? The answer depends on what you mean by spread.
A quick review. Subprime mortgages are bought by investment banks and re-packaged as Residential Mortgage Backed Securities (RMBS). These securities, which can contain thousands of loans, are split up into varying groups, or tranches, and each tranche is given a rating by one or more rating agencies, typically Moody's, S&P, or Fitch. The highest-rated levels get the first monies paid back into the fund, the next level is next in line, on down to the bottom rung which is last in line to get repaid and first in line to get shot if there are losses in the mortgage revolution. The bottom rung is called the equity tranche, also called toxic waste by those in the industry.
Let's look at a chart of an index of BBB-rated tranches, which is typically the mid-level tranche. They have dropped 42% since January. This is not a pretty chart, but it tells the story. (www.markit.com) Some BBB tranches from 2006 are down as much as 60% already.
can't reproduce the graph here .. dub
If you looked at a chart of the high-yield bond index, it is only off a point or so, which suggests that the problem is not spreading into other major bond categories, or at least not yet. In fact, we should take comfort in this, but looking into the index does raise an eyebrow. The index is composed of 100 high-yield bonds. My assistant, Majed, looked at all 100 issues and found that roughly 60% had negative free cash flow per share. While that is not unusual for growth companies, a lot of those names are old-line corporations, with large debt-to-equity ratios. This is definitely not a low-risk market, but it only pays 320 basis points over US government debt. You can look at the various stocks in the index for yourself at http://www.markit.com/information/affiliations/cdx.
This week Bear Stearns felt it necessary to inject $1.6 billion in loans into two of its hedge funds that deal in the higher-risk portion of the subprime mortgage market. These funds are down 23% at least. Merrill Lynch moved to seize some of the assets it had as collateral for loans and put them to the market. Prices offered on the better grades were a shock, and there was no bid on most of the securities. Not low bids, but no bids. Merrill decided to not press the sale.
Mark to Model or Mark to Market?
Here's the problem. There is not an active market in these securities, so the various funds and banks use a "model" to decide what price they should use when toting up their assets. But if these securities are actually sold, then now you have a market price, and you have to mark to market rather than use the price based on your model.
As an illustration, let's say an institution had to sell a $1 million chunk of a $20 million BBB tranche. Before that sale, other funds and institutions owning that same tranche could price it at the model price, i.e., what their accountants or bankers said they were worth. Nothing sinister here, just people making their best guesses. Typically, a fund will take three such guesses from outside firms and go with the average, but each fund or bank will have its own rules.
But if that institution sells their tranche for a 20% loss, then now there is a market price established for everyone who owns that same tranche. Losses passed all around. As an aside, a group of hedge funds has written the SEC asking the agency to be on the lookout for investment banks who would buy bad loans in order to keep them from defaulting and leading to losses in the banks' derivative portfolios. As you might imagine, these funds are short these derivatives and want to see large losses at the banks.
But this stuff is rated, right? It should all sort out. Not to panic.
Let's go to the latest issue of Bloomberg Markets magazine. In an explosive article called "The Rating Charade" (kudos to Richard Tomlinson and David Evans for a well-done piece), they discuss a Collateralized Debt Obligation (CDO) issued by Credit Suisse in 2000 (well before the current subprime crisis!) and composed of mortgages, loans, and other debt. Cutting to the chase, all three rating agencies rated the five tranches. 95% was rated investment grade. The unrated tranches went completely bust, the two mezzanine tranches (rated BBB and A!) lost everything as well. Those losses combined for $47 million.
The AAA tranche lost almost 25%, although those investors did have insurance, so they got their money back, as MBIA took the $73 million loss in the AAA tranche. A total of $120 million in losses in a $340 million CDO, with 95% of it rated as investment grade. Ouch.
As I have written since the fall of last year, these bonds are bought by various institutions because of the ratings from the credit agencies. And it is not just European and Asian institutions. Let's make a short list of some US pension funds that buy the equity (toxic waste) portion of CDOs: The New Mexico Investment Council ($222 million and another authorized $300 million for 3% of its total fund), the General Retirement System of Detroit ($38.8 million), the Teachers Retirement System of Texas ($62.8 million), Calpers ... the list is evidently long. 7% of all the equity tranches sold in the US in the past decade were purchased by US pension funds, endowments,s and religious organizations.
So, why did they buy them? Let's be clear. They were looking for higher returns, and they took comfort in the ratings. Some of these CDO-rated tranches paid 10% over LIBOR. That is a huge difference and roughly double the return for similarly rated debt. The equity portions offered as much as 20%. Bear Stearns said in a marketing meeting that "The outside agencies that oversee these structures are the rating agencies." I will bet you a few dollars against a donut that similar statements were made by other investment banks.
But what do the rating agencies say? Fitch: "It's not accurate. We don't provide any oversight." Moody's: "It's a common misperception. All we're providing is a credit assessment and comments." S&P: "We disagree. We rate the transactions that issuers bring to us based on our published criteria."
