Market Liquidity

Potential flaws in derivative markets:

FT.com / Home UK - On Wall Street: Derivatives cannot take the pressure: By John Dizard: The recent difficulties settling futures contracts in Chicago have highlighted a growing problem: there are big derivatives markets out there that aren't built to take the strain being put on them.... In June, some large holders of the June 10-year Treasury futures contract, including Pimco, demanded settlement -- taking delivery of actual bonds -- instead of, as usual, rolling their positions into the next contract. The scramble to find the necessary notes was made worse by the fact that one account, possibly the hedge fund Citadel, already held the bulk of the cheapest notes to deliver.

Whether or not Citadel, or Pimco's bond dealers, intended to make some extra money by squeezing out other bidders, it couldn't have happened if the structure of the futures contract had kept up with the times. The Chicago Board of Trade's 10-year T-note futures contract is one of the older financial derivatives around today. That's good, in that it has been tested through a range of crises and cycles, and it's bad because it was designed based on the classic requirements of grain dealers, rather than the current requirements of interest rate markets....

The 10-year futures contract is based on a theoretical Treasury note that doesn't exist... the CBOT have formulas to convert the value of existing Treasuries into the theoretical ones. For example, the cheapest bond to deliver to settle the current September contract is the August 2012 bond. So you are really getting delivery of, and contracts are priced from, the seven-year stuff rather than 10-year stuff.... There is now about eight times the number of outstanding futures contracts as bonds eligible and available to fulfil them.

The obvious solution is to go to cash settlement. That would mean the exchange changing the terms of the 10-year contract, so that actual bonds would not be delivered. Instead, those now obliged to deliver would make cash payments equal to the value of a reference price. That is how Fed funds futures contracts are settled....

The real problem is that the US economy is just too leveraged. Starting with the housing industry, the country is too dependent on derivatives markets to create the illusion that interest rate risk can be conjured away. The technical problems of the 10-year are just another early warning sign of this fundamental weakness.

I don't know enough to know whether this is a serious issue or not.

Comments

[nate]

Look to Nebraska and Buffet on derivatives:

http://www.fenews.com/fen31/one_time_articles/warren_buffet.html

"linkages" and "counterparty risk"
(per fuzzy recollection Soros talks about ideas behind linkages too - financial contagion or something like that)

excerpt from Buffet:
"In either industry, once you write a contract -which may require a large payment decades later -you are usually stuck with it. True, there are methods by which the risk can be laid off with others. But most strategies of that kind leave you with residual liability."

[nate]

same buffett article except more concise URL:

http://tinyurl.com/3yx8j

[Glen Bowman]

Stock index options like the SPX (S&P 500) and OEX (S&P 100) are already settled with cash. But if a person buys futures, whether they be soybean or single stock, the person should really have the option of actually taking delivery. Otherwise, there would be more opportunity for manipulation (that's the rationale behind prohibiting naked shorting).

Otherwise, you can have another Hunt silver fiasco. The Hunts were sunk only after the exchange made them spread out their delivery. But they did not want to own the metal; they just wanted, apparently, to manipulate the silver market.

[Glen Bowman]

Actually, squeezing the shorts by demanding delivery of the underlying (stocks like Piggly Wiggly or E.L. Bruce or commodities like silver and wheat) is an ancient trick. At the same time, that doesn't sound like something Bill Gross would do. He has a sterling reputation to protect.

[bradsetser]

I don't know as much about derivatives markets as i would like, but isn't part of the issue here that the size of the derivatives market has expanded dramatically relative to the market for the underlying instruments? Particularly since a decent share of treasuries are held by accounts that do not trade actively (i.e. foreign central banks). So in some sense, there are not enough securities available if everyone demanded delivery of the underlying instrument. the market works so long as they don't, but in circumstances where someone demands the underlying security, or where you need the underlying security to hedge your risk, there could be problems.

at least that is my (not that well informed) sense. wonder if the counterparty risk management group covered this in their report

[Glen Bowman]

Most people who trade derivatives are not interested in taking delivery. They are speculators or hedgers wanting to build/protect their portfolios. On the other hand, Bill Gross is the manager for the biggest bond fund in the world, so he might well really want to take delivery.

Ten-year note futures have more trading volume, I think, than any other futures contract traded in the United States. It has really taken off since the stock market bubble crashed (and the real estate market took off).

[Jonathan W. King]

Brad DeLong writes:
>
> I don't know enough to know whether this is a serious issue or not.

Well, if there is now already 8 times the value of outstanding bonds in the futures market, you could foresee nightmare delivery situations occurring, especially down the line when the multiple might be 20 or 50. So isn't there a technical issue to solve here anyway?

