Wednesday, October 28, 2009

Pair Trade of the Day: Lockheed vs General Dynamic

The costs of CDS protection on Lockheed and General Dynamic are strongly correlated and currently are equal, indicating that credit markets do not consider one firm as more as risk of a major event than the other.



Their stocks are strongly correlated too (R-squared of 0.848, i.e. a correlation of 92%):



However, today's market (shown as a red star on the plot) is an outlier: GD appears way too rich relative to LMT. I would short GD and get long LMT in proportions guided by the slope of the regression line. The upside is about $15 per LMT share (if GD stays around $65, LMT should revert from ~$70 to a regression line level in the mid $80's), or equivalently $15 to $20 per GD share on the short side (if LMT stays around $70, GD should, according to this linear regression, revert from $65 to its trend line in the high $40's).

Tuesday, October 27, 2009

CSMC 2007-C4

The A1Am note of CSMC 2007-C4 was rated AAA by S&P and Moody's, and still is by the latter:



Almost 9% of its collateral is already delinquent. (BTW, 2600 Michelson is one of Maguire Properties discussed in another post.) An additional 25% is under watch:



If we consider a CDR of 20 (which is in line with the latest remittance reports on ABX collateral of 06 and 07 vintage), Bloomberg's model projects a 49.84% loss on both AM and A1AM notes.

Saturday, October 24, 2009

Lehman Mortgage Trust 2008-4

Today, Standard & Poor's Ratings Services downgraded pieces of a 2008 re-REMIC from AAA to just CCC... The deal is Lehman Mortgage Trust 2008-4, and the class is A2. That their ratings can change so dramatically and so suddenly is shocking (Moody's had cut its ratings on the same paper to just C more than 5 months ago). The press release indicates that "although this performance deterioration is severe, the credit enhancement within LMT 2008-4 is sufficient to maintain the rating on class A1." S&P's current projected loss for the pool already stands at 14.74%. Is there enough subordination under class A1 to make any loss on it highly improbable? I don't think so. And in fact, Class A1 was downgraded by Moody's to just B3 five months ago already...

Let's note also that it took only 11 months for the top notes of the re-REMIC to lose its AAA rating from Moody's and S&P: the deal was issued in June 08, and Moody's downgrade took place on May 15 this year.

Thursday, October 22, 2009

Follow Up on the Bombardier/Textron Pair Trade

On the graph below, the white dot is where the pair of stocks was standing when I posted the regression. Since then, it moved to the red star. The P&L on the short is 0, but the gain on being long Textron is $2 per share.

Tuesday, October 20, 2009

For S&P, Gemstone Still Is A Jewel

Moody's announced today it downgraded four notes by Gemstone CDO, an ABS CDO: notes A-1, A-2, A-3 and B. Shares A-1 were initially rated AAA, and now end up below investment grade. In its press release, Moody's notes that "the ratings of approximately 19% of the underlying assets have been downgraded since Moody's last review of the transaction in March 2009. The trustee reports that the WARF of the portfolio is 1,354 as of September 30, 2009 and also reports defaulted assets in the amount of $48.5 million. Securities rated Caa1 or lower make up approximately 32% of the performing portfolio. The trustee reports that the Class C Overcollateralization Test is currently failing."

Now, if we look at the rating history on this paper, we see S&P hasn't touched the rating since 2004! According to S&P, it is still AAA...

Other Possible Pair Trade: XTO and Anadarko

XTO Energy and Anadarko have had almost the same credit-worthiness for the past two years, as shown by their CDS spreads:



Their shares have been strongly correlated (for the past 2.5 years, at least) -- but they've recently been out of whack.

Ideal Pair Trade Candidate: Berkshire A and B

Here's a perfect candidate for stat arb/pair trading: Share classes A and B of Berkshire. Correlation between the two has never been lower than 95%. Still, there's some volatility between the two. On the screen shot, the top pane shows the prices separately (Class A in white, Class B in amber, at the bottom); the bottom pace shows the ratio between the two. Its oscillations show the opportunities offered by the (bounded) volatility between the two share classes.

Thursday, October 15, 2009

Interest Rate Swaps Through Clearinghouses

This could be game-changing: LCH.Clearnet, Europe’s largest clearinghouse, said it has processed more than $1 trillion in overnight index swaps 11 weeks after starting the service. Interest-rate swaps are the biggest part of the over-the-counter derivatives market.

Follow-Up on the AUD-USD Trade

Since my post predicting the Australian Dollar would keep rising relative to the USD because of interest rates differential, the AUD gained 5.66%.

Wednesday, October 14, 2009

Follow-Up on CHK vs FST

The pair trade suggested the other day has paid off: CHK is slightly lower, giving a small profit on the short side, and FST is up, giving a nicer gain on the long side.

