David Henderson reviews the new Alan Blinder book

The review is here, here is one interesting paragraph:

Mr. Blinder omits a crucial fact about Lehman, one that, by itself, explains why the huge drop in value of Lehman’s mortgage-backed securities led to its collapse: the effect of changes in federal bankruptcy law. Thanks to the 2005 Bankruptcy Abuse Prevention and Consumer Protection Act, when Lehman went bankrupt it could not simply, as in earlier days, pay holders of derivatives as much as possible with its assets. Instead, it had to give each derivative holder a new contract identical to the one it had signed with Lehman, but with a different counterparty. Lehman would also have to pay the transaction cost of the new contract. Such costs are typically about 0.15% of the contract’s total value. That’s small, right? No. When Lehman went bankrupt, the face value of Lehman’s derivative contracts was $35 trillion—with a “t.” The transaction costs alone were $52.5 billion. That is what sank Lehman.

The strange and indeed unjustified senior status of derivatives contracts remains an under-discussed area for financial reform.  Here is a relevant Bolton and Oehmke paper (pdf).  The Blinder book you can buy here.


Why are transaction costs a 0.15% of the contract's value? Are there billions of contracts and it really takes $52.5 billion in labor to rewrite them? Or are a few firms running the whole show on derivatives and taking their cut?

Contracts with a face value of $35 trillion means there was a lot of leverage involved; that is the reason Lehman went down. It wasn't the transaction costs, it was the ridiculous amount of leverage that made .15% of the contracts value wipe them out. .15% is small.

Given, then, this fact of superiority of derivatives over other security interests, shouldn't creditors then restrict companies derivative exposure to protect themselves, ie, writing into loan covenants that would limit exposure.

What am I missing from self-help options available to creditors before they become creditors?

If you change the rule by legislation now you would be probably giving a windfall to these otherwise junior creditors (they would be given more security for transactions written under the old rule).

Isn't the issue one of disclosure, transparency, and covenants?

By covenants I mean: if I give you a loan, you agree not to take on leverage exceeding a certain level of exposure, or else I don't give you the loan.

If you know what the rules are regarding bankruptcy priorities, and you are considering giving someone a loan, you would write in a loan covenant limiting how much derivative exposure the company could take on.

Did the law apply retrospectively to derivative contracts already in existence?

That Lehman could have $35 Trillion in derivatives contracts outstanding is prima facie proof that derivative transactions are shams.

If I was long 17.5T and short 17.5T for a net position of 0, why would it be a sham? In fact, that might be the natural outgrowth of taking a long position when a client wanted a short position and hedging the long with another short while also taking short positions to allow clients to take long positions while hedging those shorts with other longs. If not all clients wanted to trade exactly the same securities at exactly the same time, then might we not expect that a market maker would end up with large gross long positions offset by similar-but-not-exactly-the-same large gross short positions?

It might even be the case that under normal liquidity conditions, losses/gains in the longs actually are pretty well offset by gains/losses in the shorts. It's certainly the case that, globally, the net position across all investors in all derivatives is zero, any dollar lost on derivatives by one person is necessarily gained by another, so there can be no net global loss in wealth from derivatives. However, under illiquid conditions, if I am forced to unwind both longs and shorts at unfavorable prices, then that could push me into insolvency. Furthermore, if no one is sure who is suffering losses and who is gaining from these derivative contracts, then that can cause liquidity to dry up as people fear that their counterparties might not be able to meet their obligations.

Should bankers and other financial professionals know better than to assume that there will always be unlimited liquidity when they need it and should they anticipate and plan for liquidity crises? Certainly yes. That fact that they didn't demonstrates their incompetence, not that there was some "sham".

This also highlights the difference between government interventions intended to alleviate liquidity crises, for example, to allow the orderly unwinding of derivatives positions knowing that in the end the gains and losses will net out to zero globally, vs government interventions designed to save particular insolvent firms (like an automaker as a hypothetical example), firms where even with unlimited liquidity their assets would be insufficient to meet their liabilities.

'That fact that they didn’t demonstrates their incompetence, not that there was some “sham”.'

Sure, but now what about the larger implications of that: it's not exactly an endorsement of the wonderful workings of free markets. If I stupidly decide that it's a good idea to spend $52 for a Rolex that that guy on the street is offering, well caveat emptor. If an investment bank stupidly decides it's a good idea to accumulate $52T of bad debt, it's caveat world (caveat mundi?).

"If I was long 17.5T and short 17.5T for a net position of 0, why would it be a sham?"

What reveals it as a sham is the disproportion between amount of derivatives outstanding and the size of the underlying market. We have over $700 Trillion in derivatives outstanding on on underlying global economy of about $60 Trillion per year. You are using as a hypothetical that one bank would write derivatives on something equaling the national debt of the United States.

