Decision markets l.jpg
This presentation is the property of its rightful owner.
Sponsored Links
1 / 61

Decision Markets PowerPoint PPT Presentation


  • 58 Views
  • Uploaded on
  • Presentation posted in: General

Decision Markets. Robin Hanson Department of Economics George Mason University. “Pays $1 if Bush wins”. Will price rise or fall?. sell. E[ price change | ?? ]. buy. price. sell. Lots of ?? get tried, price includes all!. buy. Buy Low, Sell High. (All speculation is “gambling”!).

Download Presentation

Decision Markets

An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Presentation Transcript


Decision markets l.jpg

Decision Markets

Robin Hanson

Department of Economics

George Mason University


Buy low sell high l.jpg

“Pays $1 if

Bush wins”

Will price

rise or fall?

sell

E[ price change | ?? ]

buy

price

sell

Lots of ?? get tried,

price includes all!

buy

Buy Low, Sell High

(All speculation is “gambling”!)


Today s prices l.jpg

Today’s Prices

10-12% GMU wins DC NCAA bracket

60-62% Bird Flu in US by Sept 30

17-20% Hamas Recognize Israel by ‘07

11-13% Bin Laden caught by end ‘07.

3-7% Soc. Sec. private account bill by end ‘06.

44%  Hillary Democrat Pres. Nominee in ‘08.

47-49% Democrat President in ‘08.

TradeSports.com


In direct compare beats alternatives l.jpg

In direct compare, beats alternatives

  • Vs. Public Opinion

    • I.E.M. beat presidential election polls 451/596 (Berg et al ‘01)

    • Re NFL, beat ave., rank 7 vs. 39 of 1947 (Pennock et al ’04)

  • Vs. Public Experts

    • Racetrack odds beat weighed track experts (Figlewski ‘79)

      • If anything, track odds weigh experts too much!

    • OJ futures improve weather forecast (Roll ‘84)

    • Stocks beat Challenger panel (Maloney & Mulherin ‘03)

    • Gas demand markets beat experts (Spencer ‘04)

    • Econ stat markets beat experts 2/3 (Wolfers & Zitzewitz ‘04)

  • Vs. Private Experts

    • HP market beat official forecast 6/8 (Plott ‘00)

    • Eli Lily markets beat official 6/9 (Servan-Schreiber ’05)

    • Microsoft project markets beat managers (Proebsting ’05)


Growing corporations trying l.jpg

Growing # Corporations Trying


Slide6 l.jpg

Inputs

Outputs

Prediction

Markets

Theory

Foul Play

For Same

Compare!

Status Quo

Institution


Not experts vs self chosen amateurs l.jpg

Not Experts vs. Self-chosen Amateurs

  • Forecasting Institution Goal:

    • Given same participants, resources, topic

    • Want institution with accurate forecasts

  • Separate question: who let participate?

    • Can limit who can trade in market

  • Markets have low penalty for add fools

    • Hope: get more info from amateurs?


Estimates from prices l.jpg

$1 if A

p(A)

$1

$1 if A&B

p(A&B)

$1

$ x if A

E[x|A]*p(A)

$1

$ x

E[x]

$1

$1 if A&B

p(B|A)

$1 if A

$ x if A

E[x|A]

$1 if A

Estimates from Prices


School voucher decision advice l.jpg

School Voucher Decision Advice

$1 if Scores Up & Voucher

P(V)

$1 if

Voucher

P(SU | V)

$1

Compare!

P(SU | not V)

$1 if Not Voucher

$1 if Scores Up & Not Voucher

P(not V)


Decision market applications l.jpg

Decision Market Applications

E[ Test scores | School vouchers?]

E[ SA civil war | US moves troops out? ]

E[ Electricity consume | Price by hour ]

E[ Sea level | CO2 Treaty adopted ]

E[ firm stock price | fire CEO? ]

Market costs start high, not depend on topic, so do big value questions first.