Tuesday, October 13, 2009

"Credit’s Divorce From Stocks Signals End of Rally"??

Interesting "chart of the day" today on Bloomberg: it overlays the S&P500 index with the iTraxx Europe 5Y index (inverted, so that a upward curve also indicates positive performance). The resulting graph shows two curves that look very similar -- until very recently:



The conclusion reached by the strategists who published that graph is that the S&P500 went ahead of itself, as judged by the relative under-performance of the credit index. To quote the press release, "Credit’s Divorce From Stocks Signals End of Rally."

Although I do believe the equity index went ahead of itself, I have issues with the reasoning.

First, we have to be mindful of units: one time series is expressed in S&P points, the other in the inverse of bp; so a more meaningful comparison would be to compare time series of returns (for instance, day-to-day returns).

Second, the time scale is short: their time series begins in early August. If we look at a longer period, we get the graph below. The time frame is still relatively short (we pulled data starting on March 02, 09) but long enough to show that there's been other "credit's divorce from stocks" in recent past:



And third and foremost, how can we tell: Is it S&P500 that's too high, or iTraxx that's too low?? The two indices do have a (relatively short) history of moving in tandem, but there have already been times when the two plots diverged before merging again -- i.e., there have been other "divorces of credit from equity," and they were temporary: None of these instances was followed by an end of the equity rally! On the contrary, the trend has steadily been upward! So I find the technical chartists a bit short on convincing arguments here.

Sunday, October 11, 2009

Dollar Flows at Banks

On Bloomberg news, Oct. 12:

"Central banks flush with record reserves are increasingly snubbing dollars in favor of euros and yen, further pressuring the greenback after its biggest two-quarter rout in almost two decades. Policy makers boosted foreign currency holdings by $413 billion last quarter, the most since at least 2003, to $7.3 trillion, according to data compiled by Bloomberg. Nations reporting currency breakdowns put 63 percent of the new cash into euros and yen in April, May and June, the latest Barclays Capital data show. That’s the highest percentage in any quarter, with more than an $80 billion increase."

Friday, October 9, 2009

Rate on Non-Financial CP: Lowest in 9 Years

Commercial Paper issued by non-financial companies pays the lowest rate in 9 years. However, the rate is just tracking LIBOR, and the spread between the two is simply coming back to normal levels.


(Click to enlarge.)

In the top pane, the CP rate is plotted in white and 3-month LIBOR in amber. The spread is shown in yellow in the bottom pane.

Too High Too Fast

The CS HY index is up 56% from its trough at the end of last year. Of course it's too high too fast... Default rates are not yet as high as feared earlier this year, but HY bonds are in what looks and feels like a bubble. (Sorry, this is as of Wednesday -- didn't have time to upload the chart earlier.)

Tuesday, October 6, 2009

Merrill Lynch Posts $250 Million of Mortgage-Issue Trading Losses

The article below was published in 1985! All the financial fiascos of the last 2 years had already happened then, including huge bank losses (by 1985’s standards…); process failures; rogue traders (or presented as such by their employers); untested financial engineering & lack of proper modeling of newly created securities; sudden markets illiquidity and investment banks stuck with their inventory (think of CDO warehouses); and the question of whether risky investments are appropriate for all investors.... I highlighted a few juicy sentences and added a couple of comments. Enjoy.

Merrill Lynch Posts $250 Million of Mortgage-Issue Trading Losses
by Steve Swartz
The Wall Street Journal, April 30, 1985

NEW YORK Merrill Lynch & Co. said it sustained an estimated $250 million pretax loss, largely because of unauthorized mortgage-securities dealing by a senior trader. The rest of the loss came from “subsequent market volatility" in the securities, the company added. The trader, Merrill executives said, had far exceeded his limits in acquiring mortgages that were packaged into a particularly risky form of securities. The value of some of those securities plunged recently when interest rates went up. Market experts said the trading setback was by far the biggest in recent memory, and was likely the largest in securities industry history. “This is a new world record," said the head of mortgage trading at a rival firm.
[..]

Merrill Lynch said the senior trader, whom it didn't identify, had been fired. However, executives at Merrill Lynch privately identified the trader as Howard A. Rubin.
[..]

The securities said to have caused the problems are created by splitting of the interest payments on the mortgages from the principal and selling each separately. They are known as “interest-only/principal-only" securities, or IOPOs.
[..]

People at Merrill Lynch said Mr. Rubin had accumulated for the firm's account an unusually large portion of the principal-only securities without notifying his superiors. When interest rates went up earlier this month, principal-only securities lost as much as 15% of their value over two weeks, traders said. “He just put them in his drawer," said one senior Merrill Lynch executive. “We didn't know we owned them." Mr. Rubin eventually told his superiors about the unauthorized trades, according to people at Merrill Lynch. Securities firms and banks usually allocate a specific amount of capital that can be risked in each area of trading.