You still aren't grasping the difference between notional value and contract (or fair market) value. You can write a $1 trillion notional amount interest rate hedge contract that has an at-risk contract value of $1 million. That's not a sham...that's the nature of a hedge.

It's sort of like in the insurance world comparing policy face amount with liabilities/reserves.

Norman I think I understand the significance of the notional amount well enough. Even though the exposure on an interest rate swap is only on moves in the interest rate, who has a legitimate need to buy protection against interest moves on 15 Trillion of principal? I will confess that I find it difficult to imagine that anyone would or could write a legitimate transaction with a notional value of $1,000,000,000,000 that has only $1,000,000 at risk.

Joe, it doesn't work that way for an investment bank. On day one a customer want to buy say $10M notional 5Y standard swap and Lehman does that. Lehman doesn't like the interest rate risk so it buys the same (slightly different rate) swap from someone else. Another customer comes the next week and want to sell a $10M swap and Lehman again goes to the market to offset the risk. The offset isn't always exact and you can offset two customer deals when you are lucky, but the above can easily happen and we have just added $40M to Lehman's derivatives portfolio even though Lehman has (and specifically worked to avoid) no net interest rate exposure. These kind of trades add up quickly. Lehman did of course also do their own long/short trades of slightly different derivatives like the three major synthetic CDO indices (ITRAXX) but the next market risk was not dramatic.

These trades do of course carry counterparty credit risk and that should be the main concern. Banks spend a lot of effort trying to measure this and this is an area where the regulators both here and in Europe has gotten a lot tougher.


"On day one a customer want to buy say $10M notional 5Y standard swap"

I understand that it is improbable that anyone would want to write a swap on $15 Trillion all in one go.

I am not a big city big finance lawyer. I could understand there being a few customers who wanted to write $10 Million or even $100 Million swaps. But this former physics student, now small town lawyer, does not see where you get enough people wanting to do swaps that you get to $400 Trillion notional in interest rate swaps - at $10 Million a pop you would need to write 40 Million contracts - at $100 Million each, you still need four million contracts. I know that there will be some double counting but you need a HUGE number of contracts.

Law and physics both taught me to do reality checks. If something makes no sense then there is something wrong. It makes no sense that the derivatives market is $700 Trillion in total - there is something very wrong. As you note there is counter party credit risk and it is a concern when the notional is $700 Trillion even if the value at risk is SUPPOSED to be vanishingly small (at least relative to $7 * 10^14). The banking system could be one wrong spreadsheet formula from collapse.

I'm a finance lawyer who works with derivatives and I have no f'in idea what he is talking about. For one thing, his statement doesn't even make sense. He ends by saying that the transaction costs is what sank Lehman, but earlier he says that the transaction costs of re-writing the derivatives contracts (what?) were brought about by the bankruptcy itself. So by his own statements, the transaction costs did not sink Lehman; it was already sunk. And I fear he might be confusing "contract value" with notional amount. There is no way it costs $52 billion to write contracts (as a lawyer, that would be nice).

Good point.

And isn't the $35 trillion figure meaningless unless we know how the contracts netted out?

Excellent point Counselor. I was just about to write that the sentence "Thanks to the 2005 Bankruptcy Abuse Prevention and Consumer Protection Act, when Lehman went bankrupt it could not simply, as in earlier days, pay holders of derivatives as much as possible with its assets" make s no logical sense. "when Lehman went bankrupt"? In other words, is the issue what the creditors of Lehman would get, and how unfair this is? Why is that important? And that precipitated bankruptcy? Cart before the horse. It does not explain why Lehman went bankrupt in the first place. No firm wants to go bankrupt, and the fact bankruptcy law may be unfair to certain parties is not a reason firms go bankrupt. This link is probably a better source of why Lehman went bankrupt (and notice nobody wanted to buy them before bankruptcy either, showing they were salvageable): http://en.wikipedia.org/wiki/Bankruptcy_of_Lehman_Brothers#Exposure_to_the_mortgage_market

From the review:

This bailout, according to Mr. Blinder, was "sorely needed" to stem the fire sale of commercial paper. But that's highly debatable. Had the Treasury made clear that it would not bail out MMFs, then many of them would have also had to "break the buck." Once depositors knew there was no gain from getting their funds out early, the run on MMFs would have ended, thus stopping, or dramatically slowing, the plunge in value of commercial paper.

The subject under discussion is the insurance provided money market funds after Reserve Primary was forced to "break the buck" as a result of holding some Lehman commercial paper. This was done to stop a developing run on MMF's. Apparently Henderson thinks the run would have simply stopped of its own accord without intervention, though his logic looks fairly suspect. Why wouldn't there potentially be a gain from getting out early? And are MMF investors really going to calculate things that carefully or they going to cash out now and worry later?