Typical problems l.jpg

Typical Problems

  • Legal Barriers

  • Moral/Culture Concerns

  • Not really want to know

  • Hard to find precise related events

  • Can not get participation for cheap

  • Not enough events to validate, learn


Concerns l.jpg

Concerns

  • Self-defeating prophecies

  • Decision selection bias

  • Price manipulation

  • Inform enemies

  • Share less info

  • Combinatorics

  • Moral hazard

  • Alarm public

  • Embezzle

  • Bozos

  • Lies

  • Rich more “votes”

  • Risk distortion

  • Bubbles


Ask the right questions l.jpg

Ask the Right Questions

  • Cost independent of topic, but value not!

    • Seek high value to more accurate estimates!

    • Relevant standard: beat existing institutions

  • Where suspect more accuracy is possible

    • Suspect info is withheld, or not sure who has it

  • Prefer fun, easy to explain and judge

  • Prefer can let many know best estimates

    • Not fear reveal secrets, use fear to motivate

  • Avoid inducing foul play


Eight design issues l.jpg

Eight Design Issues

  • How avoid self-defeating prophecies?

  • How handle billions of possible combos?

  • What if bad guys lose $ to mislead us?

  • What if bad guys gain $ by give us info?

  • How not alarm public, inform enemies?

  • Price can mislead if deciders know more.

  • Will markets induce people to lie?

  • Will markets help employees embezzle?


Slide15 l.jpg

The Fuss:

Analysts often use prices from various markets as indicators of potential events. The use of petroleum futures contract prices by analysts of the Middle East is a classic example. The Policy Analysis Market (PAM) refines this approach by trading futures contracts that deal with underlying fundamentals of relevance to the Middle East. Initially, 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.

[Click here for a summary of PAM futures contracts]

The contracts traded on PAM will be based on objective data and observable events. These contracts will be valuable because traders who are registered with PAM will use their money to acquire contracts. A PAM trader who believes that the price of a specific futures contract under-predicts the future status of the issue on which it is based can attempt to profit from his belief by buying the contract. The converse holds for a trader who believes the price is an over-prediction – she can be a seller of the contract. This price discovery process, with the prospect of profit and at pain of loss, is at the core of a market’s predictive power.

The issues represented by PAM contracts may be interrelated; for example, the economic health of a country may affect civil stability in the country and the disposition of one country’s military may affect the disposition of another country’s military. The trading process at the heart of PAM allows traders to structure combinations of futures contracts. Such combinations represent predictions about interrelated issues that the trader has knowledge of and thus may be able to make money on through PAM. Trading these trader-structured derivatives results in a substantial refinement in predictive power.

[Click here for an example of PAM futures and derivatives contracts]

The PAM trading interface presents A Market in the Future of the Middle East. Trading on PAM is placed in the context of the region using a trading language designed for the fields of policy, security, and risk analysis. PAM will be active and accessible 24/7 and should prove as engaging as it is informative.

Became:


Pam press l.jpg

PAM Press

Of 500+ articles, these gave more favorable PAM impression:

Article: later in time, more words, mentioned insider, news (not Editorial) style, not anonymous

Publication: finance or science specialty, many awards, many readers


Info institutions features l.jpg

Academia

Expert Panel

News Media

Web & Blogs

Elections

Rumor Mills

Speculative Markets

Can ask direct question?

What $ per question?

How clear answer?

How precise, accurate?

How fast/often updates?

How open/equalitarian?

Incentive: truth or what folks want/expect hear?

Info Institutions & Features


Slide19 l.jpg

Cosmo constant > 0

Cancercuredby 2010

Science 291:987-988, February 9 2001


Hollywood stock exchange l.jpg

Hollywood Stock Exchange

Science 291:987-988, February 9 2001


Track odds beat handicappers l.jpg

Fourteen professional handicappers

Estimated using 146 races, tested on 46 races

Track Odds Beat Handicappers

Figlewski (1979) Journal of Political Economy


Economic derivatives market l.jpg

Economic Derivatives Market

Wolfers & Zitzewitz “Prediction Markets”

(2004) Journal of Economic Perspectives


Slide23 l.jpg

NFL Markets vs Individuals

Average of Forecasts

Servan-Schreiber, Wolfers, Pennock & Galebach (2004)

Prediction Markets: Does Money Matter? Electronic Markets, 14(3).

1,947 Forecasters


Iowa electronic markets vs polls l.jpg

Iowa Electronic Markets vs. Polls

“Accuracy and Forecast Standard Error of Prediction Markets”

Joyce Berg, Forrest Nelson and Thomas Rietz, July 2003.


