Where Does Collateral Come From, Daddy?
Last Friday I was part of a group of 8 economists who made presentations to the Board of Governors of the Federal Reserve. Once or twice a year the Board invites academics to discuss a policy-relevant economic issue. The subject of Friday’s session was Network Economics, and Alistair Milne and I spoke about network issues in clearing and trading.
The Parthian shot in my presentation was that it was essential not to ignore the unintended effects of derivatives regulation. In particular, clearing mandates, capital and collateral rules, and the like are all intended to reduce the amount of leverage in the financial system. But even if you require extensive collateral against derivatives trades, market participants can offset the reduction in leverage inherent in derivatives trades by using the freed up debt capacity to add leverage in other forms. What goes out through the front door can return through the back door. This can be particularly dangerous if you think you’ve solved the problem when Mr. Leverage walks out the front and ignore that he’s snuck back in.
Put differently, it is very hard to reduce leverage in the system as a whole by regulating just a piece of firms’ capital structure.
Another example. Reducing derivatives counterparty credit risk through clearing and netting and collateralization works by increasing the priority of derivatives in bankruptcy. But if you move derivatives towards the head of the line, somebody else ends up further back. Like unsecured creditors, including money market funds who buy short term bank debt. They can be systemically important, too, meaning that the regulations have primarily shifted the locus of systemic risk, and not really reduced it.
Focusing on the collateral issue, the numbers here are huge. Tabb Group estimates that interest rate derivative collateral alone will total $1.4 trillion. Credit, equity, and commodity derivatives-related collateral will likely add hundreds of billions to the total: although these markets are smaller, they are riskier, requiring more collateral per dollar of notional.
So where is all this collateral going to come from? I’ve written about collateral management services, by which banks will effectively lend cash that can be used for margins, with the loans collateralized by assets that CCPs will not accept directly. In other words, Mr. Leverage is a Back Door Man.
CME Group’s Craig Donohue talked about the need for such services:
“There will be a need and a value for the collateral transformation services,” Donohue, chief executive officer of CME said today during a panel discussion at the Futures Industry Association annual conference in Chicago.
The types of accepted collateral may need to be expanded as thousands of new users seek to back trades. While cash and government securities will always be the least costly way to pay for margin, some who want to pledge “three chickens and an old Ford” will find it more expensive to have non-standard collateral transformed into acceptable form, said Jeffrey Sprecher, chief executive officer of Atlanta-based Intercontinental Exchange Inc.
The demand has created a “cottage industry” of firms and units within banks that will charge a fee to hold securities or other assets for investors in return for acceptable margin, said Richard Prager, managing director and head of global trading for BlackRock Inc. (BLK)
Cottage industry. Somehow this brings to mind all the stories about shadow banking system in China that mushroomed after the government tried to crack down on bank lending.
And CCPs are looking to expand the assets they are willing to take as collateral. This exposes the CCPs to higher risks, including valuation risks. As an example, LCH.Clearnet (which has already begun accepting gold) is going to accept some corporate debt as collateral (no link yet, but it’s from a Matt Leising Bloomberg story that Matt emailed to me):
LCH.Clearnet Ltd., the biggest interest-rate swap clearinghouse, is considering accepting corporate bonds for the first time to meet an anticipated surge in demand for a broader array of collateral in the $601 trillion over-the-counter derivatives market.
“We want to be able to have more clients come to clearing,” Floyd Converse, head of U.S. sales and marketing for London-based LCH.Clearnet, said in an interview in Chicago this week at the Futures Industry Association annual conference. The world’s biggest swaps clearinghouses are expanding the
types of securities they accept as regulations force thousands of new users in the U.S. and Europe to use clearing. Among them, CME Group Inc., the world’s largest futures exchange, is broadening the variety of company debt it accepts as margin for its customers.
(Quick thoughts off the top. “Investment grade” doesn’t really give the warm and fuzzies, does it? Will corporate debt be accepted as collateral against CDS trades? I know LCH.Clearnet doesn’t clear CDS, but there is a reason to wonder whether competitive pressure would lead to such an outcome. Wouldn’t this expose a CCP to substantial credit correlation risk, which is devilishly difficult to quantify and manage?)