Merrill Lynch's announcement also raised questions about the firm's controls. “This would suggest to me that Merrill's international controls are far less than they should be," said Perrin Long, an analyst at Lipper Analytical Securities Corp. and a frequent critic of Merrill Lynch management. Although Merrill Lynch said that the greater part of the loss stemmed from unauthorized trading, it declined to say how much. However, rival mortgage traders said that Merrill Lynch came to market early this month with more than $900 million in IOPOs at a time when the market was becoming unreceptive to IOPOs. The competitors say Merrill Lynch was largely stuck with the principal portion of the securities in its inventory when the price plummeted.

Merrill Lynch officials say they sold most of the $900 million to customers, except for a portion that was kept in the belief that the price would go up. [my comment: does that imply the other portion was dumped on customers knowing its price would go down??]
[..]

Principal-only mortgage securities are attractive in a market where interest rates are low and mortgage holders have an incentive to pay off early. That produces a quick profit for the security holder, who buys at a discount to principal. When interest rates rise, however, early payments on fixed-rate mortgages become less attractive and the profitability schedule of these mortgage securities is stretched out.

Merrill Lynch officials said Mr. Rubin was suspended a couple of days ago after disclosing his unauthorized trading to his superiors. The officials said Mr. Rubin kept the trades secret for 10 days. They said they were unable to independently monitor Mr. Rubin's trades because he never wrote out tickets for them, a violation of standard procedure.
[..]

The risk of stripped mortgage-backed securities has been a hotly debated topic on Wall Street since the technique was developed last summer. Many investment bankers believed the extraordinary volatility of the security wasn't well understood, and have been predicting that holders could get badly hurt if interest rates moved sharply. However, nobody was expecting a hit as large as the one Merrill took. Although the price of the securities the firm held has dropped dramatically in the last couple of weeks, it didn't fall more than about three points in any one day. Even if Merrill had known it had all the securities, however, selling them when the market began to drop might not have been as easy as it looked. An official at one major firm called the market for stripped securities, of which IOPOs are by far the most popular form, “the most illiquid $11 billion market in the financial world." Merrill Lynch has been selling its position gradually over the last 48 hours, and made its announcement after most of it was sold, market sources said. The firm was extraordinarily active in the Treasury market Tuesday, they added, leading to speculation it was trying to hedge what it hadn't yet sold.

Merrill Lynch officials won't say where they sold the securities, but the firm is rumored to have disposed of at least part through its retail system. Merrill Lynch has sold portions of other stripped offerings retail, but the question of whether retail investors understand a product that complex is a much-debated one.

Monday, October 5, 2009

The Curious Case of Lincoln Avenue -- A Structured Credit Medley

If you find this post complicated: it's the whole point. Thanks to the Lincoln Avenue deal, we are about to encounter in this post almost every contraption created in the gogo years of credit. In fact, it is so complicated that I doubt any rating agency or investment bank has the models and the infrastructure to properly analyze such deals. Of course, we could also say that confusing structured credit investors is the whole point as well!

So judge by yourself. And BTW, everything here is public data, you just need to google for "Lincoln Avenue CDO" and do some research on the Bloomberg terminal.

The Lincoln Avenue ABS CDO, issued in July 2006, has a rather eclectic collection of paper as collateral: other collateralized debt obligations (CDOs), some U.S. commercial mortgage-backed securities, and U.S. residential mortgage-backed securities. So it is in part a CDO-squared, a CDO of CDO.

The maturity of Lincoln Ave is nothing less than 40 years. Its senior tranche, the Class A-1 notes, was originally rated AAA by S&P and Aaa by Moody's; it had $1.094bn of notional. The next notes in subordination order is the A-2, with $77mm of face value; then $26mm of Class B notes, $21mm of Class C, and $19mm of Class D, "equity" or "first-loss" notes.

According to totalsecuritization.com, Lincoln Avenue ABS CDO had an "event of default" (EOD) as early as Sept 12, 2008. In fact, a majority of Lincoln Avenue's collateral has today a rating of just C (i.e., junk) by DBRS.

So not surprisingly, everyone seems to be trying to get rid of this stuff.

First, in December 2007, Barclay's repackaged $7.5mm of face value of Class C notes (ISIN: USG5490EAD07) of Lincoln Avenue, together with 9 other CDO securities, into $70MM "Principal Protected CDO Portfolio-Linked Notes" and into a $20MM so-called "Credit Linked Note Programme." Looking at the OM, it seems that these securities had ISIN XS0336216634, and were marketed mostly in Europe.