The cash settlement answer is easy enough, but now I begin to wonder if yet another way for the world's derivatives markets to blow up, based on instruments that could combine real commodities with naked bets on stuff like the T-note. So, what if, to unwind some position in a crisis, you needed to deliver a countably infinite number of porkbellies? Or what if somebody started a cascade of ruin because a short squeeze on T-bills forced a number of big players to unwind? As usual, my thinking about derivatives markets gets really foggy really quickly since, right now, you have no idea what bizzaro instruments or structured notes or whatever are really out there. Somebody messing up and ending up with the entire 2006 winter wheat crop getting dumped on their desk has always been a theoretical possibility, but it will never happen. Some large hedge fund that ends up busted because the Candian dollar went up a nickel while oil futures are down $3 might be just as unlikely, but it would be really hard to tell at this point.

[John H]

Come on - in this case the Treasury would solve the problem. They would LEND additional 10year notes collateralised by (excessive amounts) of other government bonds or cash.

The Treasury would naturally make (quite a nice) spread on it - but the losses to the Primary Dealers (who presumably need to borrow because they or clients fail to deliver) would be small (relative to their balance sheet).

Fails in Treasuries are the least of our worries.

The problem is in commodities or equities. People can (at a pinch) create more GM shares if there is a failure to deliver GM shares. At a high enough price GM will do it for you (they need the money). But a high enough price might involve substantial financial market losses.

Central banks might instantly supply some gold to solve a failure to deliver crisis (but nobody is creating more).

Instantly creating copper or oil however looks very tricky.

JH

[monkyboy]

A quick Google of Long-Term Capital Management reveals they used about $5 billion in capital to purchase control of $1.25 trillion worth of commodities. When their "foolproof" scheme unraveled, the Fed stepped in with a bailout package

[You've got the story wrong]

[Peter]

I think there are two answers to your question, Brad. One is technical, and I think the short version of it is, no it's not really a problem. Cash-settled futures are relatively recent, I believe since about 1981, and they have fairly effectively 'solved' the problem of attempted corners. What's kind of wild about this example is that while we've seen corners and attempted corners via futures markets for physical commodities (silver most recently, wheat most famously), there haven't been a lot of attempted corners on financials. Whether or not this is an attempted corner or a real position is another (also interesting) matter.

The second question is more interesting. It's something like, do futures markets still work the way they were designed to? Not as far as individual firms being able to wipe out liquidity (LTCM being most recent), but the fundamental assumption that futures markets still correctly and consistently work as they are supposed to in the face of full-throttle advanced financial capitalism? I suspect the answer is, we don't know. These things are based on premises from the 1870s that haven't really changed much, while the world so dramatically has.

[wcw]

Who cares how futures were designed back when?
The question is whether futures markets continue
to offer utility to the practitioner. When
well-regulated under normal circumstances, the
answer is pretty apparently 'yes'.

I think all this article is really saying is,
"cash settlement of index futures is best."
That is a sentiment with which I agree.

[tjallen]

Cash settlement won't work if you are ADM or General Mills, and bought futures with the old-fashioned idea that you will get their corn at the end of the season. You need and want that corn (to make gasoline or corn flakes), and will lend the farmers seed money via a futures contract, to guarantee not only a stable price, but to guarantee you get the corn!

A futures contract on corn that is settled in cash is a futures contract on cash, not corn.

[wcw]

"cash settlement of index futures is best."

Corn futures are commodity futures.

If you are arguing that treasury futures should
for some reason be settled like commodities rather
than indexes, I shall be interested to hear your
view. Perhaps you merely misread, though.

[monkyboy]

If I got the story "wrong" on LTCM, maybe someone could post the "right" story.

I was reading another story about a bank in England that wanted to take delivery on some bonds recently, but there were 2 problems:

1. The bonds they held futures on don't actually exist.

2. For the bonds that most closely matched what they held futures on, there were 8 futures contracts for every actual bond.

Futures without any underlying commodity or security are just bets and should be traded in Las Vegas, not Chicago...

[Barkley Rosser]

To Glenn Bowman: The repo market in Treasuries far exceeds the direct market in them. Indeed, that is the market the Fed operates in for its open market operations.

There are many financials where the derivatives far exceed the assets. Cash settlement helps, and in most situations the derivatives vitiate risk. The problem is that while local risk in specific markets is reduced, global risk or fragility has arguably increased. If many of these markets are facing crisis, the related transactions can make it impossible to resolve many of them. We have already seen this in the case of LCTM, which did require an extraordinary intervention by the Fed.

[meno]

Only a serious issue of the Treasury lets it be one.

So you can make dosh this way, but it get too greedy then Treasury will step in, issue some special bonds that are far and away the cheapest way to settle these virtual bonds, and you'll lose out.

[monkyboy]

Hehe, the bank in England story was a link from this very thread...duh!