And it is becoming far more widely accepted that CCPs will require central bank liquidity during periods of extreme systemic stress.
All of these stories suggest that mandates are first and foremost just leading to a reshuffling of leverage and risk. There is room for deep skepticism that this reshuffling will reduce systemic risks.
Indeed, if Gary Gorton is right, and shadow banking was first and foremost a mechanism designed to create informationally insensitive debt that could be posted as collateral, the potential increase in demand for collateral could catalyze a similar process. This process is systemically risky because it increases fragility and creates substantial tail risks, as this new Gorton paper discusses.
There is also reason to doubt that this is a good use of capital, and could lead to lower growth. Just what we need. The cost-benefit trade-off appears particularly unfavorable with respect to collateral posting by long-only buy side investors (e.g., pension funds) and “end users” who do not pose substantial systemic risks.
So, do Ben et al get it? My sense is that they do, although Bernanke was pretty inscrutable through the entire morning.
One person who obviously doesn’t get it, or is pretending that he doesn’t (which would be even more disturbing) is Gary Gensler. October is sprinkled with major conferences, and Gensler has been doing the rounds. Despite all of the advances in understanding of the complexity of the issues involved, Gensler continues to peddle his cartoonish morality play:
“While the crisis had many causes, it is evident that swaps played a central role,” Mr Gensler said. “Swaps added leverage to the financial system with more risk being backed by less capital.
“They contributed, particularly through credit default swaps, to the bubble in the housing market. They contributed to a system where large financial institutions were thought to be not only too big to fail, but too interconnected to fail. Swaps – developed to help manage and lower risk for end-users – also concentrated and heightened risk in the financial system and to the public.”
No recognition of the ability of firms to substitute swap-based leverage with other types of leverage, and their willingness to do so. No acknowledgement of how mandates re-order priorities in bankruptcy, and the implications thereof. No appreciation for the potentially destabilizing effects of rigid variation margin mechanisms, and the pro-cyclicality of these mechanisms. And yeah, Gary: clearing mandates don’t concentrate risk and create their own interconnections.
Gensler is also still pushing the transparency nostrum. He links the crisis of ’08 to the opacity of the swaps market. Well, lack of pricing information was arguably an important problem with AIG in particular, his proposed cure–the SEF mandate–will have no effect on this. The problem with the instruments that AIG bought wasn’t that they weren’t traded on transparent markets: it was that they weren’t traded. Markets are not the Field of Frickin’ Dreams: If you build it, there’s no guarantee anybody will come. Comparisons of markets for illiquid products to exchange traded futures are superficial and silly.
In sum, Gensler continues to say foolish things. Is he a fool, or does he just believe his audience is? He batters straw men. He sticks to a cartoonish narrative of the events of 2008. He pushes nostrums. He ignores deep complexities, and the huge potential for unintended consequences of the policies he advocates.
Nonetheless, he is not shy about telling others what to do. @OTCderivatives tweets that Gensler said the CFTC should be “top dog” in IOSCO because of the size of US markets. In London, he chided the Europeans for not including exchange traded derivatives in Emir:
It is critical, though, that those swaps traded on regulated exchanges also are subject to the clearing requirement to bring the benefits of clearing to swaps, regardless of where they are traded – on exchange or off exchange.
This will no doubt thrill the Germans, because an expansion of the Emir remit would also force open access to exchange CCPs.
But given that the Germans have been quite adept at slapping Timmy! around, I doubt that a mere agency head will give them much trouble.
[Interesting aside. Risk Magazine’s Laurie Carver tweeted that Gensler blew him off to allow “other less aggressive journalists ask him comfy things.” Go figure.]
To sum up. Dodd-Frank has set off a very complex dynamic of unintended consequences, many of which will undermine the intent of the law. One of the figures primarily responsible for implementation of Dodd-Frank has a very simplistic view of the world that ignores these complexities. What could go wrong?
