I was forwarded the following Bloomberg news article today:
July 28 (Bloomberg) -- The U.S. military said it's
reconsidering its plan for a worldwide on-line futures market to
help it predict events in the Middle East.
A statement issued late today said the Pentagon ``will
continue to reevaluate the technical promise of the program
before committing additional funds beyond fiscal year 2003.''
Just hours earlier, two Democratic U.S. senators held a
press conference to demand cancellation of the project, which
would allow traders to bet on the likelihood of events ranging
from the overthrow of a government to the collapse of an economy
or the assassination of Palestinian leader Yasser Arafat.
``Clearly, this is morally wrong,'' Oregon Senator Ron Wyden
said on Capitol Hill in Washington. Senator Byron Dorgan of North
Dakota called it ``the most Byzantine thing I have ever seen
proposed by a federal agency.''
The Pentagon requested $3 million for the project in its
fiscal 2004 budget. The Senate said no; the House said yes. The
budget is now before a conference committee of both chambers. If
the Pentagon doesn't end the project, ``we have to find a way in
conference,'' Dorgan said. ``This is a hare-brained scheme.''
The Policy Analysis Market is part of a Pentagon effort to
anticipate terrorist attacks. The market's Web site says traders
can register starting August 1 with trading to begin Oct. 1.
`Small Research Program'
The market is to be managed by the Pentagon's Defense
Advanced Research Projects Agency, or DARPA. DARPA spokesman John
Jennings said the agency was caught off-guard by the senators''
press conference.
DARPA's ``statement regarding the future'' of the program
said there still are ``a number of major technical challenges and
uncertainties'' facing this ``small research program.''
``Chief among these are: Can the market survive and will
people continue to participate when U.S. authorities use it to
prevent terrorist attacks? Can futures markets be manipulated by
adversaries?'' it said in its four-paragraph statement.
The Pentagon planned to ask for $5 million for the program
in fiscal 2005 on top of the $3 million requested in fiscal 2004.
The statement left unclear whether any of this money is still
sought, and a phone call to Jennings for clarification was not
immediately returned.
`How Would You Feel?'
Wyden and Dorgan said the market was a terrible idea.
``How would you feel if you were the king of Jordan and
learned that the U.S. Department of Defense was creating a
futures market in whether you're going to be overthrown,'' said
Dorgan, a member of the Senate Appropriations Committee.
The Web site for the Policy Analysis Market cites
``Overthrow of Jordanian Monarchy'' as an example of the sort of
event traders might speculate on. Others include Arafat's
assassination and the likelihood of a North Korea missile attack
on the U.S.
``PAM will focus on the economic, civil, and military
futures of Egypt, Jordan, Iran, Iraq, Israel, Saudi Arabia,
Syria, and Turkey and the impact of U.S. involvement with each,''
according to the ``Concept Overview'' on its Web site.
``The issues represented by PAM contracts may be
interrelated'' and traders can ``structure combinations of
futures contracts,'' it says. The market ``will be active and
accessible 24/7 and should prove as engaging as it is
informative.''
Dorgan didn't see it that way. ``Futures markets almost
always relate to a commodity in this country,'' he said.
``There's not a commodity here. It's wagering.''
Wyden, a member of the Senate Intelligence Committee, was
scornful of the project.
``They have this notion about the predictive capabilities of
intelligence,'' he said. ``We think it's absurd; we think when
you look at what happened in 9/11. You ought to go on the basis
of real evidence from the real world. They have these ideas about
markets and predictive capabilities which are more of fantasy
land.''
--Tony Capaccio in the Washington newsroom (202) 624-1911 or
acapaccio@bloomberg.net. Editor: Schmick
Story illustration: To review DARPA Policy Analysis Market Web
site type http://www.policyanalysismarket.org/
It's a shame this tiny project has become a political/budget fight.
First, why are two senators making a capital case over $3 million, when the entire US 2004 Defense Budget request is $399.1 billion? That only accounts for 7.5 ten-thousandths of one percent of the total budget.
Secondly, it's (unfairly) made to look absurd as part of this politically motivated ambush. In fact, the government has funded projects to achieve these same goals since the 1960's.
I've no interest in thinking about politics, let alone writing about them. But I am chronically interested in markets, and this looks like an interesting and sophisticated one.
But first, I think it's important to understand that this idea stems from a line of research that began during the Cold War.
The Delphi Method
For the past several decades the government has struggled with ways to "produce a tool in which a group of experts could come to some consensus of opinion when the decisive factors were subjective, or less knowledge-based."
In the 1960s the RAND corporation developed the Delphi Method.
As quoted in the desription: "Delphi is particularly appropriate when decision-making is required in a political or emotional environment, or when the decisions affect strong factions with opposing preferences. The tool works formally or informally, in large or small contexts, and reaps the benefits of group decision making while insulating the process from the limitations of group decision-making; e.g., over-dominant group members, political lobbying, or "bandwagonism". "
Certainly matters of the Middle East fall into the "political and emotion environment" that involve "strong factions with opposing preferences."