(A lot of these "investments" were made as principle-protected notes. That means, as an example, that an investment bank would couple a 12-year zero-coupon bond with the equity CDO tranche. The pension fund will not lose money if the equity tranche blows up, but will make nothing on its asset for 12 years. Not a good deal. But don't get me started on principle-protected notes. That's one of my pet peeves.)
The rating agencies all have disclaimers which read something like this: "You should not make an investment based on our ratings." That is of course laughable, as the only reason that anyone buys these investments is the ratings. It is going to be interesting to see what the courts say, as you can be sure of one thing: this is going to end up in court.
How large are the losses we are talking about? They could be quite large. From the Bloomberg article:
"As foreclosures increase, the subprime-backed securities in CDOs begin to crumble. Subprime mortgage securities make up about $100 billion of the $375 billion of CDOs sold in the U.S. in 2006, according to data from Moody's and Morgan Stanley. Seventy-five percent of global CDO sales are in the U.S. Moody's reported in March that about half of the CDOs sold in the U.S. last year contained subprime debt. On average, 45 percent of the contents of those CDOs consisted of subprime home loans, Moody's said.
"In a certain class of CDOs, the concentration of subprime is even higher. S&P and Fitch estimate that subprime mortgage securities make up more than 70 percent of the debt in so-called mezzanine asset-backed CDOs, a type of CDO that repackages bonds, mostly mortgage debt, with low credit ratings. Investors bought $59.5 billion of these CDOs in 2006, according to Morgan Stanley. On average, as with all CDOs, more than 90 percent of the value in them is rated investment grade."
The Center for Responsible Lending estimates that 2.2 million borrowers who got subprime loans since 1998 either have lost or will lose their homes through foreclosure over the next few years. This includes one of every five borrowers who got subprime loans in 2005-06, a default rate unmatched in the history of the modern mortgage market.
You can go to your Bloomberg quote machine and pull up residential subprime structured finance deals. What you find is one RMBS that was issued in 2006 that already has over 54% of its loans more than 60 days delinquent and 17% of them in foreclosure. Think the buyers of that equity tranche stand a snowball's chance of getting anything?
Has this security been re-rated? No, because the ratings agencies say they cannot re-rate something until they know for certain there are losses. They can't act on suspicion. However, I do remember them putting out warning notices for various bonds and corporate offerings prior to re-rating. I would think those are coming.
$250 Billion in Subprime Losses?
The problem is that if these offerings lose their investment grades, many institutions will be forced to sell, as they are limited by their charters to only invest in rated paper. But who will buy until the smoke clears? It will get ugly. This will end in tears.
"The highest default rates on home loans in a decade have reduced prices of some bonds backed by mortgages to people with poor or limited credit by more than 50 cents on the dollar and forced New York-based Bear Stearns Cos. to offer $3.2 billion to bail out a money-losing hedge fund. Almost 65 percent of the bonds in indexes that track subprime mortgage debt don't meet the ratings criteria in place when they were sold, according to data compiled by Bloomberg.
"That may just be the beginning. Downgrades by S&P, Moody's and Fitch would force hundreds of investors to sell holdings, roiling the $800 billion market for securities backed by subprime mortgages and $1 trillion of collateralized debt obligations, the fastest growing part of the financial markets." (Bloomberg)
How much could the losses be? It depends on who you ask, but estimates of $150 billion or more are quite common. Institutional Risk Analytics, a Hawthorne, California-based company that writes computer programs for accounting firms, says 25% of the face value of CDOs is in jeopardy, or $250 billion. But no one really knows. It is all guesswork, except that everyone seems to agree it will be large.
The rating agencies use various models to come up with their ratings. But a recent study suggests that if you "take the 300 bonds that are used in the ABX indexes, the benchmarks for the subprime mortgage debt market, 190 fail to meet the credit support standard, according to data released in May by trustees responsible for funneling interest payments to debt investors.
"Most of those, representing about $200 billion, are rated below AAA. Some contain so many defaulted loans that the credit support is outweighed by potential losses. Fifty of the 60 A rated bonds fail the criteria, as do 22 of the 60 AA rated bonds and three of the 60 AAA bonds.
"All but five of 120 securities in BBB or BBB-rated portions of the mortgage-backed securities would have failed S&P's criteria, according to data compiled by Bloomberg."
Bottom line: no one knows how large the losses will be. Right now the "hope" is that they are spread out among hundreds or thousands of various funds, institutions, and agencies, and thus there will be no major failures or systemic risk.
So while it may not directly affect other types of bond offerings, it is definitely going to affect the appetite for risk, which in the long run is a good thing, as deals were getting done that were questionable. Nothing helps concentrate the mind like the prospect of a hanging, said Judge Roy Bean. And nothing helps you focus on risk like a serious loss in your portfolio.
We will continue this theme next week, as this letter is getting long and it is getting late. There is obviously a whole lot more to say, but this is one game that is going to take a long time to play out. It won't be like breaking an egg. It will be more like watching ice cream melt. .....
dub
To subscribe to (the free) John Mauldin's E-Letter go to http://www.frontlinethoughts.com/subscribe.asp
(20min delay)