As I understand your argument, I couldn’t agree more. I would like to make a comment or two on the 2007-9 debacle. I am not an economist, however, nor even a quant, so forgive me if I am wrong or over simplistic in my judgements, but I was in a SIV consulting practice (just a bottle washer)and could see the following as clearly as if it had marched naked up 6th Avenue with a brass band.
The issue is, as always, leverage. CDS, however, did not by themselves lever the system, nor did it “allow” banks to lever themselves: What happened in early 2006 is that there was a fundamental mis-pricing of CDS relative to the calculated cost of capital of the (mostly) European Banks, along with the pernicious use of joint probabilities to calculate default risks. Also it was assumed that each deal was stand alone, when in fact all the collateral was cross correlated as to default risk (If one went then almost all would blow at once). It was true that some were less toxic than others, but not by much. As you pointed out there was no market for AIG – that wasn’t a bug in the system, it was the feature that allowed it to exist. Existing as it did, allowing the employees to show huge “profits”, hence bonuses all around.
On joint probabilities and mispricing: If you buy AA security x, insure it with a AAA monoline (remember those?) AAA or CDS AAA, the risk you have is supposedly rAA * rAAA, or virtually nil. Almost no capital is required. It got so ridiculous that ACA, an A- rated insurer, made money for a while wrapping its A- around AAA securities! rAAA *rA- < rAAA!?!
You will do this forever as long as the net interest income is greater than 0 plus the de minimus cost of capital ( Say .25% of capital on the whole position). If there had been no arbitrage there would have been no leverage. Add to this that AAA issuers (AIG) did not have to post collateral and you have the perfect perpetual motion machine – until someone wants to get paid. In this environment we produced crap, I mean collateral that would pass muster until we choked. And, by the way all the monolines' risk had were internally correlated to each other and the collateral – which was itself correlated to itself. All the insurance was on one big pool of bad loans – not 1000 separate structures. When one went, they all did. In other words the modeling and assumptions were wrong: we created systemic risk by failing to note the systemic or unified nature of the pools of assets being packaged and insured.
Now as to clearing, particularly CDS: will collateralization guarantee proper arbitrage free pricing? Beats me. What is the cost of collateral when interest rates are near 0 – not much. Cash calls, you bet, preferably financed off of collateral held by your trustee. Financing – voila a new customer base that takes all the cash we are choking on says the bank! This may be a good part of the motivation under the clearing boom (in addition, of course, of keeping the desks trading).
Future potential disasters:
1. Massive price movements causing variation marks that overwhelm the borrowing power of the collateral available.
2. Deterioration of collateral quality.
3. Liquidation panics that can trigger either of the above.
Your point, if I understand it is completely correct – all we may be doing is creating new borrowers for the lender of last resort.
Comment by Sotos — October 14, 2011 @ 1:17 pm
@Sotos-you understand my point perfectly. Your list hits some important points. Earlier posts have touched on some of these.
I found your description of what was happening pre-08 fascinating. Wrong way risk run amok.
Did anybody bother to ask why the CME keeps playing with the gold margins? Hmmmm?
Comment by Mr. X — October 15, 2011 @ 3:02 am
Dear Prof:
Glad you liked it – but of course it is a gross oversimplification. In addition, banks and regulators engaged in a “fan dance” of getting things off balance sheet when in fact they were on balance sheet. For example Citi had to bring almost all of its sponsored SIV’s back on balance sheet once no one would buy the CP that supported them. One little known area were the conduits – some of them suffered the same fate, while a few of them still are outstanding with heavily implied guarantees. The rules as to what is on or off can be truly bizarre. In conduit world, last time I looked if a pool of assets went immediately from investment grade to default it was on the conduits hands – if there was a downgrade to a lower rating, but greater than default, the sponsor has to but back. Can anyone explain why this is (was) somehow a lower risk trade to the bank in 99% of the cases? When the dung hit the fan, many of these fictions mostly blew up.