The reason to get out early was to get $1 per $1. Once MMF's made clear that they would break the buck and not give $1 per $1 but, say, 97 cents per $, the incentive to get out early goes away.

I still don't follow. Suppose you believe that your MMF is sound, but you also believe that there are a lot of panicked investors in the fund (and other funds holding similar securities) that will redeem their shares. Those redemptions will cause the MMFs to sell commercial paper into an illiquid market at unfavorable prices to meet redemptions, which will in turn lead to drops in NAV below $1. The panicked selling can cause additional price drops and even less liquidity. Thus, you are incentivized to redeem ahead of the panicked investors even if you believe that their panic is unjustified. How was the MMF situation different from bank runs that FDIC is intended to stop? Keep in mind that investors redeem at NAV, but the selling of securities to meet redemptions can actually affect NAVs after the initial redemption date. When one fund sells one security at an unfavorable price, it affects the mark-to-market valuation of identical and similar securities that fund and other funds continue to hold.

With finite liquidity, it seems like there needs to be some mechanism by which un-panicked investors can be assured that the fears of irrationally panicked investors won't be self-fulfilling.

Unless you fear that if you wait you will get, say, 92 cents instead of 97 cents.

Henderson would have appeared on TV reading the Jimmy Stewart lines if only the government had given him a chance to arrange the TV presser and dress up and rehearse the lines.

Ignore that review. The quoted paragraph is nonsense (not a "crucial fact").

It's really hard to guess what the reviewer has in mind. The line about paying out as much as possible out of assets before 2005 and instead having to pay fees to write new contracts after has no basis in reality. Despite some popular accounts, there was no change in the 2005 law that bears here, and the priority of derivatives counterparts owed money on BK is the same as general unsecured creditors, not any higher. No one who understands derivatives issues talks about $35 trillion in face value, and $52.5 billion in fees is an absurd fantasy. Not trying to slam or be harsh but just pointing out you should completely disregard anything by that author on derivatives or BK.

About the write the same thing, this paragraph is a mess. Every sentence but one in the quoted paragraph contains a mistake and the one sentence that isn't just wrong ("Lehman would also have to pay the transaction cost of the new contract.") is almost certainly misunderstood by the author as the author does not realize that derives are the same as other senior unsecured. Lehman went bankrupt because their assets went down in value by a lot, not because of the shady changes in bankruptcy laws.

Also, as others have mentioned, it just doesn't make much sense that rules regarding claim values once in bankruptcy are what drove Lehman into bankruptcy. Now, the ISDA rules might have explained why the recovery rate was so low once they got to bankruptcy, but that isn't the reason anyways as the deriv claims were not so large compared to the rest of debts.

The theory that Lehman's bankruptcy somehow caused the "financial collapse" has been absurd since day one. The "collapse" was caused by the faulty belief that American real estate wouldn't decline. This belief was shared by virtually all investors, homeowners, lay people, regulators, and humans.

Lots and lots of austrian economists said it was about to crash. Some even went on TV and said it.

Jim F-ing Cramer even went on TV about a year before the crash and said that there was a looming foreclosure crisis.

President Bush went to congress and asked them, over a year before it happened, to reign in Fanny and Freddy, because his advisors said that there was a bubble

The most popular econ-writer on the planet (Krugman) saw it coming.

Dude, it seems the only people who didn't see it coming were people who don't believe in bubbles, and regulators, and banks.

Rather than layer on another critique of the review, which had just enough balance to keep me reading it. I thought I would post the highlights of the book, it seems, from the author. (Sorry for the repost, but I guess I was a little impatient yesterday and it fits better here.) Blinder's op-ed: http://www.nytimes.com/2013/01/20/opinion/sunday/financial-collapse-a-10-step-recovery-plan.html?partner=rss&emc=rss&_r=0 In his review, Henderson criticizes Blinder for going to easy on government and regulations. I have not read the book, in my pre-order list now, but I suspect the differences in Blinder's story runs more as business with sins of commission and government with sins of omission. Surely financial-housing regulations added some fuel to the fire, but there were some regulators who could have thrown a lot of water on the whole blaze and did not. Why not? And are we really set up well to avoid the next episode?

" but there were some regulators who could have thrown a lot of water on the whole blaze and did not. Why not? And are we really set up well to avoid the next episode?"

George W. Bush asked them to, but the Congressional Black Caucus blocked it, as housing was their pet issue.

Notice the plural on regulators in my statement. There were many who stood by and no one political party is to blame. The markets self-correct or self-regulate view was also alive and well at that time, not just the push for social policy in housing.

Comments for this post are closed