The lessons of CELF

By Felix Salmon
November 26, 2010
Jesse Eisinger has the story of CELF, which has some interesting implications. Essentially, Goldman Sachs took a bunch of leveraged loans it had lying around on its balance sheet, and bundled them into a CLO called CELF which it sold in July 2008.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Jesse Eisinger has the story of CELF, which has some interesting implications. Essentially, Goldman Sachs took a bunch of leveraged loans it had lying around on its balance sheet, and bundled them into a CLO called CELF which it sold in July 2008.

The transaction was clearly profitable for Goldman — if it wasn’t, the bank wouldn’t have done it. And like all CLOs, the reason was that there was insatiable investor demand for triple-A-rated securities. As a result, by bundling up a bunch of loans and tranching them so that a triple-A-rated security fell out the other side, Goldman could make money: demand for AAA debt was much greater than demand for leveraged loans, so turning the latter into the former was profitable.

Then, this year, Goldman unwound the deal. It bought back those AAA-rated loans — I’m hearing at about 96 cents on the dollar — and bought a bunch of the equity in the deal as well, enough to bring its equity stake to something over 50%. With control of the CLO, Goldman then decided to liquidate it entirely, breaking it up into its constituent parts and selling off those loans in the secondary market.

This deal, too, was profitable for Goldman — for exactly the opposite reason that the CLO was. Today, there’s a lot of demand for high-yielding loans, or high-yielding anything, really. Meanwhile, there’s no appetite at all for structured products carrying AAA credit ratings which no one believes. So Goldman can make money by turning out-of-favor structured product into highly-desirable loans.

Eisinger’s point here is that Goldman reckons it can do this kind of thing — making money by structuring and unstructuring complex financial products — without falling foul of the Volcker Rule: at each step along the way, it can claim to be acting in its clients’ best interest. He’s right about that, and he’s also right that if the Volcker rule can’t stop this kind of activity, it’s likely to prove pretty toothless.

But we can also draw another lesson from this story: that securitization is still pretty dead. So long as investors prefer plain-vanilla loans to collateralized loan obligations, the securitization market — which bankers and politicians both have said is crucial to the efficient functioning of the economy — will remain moribund. Let’s hope they’re wrong, and that securitization isn’t all that necessary for a vibrant economy after all. Maybe it’s mainly just good in terms of making money for investment banks.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

Well hey, if the financial sector is 10% of GDP or something like it, and securization can boost profits 50-100%. then yes it is a big part of the economy. Just not in the way most would think…

Posted by CDN_finance | Report as abusive

Interesting find, great post!

Posted by DanHess | Report as abusive

Ok missing the outrage here… So at the request of a client it bundles up some loans and sells it to some other clients who are demanding exposure to those sort of loans. The client then changes its mind and wants out. GS then buys back the exposure and sells it off later. In what way is this not client driven? Would you have been happy if GS had refused to buy back the AAA tranche? Does Mr Eisinger know of a single market making transaction where there is NO risk for the middleman?

Posted by Danny_Black | Report as abusive

Typically clueless piece by Jesse Eisenger.
Control of CLOs (unlike CDOs) routinely pass to equity holders after a short call period. The equity holders then have the option to redeem the CLO (at par):
Typical clause:
“Optional redemption
The issuer may redeem the notes may early at the option of the subordinated noteholders after the end of the noncall period or earlier if certain tax events have occurred. This optional redemption may occur only if the proceeds from liquidating the collateral are sufficient to redeem the rated notes at par, plus accrued interest.”
This has nothing to do with alleged preferences for AAA securities or “non appetite at all for structured products carrying AAA credit ratings” it has to do with moves in credit quality of the underlying loans and simple maths regarding credit spreads.