So this is a justifiable problem to tackle.
The Market Method
It's also conceivable that using a market is a reasonable way to attack it. Without getting into details, the Delphi method is quite regimented and procedural. Markets, in their apparent chaos, theoretically have always priced in all available, useful knowledge held by anyone, not just a board of "experts".
Furthermore, some of the problems of a group-decision making process (over-dominant members, political lobbying, and "bandwagonism") tend to be cured by the market -- bad or irrational choices lead to bad consequences and good choices lead to good consequences. It's terribly, awfully, painfully fair.
The project is working towards integrating lots of opinions and knowledge into the best possible single, unambiguous opinion. That doesn't sound loony, does it?
When I originally started writing this blog entry, I didn't realize that this market was already being developed. I thought it was just a research paper. No! It turns out it's a live electronic market that allows trader registration starting August 1, and begins live trading October 1. Furthermore it appears that it will be a cash-settled market -- no 'mock book bullshit' here!
The site only provides enough details to be titillating.
Contracts will be settled by "observable events" on a quarterly basis. EG. "The Jordanian Monarchy falls by Dec 31st 2003" or "No US forces remain in Iraq by September 30, 2004." The contract specification is very summary, but I assume there is a fuller definition that covers all the yucky exceptions.
That there are several periods for each contract means the market will have a structure, and that means trading and arbitrage opportunities, which means a good market.
Another very interesting, and I believe unique, characteristic of this market is the ability to trade products of two different contracts. In the example given online, you would be buying or selling one of the four combinations of: "The Jordanian Monarchy Falls" and "Iraqi Regime crumbles after 1 month of hostilities". The idea is that if you only have good knowledge of one of those interelated topics you can still express your expert opinion while hedging yourself against your blind spot. The market keeps the the odds of these four possible outcomes always summed to 1.
This is quite an interesting concept. The Delphi results are good for doing things like rank ordering preferences. But looking at the shape of the futures curves, the volatility of those curves, and the spreads between contracts could paint a very toned picture of the market's "best opinion." It needs to be, questions about the middle east are far more complicated than deciding for which of four public health centers to cut funding.
I'm quite interested to participate in this, just to experience how it works/doesn't work. Whether the DARPA part of it survives or not remains to be seen, but fortunately some good companies also back it, including The Economist and Net Exchange. Hopefully even if it loses government funding, commercial funding will still be available.
===
Wednesday, July 30th update.
Dead
So the project is absolutely dead. I was shocked by the hew and cry of the public and media over this. Get the impression maybe this was aimed more at John Poindexter than anything else.
The thing that suprised me is how everyones' attention was fixed on a basically ludicrous argument -- that this market would somehow incentivize a terrorist army making millions for themselves by buying futures and then assasinating world leaders.
First, I think the whole idea of derivative-trading mercenary assassins is ludicrous. The terrorists already have plenty of far less ostenatious business schemes running.
But if you insist, terrorists wanting to enrich themselves by driving current events, they wouldn't be using an exchange like this. How much money, open-interest, could there possibly be in this exchange, compared to something like the New York Stock Exchange or the International Petroleum Exchange? I'll tell you the answer: SQUAT.
If you can kill the Saudi Royal family, believe me, the price of oil futures will make you more-than-rich. There is no need to to be trading tiny "Decision Information Markets."
Back when I started my career in oil, the Iraqi Oil-for-Food scheme was just being negotiated I remember a particular (particularly odd?) consultant consistently asserting that Saddam was manipulating the negotiations and headlines so that his gnomes in Zurich Banks could expand his fortune trading oil futures.
So this idea is nothing original, it already goes on, and this whole project is a total tempest-in-a-teapot.
Post Mortem
Another thought about this market is, "why is this exposed to the public at all? It's the academics, politicians, military, and industrialists who would have anything to add, anyway, not John Q. Public who only reads Time magazine."
It occurred to me that that was the whole point of the market. In allowing anyone to participate, small slivers of knowledge from obscure places would otherwise be ignored.
In the Delphi method, yes, only some "inner circle" types are going to participate. Now yes, they know much, much more about the topic of Iraqi-Jordan relations than the average citizen. But where the Delphi method loses out is when a Coptic-Christian emigre from Iraq, living in Dearborn Michigan, who has cousins in Jordan, hears an incredible slice of news from his relatives that provides good insight into the situation.
This is where the market shines -- it allows this obscure, useful-once-in-a-lifetime guy to enter the market, and by entering the market, his tiny information contribution enters the universe of collective opinion forming.
That's the magic of the market. Unfortunately this whole research project got off to a very clumsy start.