Then, recently, even the holder of the top of the capital structure of the Lincoln Avenue CDO, the holders of Class A-1 Notes, seems to try to swap the toxic asset away: DBRS announced today that it "has assigned a rating of BBB (low) to the total return swap (TRS) referencing Lincoln Avenue ABS CDO, Ltd.'s Class A-1 Notes, pursuant to the transaction documents dated April 28, 2009, with a current notional amount of $82,860,028." As an aside, DBRS indicates that "this rating is being provided at the request of the majority Class A-1 Noteholder." I.e., their client is not the CDO, but indeed an investor in the CDO's top tranche -- the holder trying to swap its position away.

Now, listen to this: "The TRS benefits from approximately 92.2% subordination" from the Class A-1 Notes and from a "commitment letter" (a credit enhancement). So, even with 92.2% of subordination and the credit enhancement, the TRS could barely get an investment-grade rating? Personally, I consider ratings in structured finance have lost all and any credibility. (In fact, I find suspicious the sheer fact that the TRS barely passed IG.) But the low rating given the subordination and the commitment letter shows how toxic the collateral and the structure is. In other words, if the TRS on the A-1 notes are BBBL-quality, the Barclay's note probably couldn't receive a rating at all.

Friday, October 2, 2009

Possible Stat Arb on Bombardier and Textron

Bombardier and Textron are in the same industry, and have almost identical creditworthiness according to the CDS market:



Their shares, too, have moved in tandem. The linear regression of their share prices shows a correlation of 89%:



Still, today's point if far from the regression line, suggesting that the shares of Bombardier are rich relative to those of Textron. Specifically, it suggests being long 0.129 shares of Textron for each share of Bombardier shorted. According to this model, if Textron's shares don't move, Bombardier's should retrace from about $4.5 to about $4; if Bombardier's shares are unchanged, Textron's may revert to its trend line, from about $18 to about $22. Any other scenario reverting to the red regression line is of course possible as well, and would also generate a profit.

Thursday, October 1, 2009

Commercial Real Estate: Delinquencies by Property Type

The percentage of commercial real estate loans that are 60 days or more delinquent in the hospitality industry shot up in September:

(Source: Bloomberg, dataforthoughts. Click to enlarge.)

In absolute $ of loans delinquent, however, the hospitality industry is on par with multi-family and retail commercial real estate loans:

(Source: Bloomberg, dataforthoughts. Click to enlarge.)

Note that the total amount of loans late 60 days or more has reached a staggering $25bn.

What Is Wrong With Re-Securitized MBSs

To see what is wrong with re-securitized MBS, we need look no further than Credit Suisse Mortgage Capital's recent CSMC 2009-10R deal.

Its collateral is 2242 loans, all ARM. 29.7% are in California, 5.8% were taken for investment. 99.6% of them were originated in H2'06 and Q1'07, at the trough of mortgage origination's standards. The performance of its collateral is catastrophic:

  • Loans delinquent for 60 days or more represent 57.34% of collateral
  • Loans with a loan-to-value between 80% and 139% are 57.96% -- not surprisingly, almost the same number as the fraction of households who stopped making payments! (Remember, an LTV above 100% means that you owe the bank more than what your place is worth.)
  • 91.8% had FICO score at or below 700, 68% below 651.
  • 11.1% of these loans currently pay usury interest rates of 10% or more! (Their ARM rates probably reset.)
But the worst may be yet to come for this security: 10% of the loans faced resets on their mortgages last month (September); almost 8% will be reset in December, then 19.7% more in Jan, and 23% in February! While again already 57.34% of the loans are late on their payments! I have no idea what this paper can be traded at, but probably not much.

The Rates Banks Charge vs. Their Funding Cost -- Continued

We already noticed examples where banks are charging their customers much more than their funding costs. Here's another example: the spread between the national average of rates charged for a home equity line of credit (HELOC) and 3-month LIBOR.


(Click to enlarge)

The screen shot is a bit busy, but the top pane shows the average HELOC rate, in amber, and LIBOR, in white. The spread between the two has been about constant until 2008, when it started to gap out. This can equivalently be seen in the bottom pane, which plots the difference between the two: that difference has risen steadily since late 2007, except for a brief period in 2008 due not to the generosity of banks but to a sudden spike in LIBOR.

Mini Panic Today, and "Riskless" Assets Rallied

The yield on the 30-year Treasury bond fell 11bp to 3.94%, the lowest level since April 29. The 10-year note yield dropped 3.18 percent, its lowest level since May 21. Yields on Fannie Mae’s current-coupon 30-year fixed-rate mortgage bonds tumbled 14bp to 4.09%, the lowest since May
21 as well.