Thursday, June 14, 2007

So... you got your reward and you're just leaving then?

In a move that surprises most Australian aboriginals and some (but not all) long-term coma victims, a high-profile hedge fund focused in the sub-prime mortgage area is liquidating. Bear Stearns put about $3.8 billion in bonds out for the bid today, a little bigger than your average BWIC. According to various people I've talked to, the bonds are predominantly AAA floaters, so despite the sub-prime stigma, these are pretty clean bonds which should garner demand from various types of investors.

I understand that Bear has riskier assets in the fund, called the Bear Stearns High-Grade Structured Credit Strategies Enhanced Leverage Fund. When these are to be sold, I don't know. Some of the CDO equity the fund was holding has already been liquidated.

This is classic Wall Street. Hang around while the product is hot, create as much in fees as you can, invent new structures or create your own investing vehicles if you have to, and cash your chips in at the first sign of trouble. You can decry it all you want, and I'm sure people like Tanta will, but I'm pretty sure that's how Wall Street has been operating since the days when there was actually a wall to keep the injuns out.

Anyway, let's talk about something more interesting.

Any time a hedge fund is liquidating, people start to think about Long Term Capital Management, and how much of a mess that created for the market. The Bear fund, which has about $500 million in capital, just isn't large enough to cause an LTCM type disaster on its own. But I think there are some real parallels to LTCM and what's going on in sub-prime today.

Long Term's primary bet was that markets would always be fairly efficient. I remember reading a quote from one of their partners, I think it was Merton, who described their strategy as picking up nickels the market was leaving behind. Their models consistently lead them to believe that owning a wide variety of high-risk assets would result in a low risk portfolio, which could then be levered 100-1.

In the fall of 1998, due in part to the Russian debt crisis, the market's risk appetite diminished markedly. This resulted in risk assets in all corners of the world to fall in value, from Taiwan to Brazil. Suddenly, assets which would have seemed to have no fundamental correlation were displaying a very high correlation. LTCM was suffering losses on everything at once, and their leverage was so high that the whole thing fell apart rapidly.

There are many parallels in various sub-prime strategies, be it CDO's, hedge funds, or REITs. First, the leverage is usually very high, because the spreads on most sub-prime securities just aren't that great. The majority of sub-prime backed bonds in the market today are investment-grade. In order to make the sexy returns the hedge fund crowd is looking for, you've got to leverage this stuff. Second, the correlation in sub-prime performance is converging toward 1. Whereas we might have once thought that the performance of a random loan in Ohio versus one in California had a low correlation, the pervasiveness of weak underwriting standards has changed that. In addition, the pervasiveness of adjustable-rate mortgages and the extreme upward move in interest rates is also creating a large degree of stress on a high percentage of loans, regardless of geographics or other factors.

As readers of this blog no doubt know, the unwinding of LTCM was extremely painful for the market. The types of bonds they owned were already trading weak, and their positions were leaking into the street. So the street was just killing their positions. When it came time to actually liquidate, the bids were scarce. The contagion was significant, if short-lived.

So are there any sub-prime funds large enough to cause a LTCM-type result? Maybe not. But its really not that hard to imagine a major contagion scenario:

1) A large number of investors in higher quality CDO tranches (A and AA) are burned by sub-prime defaults.

2) This causes a re-pricing of CDO spreads, and causes a drastic slow-down in deal flow.

3) In turn, this eliminates the "CDO Put" in the credit market. This is where any widening of credit spreads made forming new CDO's that much more attractive, thus creating a back-stop for spreads generally. If the CDO market disappears, even temporarily, this "put" is gone.

In that scenario, we finally see the widening of credit spreads everyone's been waiting for. And given that real money has been so reluctant to over-weight high-yield, the widening could be quite dramatic.

For the moment, I don't see this actually happening without the sub-prime default rate rising even higher than most expect. It takes a lot for A or AA-investors to start taking losses. And as long as the A and AA investors are mostly untouched, then a high level of defaults really just validates CDO technology rather than cause the market to question it. Watch the delinquency reports, though.