I think what we are seeing is that there is simply too much capital in the world chasing too few real assets.

Chinese workers last years saved $1.5 trillion from their meager paychecks. That's enough to buy about 10% of every publicly traded company in America. American companies' market cap represents about 75% of the market cap of every public company in the world. So, in about 10 years, at the current rate, Chinese workers alone will have saved enough to purchase every single share available on the world markets...what will they invest in then?

I don't think any economic theories provide answers when the world is awash in capital. I think we have entered the age of the permanent bubble and silly securities, oil and U.S. Real Estate are just the latest, certainly not the last bubbles.

[gundryggia]

A few points:
Bill Gross would probably have rather rolled over his position to the Sept contract; however, that would have hosed him in this case so he demanded delivery.

There is no danger of "running out" of treasury bonds in the way some seem to think. The Chicago Board of Trade lets you deliver a number of bonds for an expiring contract; the problem is that for the June contract one bond was much cheaper to deliver than all the rest, so everyone who had to deliver drove up the price of that particular bond to ridiculous levels and lost money. Horrible, I know. Think of the all starving bond traders.

Yes treasury futures should be settled in cash (simpler for all involved) but what is life without challenges?

(I do not see this has much to do with LTCM other than "there's a lot of money in derivatives.")

[monkyboy]

Just because there are "respected" financial people using fancy mathamatical models to create derivatives doesn't mean they aren't just pretend investments.

What tubed LTCM was a rush to liquidity, or "real" assets.

On the China scene today from the MoscowTimes online:

HONG KONG -- China National Petroleum Corp. has agreed to buy PetroKazakhstan for $4.18 billion, topping a bid by an Indian producer and securing supplies for the world's fastest-growing major economy

When China owns everything, all we'll have left is Monopoly money trades through Chicago...

[Kaleberg]

As an electrical engineer I remember that there are two principles used in designing stable electrical systems, like amplifiers or power grids. You can limit the range of operation or you can line up the poles, unstable situations, with the zeros, impossible or irrelevant situations.

For example, you could limit the range of operation by limiting the number of futures that can be sold to the number of underlying items, or a conservative projection thereof.

Alternatively, you could design your financial instruments to stay stable across an expected range of events. This is called hedging one's bets.

As anyone who has built an amplifier or a power grid knows, they can still go instable. All those local linearities we cherish are merely approximations. Without them, design is futile, but because of them, perfect design is impossible. Warm capacitors behave differently from cold capacitors. Short impulses and unexpected resonances destroy the best laid of plans.

The mathematics are similar for financial models, and most likely the real world instabilities are similarly unescapable.

[christo]

Whoever these wicked smart bond and futures traders are, when the bubble bursts they'll get to keep their bonuses for the last 3 to 5 years while the taxpayers get to clean up the mess. It's that damn asymetrical class warfare again, millions for you and chump change and partial ownership of a huge taxbill for me. If the market at settlement isn't reacting the way it's modeled, isn't that evidence that the black box boys don't fully understand what they're doing? They modeled everything perfectly except for the unexpected, that doesn't sound master of the universeful. I think we sound go back before the Roosevelt era when directors of banks had full liability for any debts their banks had. Personal responsibility for free market traders rather than unlimited amounts of free bail-out money. this is not something the central planner Mr. Magoo will be supporting in public testimony.

[monkyboy]

Hehe, christo, I don't think there will be another bailout for the "black box" guys. For most derivatives, there has to be a buyer for every seller. It's just musical chairs for bored, wealthy Americans.

As for class warfare, I think we should look at executives voting themselves huge retirement payouts as they line up their companies to be purchased by China...

[Peter G.]

'monkyboy' there are _three_ types of class warfare going on in the USA right now:

* Management are stealthily expropriating shareholders and capitalists.

* Management and capitalists have shifted the tax burden almost entirely on the middle class, and are thinking of going further: http://www.fairtax.org/

* The voting middle class are exploiting ferociously the non voting immigrant ''lumpenproletariat''.

[The Nattering Naybob]

Brad, it is a very serious issue as a lack of volatility indicates not only complacency buy a lack of liquidity as well. And that can lead to really big problems.

[J. Patrick Bedell]

Hi Mr. DeLong,
From the perspective of a student of Mises and Rothbard, it seems apparent that the failure to deliver the underlying is a symptom of the continual expropriation of value from the economy through the creation of paper (fiat currency) and derivative financial instruments based on that currency. The credit pyramid based on debt instruments (for example) has paid for quite a few political contributions!

Patrick

P.S.: I hope that the development of digital settlement of public key-based financial instruments (particularly at http://infoeng.sf.net) will place the truth of their perspective beyond the reach of interested sophistry... but that is yet to be seen. ;-)