One thing that I have not seen commented on is the effect of making swaps generic is the effect on Street profits and trading. It was a truism that the street likes to make things trade-able. This is true to a point: what it likes is to create markets that trade, but have enough inefficiencies so there is a big spread and or inherent knowledge gaps. Take the late unlammented CMO as an example of both. I used to ask people why CMO’s existed, and would get an answer that we were taking a MBS and turning it into pieces that better met investors individual nees, etc. WRONG. the CMO existed because the spread on MBS was a tick or two, while the most generic CMO was 4 to 8 ticks. Firms, most notably Bear, carried this process even further by creating structures with over 30 tranches on one pool of collateral and multiple sepparate deals with cross collateralized pools of underlying collateral under one wrapper. This feature was compounded by the poor wquality of cashflows that were available to most of the individual institutional investors usually on Bloo** well let us not get into naming names. A number of houses made a very good living trading on crappy cashflows, while there own were better. Let us not even consider CDO’s squared, sythetic CDO’s etc.
Moving form the PreCambrian to the present day, central crealing will make trades even more generic, thus reducing spreads earned, thus putting pressure on the desks to boost trading volumes or lose their revenue. How will this help systemic stability?
The natural tendency is to “degenericize” products so that they can still trade, but at a wider spread. what is to prevent the same stupid, unproductive and ultimately dangerous process from starting off, using generic swaps as the base for some heretofor unthought of monstriosities as they did with MBS and corporate loans? As soon as the risk appetite moves beyond it’s current state, the boys and girls will be at it again.
Comment by Sotos — October 15, 2011 @ 8:04 am
Just catching up on my weekend reading, so I am late to the discussion, but I wanted to put in my two cents. The problem I have with the Dodd-Frank mandates is that the relationship between the industry and the regulators seems to be highly adversarial. Too adversarial. Regulatory decision makers seem to be very much in over their heads in trying to craft effective regulations and the industry seems to be fighting them at every turn with the knowledge that once regulations are cast in stone, ways to negate the effects of the regulations will be employed in every instance possible. There seems to be no interest on the part of industry participants and government regulators to agree that changes need to be effected and that new principles much be established for the development of “fair and efficient markets.” While details of regulations are discussed and considered, what we need is the emergence of a leader that can start from scratch, build consensus as to why we need to effect changes, make reforming the markets a shared priority for industry as well as government and to craft regulations that will be enforced jointly by industry participants first and by federal regulators second. We are crafting sweeping regulations with the wrong leadership in place. Accordingly, industry is going to employ enormous resources to comply with the letter of the law and to negate the spirit of the law. In the end, all this approach will create is higher costs for most participants and greater systemic risk for everyone.
Comment by Charles — October 17, 2011 @ 9:45 am
Thanks for this. My thinking has evolved on this issue quite a bit. Initially, I thought if we had used CCPs for CDOs etc we wouldn’t have had a financial crisis because we wouldn’t have been able to build up leverage.
However, the OTC derivatives market is multifaceted, and I am convinced that pre crisis if we did clear plain vanilla derivatives that the banks would have invented something more exotic that would have been unable to be run through the confines of a clearinghouse and the system would have eventually blown up anyway-only worse.
Currently I am torn. Maybe the do nothing and wipe away ALL regulation is the best approach and let markets sort it out for themselves. Or, is the best approach to have a little regulation?
I am reminded of the story: A mother tells her kids they can’t go to the movies. The kids are young teenagers BTW. They wanted to see an R rated movie. The kids are mad, because they think they are “old” enough. They say, “There are only a couple of bad parts.”
The mother decides to make some brownies for them. The kids are happy. However, before they eat them she says, “I just want you to know, I put just a little bit of dog poop in them. It’s really not much at all and you won’t notice. Eat them if you want to.”
That’s how I feel about the current CFTC, except there is a lot of dog poop in what they are doing.
Comment by Jeff — October 18, 2011 @ 5:20 am
@Jeff–LOL. IMO, a dog pile with a little brownie mixed in. Maybe.