Anonymous said...

Wouldn't 1(a) be downgrades by the rating agencies?

James said...

Awesome. No one loses money and credit never tightens. Hello 5 dollar gas and a revolution.

Anonymous said...

What an extraordinarily informative post! I want to emphasize how useful this post is. Yes, I admit, it supports some of my intuitive prejudices, but my intuitive prejudices are just that: zip.

What a post!

Keep up the good work.

Anonymous said...

You know, I think it bears (no pun) mentioning that in the ordinary scheme (once again, no pun) of things, the hypothetical average man would probably consider that astronomical amounts of leverage increase risk.

Yet, as you and innumerable other writers make clear, in the current environment, astronomical leverage is supposedly aligned with decreased risk.

And pigs can fly.

Anonymous said...

You know, I've been pondering lately about risk mitigation through aggregation. ISTM that the more people rely on this, the faster we rune into difficulties. I think that housing (really mortgage) bubble CREATED corelations between crappy mortgages in Ohio and California that didn't exist before. If you create a system that works fine unless everything goes bad at once, you ensure that when it breaks, everything will go catastrophely bad at once.

Jim A.

Accrued Interest said...

Anon #1: If you own a AAA asset which gets downgraded to BBB, you've suffered pain and are probably turned off to the product. CDO's depend entirely on the ability to sell a large AAA tranche at a very tight spread.

Jim A: Its a classic question though: yes, if things go very wrong, its a huge problem. But will things ever go THAT wrong? Isn't it similar to the Fannie/Freddie argument? Or municipal pension obs?

Anonymous said...

What's happening now?

Things are going sideways.

Where are we headed?

Anonymous said...

Question: Repayment of mortgage-backed securities is typically guaranteed by the issuer. Wouldn't that prevent this sort of scenario from taking place?

Anonymous said...

Anonymous said...
Question: Repayment of mortgage-backed securities is typically guaranteed by the issuer. Wouldn't that prevent this sort of scenario from taking place?

Agency MBS from Fannie Mae/Freddie Mac is guaranteed, not subprime MBS. The banks don't guarantee anything they issue.

Accrued Interest said...

We are headed to the beach for the week. Or make that I am headed for the beach.

And indeed, only agency MBS are guarunteed.

Anonymous said...

Isn't the ability to place BBB and subordinate securities more important to the "machine" than placing the large AAA's? Even though the BBB tranches are smaller, the appetite of buyers is significantly lower. Typically, the AAA's fly out the door and, especially in the current environment, the BBB's are extrememly difficult to get done. I can't imagine going to your chief investment officer or head trader w/ a recommendation to buy another slug of BBB when the existing ones you hold are marking in the 60's. Unless of course you don't need your job and want to play golf for the rest of the year...

Accrued Interest said...

It depends on the sizes of the tranches. In high quality deals, which most sub-prime RMBS deals are, the AAA tranche is 90% or more of the deal. The BBB tranche might only be 2-3%. So even if you have to cheapen up the BBB by 100bps to make the deal work, it winds up having a very small impact on total cost of funds. However, if you have to cheapen the AAA by 10-20 bps, now the economics of the deal have completely changed.

Anonymous said...

Can someone explain the "CDO Put"? From the post: This is where any widening of credit spreads made forming new CDO's that much more attractive, thus creating a back-stop for spreads generally.

If CDO spreads widen, then wouldn't the banks have to offer higher coupons on the tranches of new CDOs that they underwrite? Since the goal for the bank and CDO issuer is to minimize cost of funding in order to maximize return on equity, how is this a good thing? Yes, investors would like the extra yield but wouldn't fewer CDO deals be economical since a new CDO is underwritten only if a desirable return on equity can be achieved?

Thanks in advance.

Accrued Interest said...

See 6/22 post.

Anonymous said...


You've got excellent information about auctions on your site. As I was scanning net for Auction resources I found your site and another site about auctions, I believe will interest you.

Check it out.