Posted by alea | Report as abusive

@Danny_Black – yes, since they did the transaction this way, Felix/Jesse/NYTimes et al will be outraged that they made money by deploying their balance sheet to unwind the trade. On the flip side, had they refused to buy the AAA tranche, it would be the auction rate securities debacle all over again and we would be hearing about their lack of client commitment and client service. Darned if you of, darned if you don’t in this new world of “journalism” (e.g. irresponsible pageview and click generation by yelping “GOLDMAN SACHS EVIL!” at every possible moment).

Posted by west-coasting | Report as abusive

west-coasting, lack of trust does that to people…

Posted by hsvkitty | Report as abusive

> Today, there’s a lot of demand for high-yielding loans, or high-yielding anything, really.

Incidentally, does this make other people as nervous as it makes me? I’ve been telling my mother-in-law that if she’s as risk-averse as I think she is, a 1% return — or frankly a 0% return — really isn’t so bad right now, and that the mindset that one is “entitled” (or something) to more is likely to cause pain at some point.

Posted by dWj | Report as abusive

Agreed 100%, dWj. We’re setting people up to be very badly burnt.

As best I can tell right now, the LEAST risky securities over a ten year horizon are top-quality dividend yielding stocks. This isn’t to say that they are any less risky than usual, just that the normally risk-averse investments are even worse.

Disclaimer: presently allocated ~75% stocks (predominantly blue chip dividend stocks), ~15% TIPS and guaranteed-principal annuities, ~10% cash/CDs with 2 year maturity or less. Conventional bonds are very scary to me at current prices.

Posted by TFF | Report as abusive

hsvkitty, so does lazy churnalism.

Posted by Danny_Black | Report as abusive

Agreed. Lazy at best (which is what I’d say for this piece by Jesse/Felix). Irresponsible and malicious at worst, which is where a lot of recent NYTimes reporting is going (if I want to read ZeroHedge’s conspiracy ramblings, I can read them there or in the HuffPo, not in America’s newspaper of record).

Posted by west-coasting | Report as abusive

Goldman can only “claim to be acting in its clients’ best interest” to people who don’t know what’s really going on.

Goldman Sachs had no client in the CELF trades, from start to finish, other than itself. 010/11/client-is-not-counterparty.html

Posted by EpicureanDeal | Report as abusive

So in other words, Epicurean, Goldman was making the deal in its own CELF interest? I was going to make another bad joke to match your ending, but I will keep it clean.

Posted by hsvkitty | Report as abusive

At the risk of losing you some sleep at night, Mr Dealmaker, I respectfully disagree.

1) Normally in underwriting the bank is taking a hopefully small risk that between closing the price and actually selling on the securities the market does not change too substantially. It does and has happened that the underwriting banks lose substantial amounts of money even in vanilla equity underwritings. I also disagree that there are terribly many reputational issues with bringing crap to the market, as long as it is crap du jour such as Internet stocks or dare I say it CDOs and CLOs.

2) With the original CLO, I suspect GS bought the loans off some clients packaged them up and sold them to other clients. Yes for a period of time those loans were warehoused but I highly doubt GS bought the loans because they thought qua loans there were a great investment they wanted to take a view on, it was just balance sheet rental for another client or clients. Classic middleman. Just like the pharmacist bought the box of Preparation H, held it for a short time till another client – you – came in and paid a slightly higher price. It was hardly a principal investment by the pharmacist.

3) I also suspect that in buying back the AAA tranche that it was not quite as arms length as you suggest. I suspect GS wanted to call, needed to buy out the AAA tranche and to make sure it seemed pucker got a third-party to run an auction which, within reason, GS would always have won. The third-party and the auction was so GS and the client could wave a piece of paper around saying look it was kosher.

The only vaguely nefarious way this can be viewed is possibly the reality was that GS in 2008 was trying to shrink its balance sheet, offered a meaty coupon to clients with the nod-nod-wink-wink we will call the notes when all this blows over. Of course the journalist in this case didn’t bother chasing up the interesting bit because he is more concerned in the moral outrage that a firm isn’t aiming to provide a service at a loss – something the NYT has been doing well for a while now.

Posted by Danny_Black | Report as abusive