Iowa electronic markets l.jpg

Iowa Electronic Markets

“Accuracy and Forecast Standard Error of Prediction Markets”

Joyce Berg, Forrest Nelson and Thomas Rietz, July 2003.


Beating official forecasts l.jpg

Beating official forecasts

  • Drug sales

  • ~12 traders

  • ~14 days

Source: eLilly & NewsFutures


Theory i old l.jpg

Theory I - Old

  • “Strong Efficient Markets” is straw man

  • No info - Supply and Demand

    • Assume beliefs not respond to prices

    • Price is weighted average of beliefs

    • More influence: risk takers, rich

  • Info, Static - Rational Expectations

    • Price clears, but beliefs depend on price

    • No trade if not expect “noise traders”

      • Price not reveal all info

    • More influence: info holders


Theory ii market microstruture l.jpg

Theory II - Market Microstruture

  • Info, Dynamic – Game Theory

  • Example – Kyle ’85

    • X - Informed trader(s) – risk averse

    • Y - Noise trader – fool or liquidity pref

    • Market makers – no info, deep pockets

      • If many compete, Price = E[value|x+y]

  • Info markets – use risk-neutral limit

    • If Y larger, X larger to compensate more info gathered, so more accuracy!


Theory iii behavioral finance l.jpg

Theory III – Behavioral Finance

  • Humans are overconfident

    • Far more speculative trade than need

    • Mere fact of disagreement shows

    • Overconfidence varies with person, experience, consequence severity

  • Implications

    • Price in part an ave of beliefs?

    • Adds noise to price aggregates?

    • Prices more honest than talk, polls, …


We can t agree to disagree l.jpg

Aumann in 1976

Re possible worlds

Common knowledge

Of exact E1[x], E2[x]

Would say next

For Bayesians

With common priors

If seek truth, not lie

Since generalized to

Impossible worlds

Common Belief

A f(•, •), or who max

Last ±(E1[x] - E1[E2[x]])

At core, or Wannabe

Symmetric prior origins

We Can’t Agree to Disagree


Explanation we self deceive l.jpg

Explanation: We Self-Deceive

  • We biased to think better driver, lover, …

    “I less biased, better data & analysis”

  • Evolutionary origin: helps us to deceive

    • Mind “leaks” beliefs via face, voice, …

    • Leak less if conscious mind really believes

  • Beliefs like clothes

    • Function in harsh weather, fashion in mild

  • When see our self-deception, still disagree

    • So at some level we accept that we not seek truth


Slide32 l.jpg

Rafi Eldor and Rafi Melnick (2004)

“Financial markets and terrorism”

European Journal of Political Economy

20:367–386


Terror concept red teams l.jpg

Terror Concept: Red Teams

  • Key: want enough events to validate tech, let participants see their ability

  • Red Team tries to get items onto planes

  • Vary: item, demographic, airport, time

    • Chosen by team, or at random

  • Forecast chances of red team success

    • Given variation, or changed policy, budget

    • Invite others to also forecast, compare


1 self defeating prophecies l.jpg

1. Self-Defeating Prophecies

  • Example

    • Speculators learn of possible airport attack

    • Market price of attack chance rises

    • Airport officials see price, close airport

    • Speculators are punished for their insight

  • Solution: predict outcome given choices

    • Predict terror attack given airport open


2 useful attack forecasts distinguish l.jpg

2. Useful attack forecasts distinguish:

  • Thousands of possible locations

  • Hundreds of possible times

  • Dozens of methods of attack

  • Dozens of terrorism demographics

  • Dozens of policy responses

  • Many combinations for each attack

  • More combinations for many attacks


Slide36 l.jpg

Policy Analysis Market

  • Every nation*quarter:

  • Political stability

  • Military activity

  • Economic growth

  • US $ aid

  • US military activity

  • & global, special

  • & all combinations


Trade scenario l.jpg

Return to Focus

?

Trade

IQcs4

IQcs4

<

85

85

03

03

SAum3

105-125

03

Update

Payoffs:

If

&

Ave. pay

Select

New Price

65%

Max Up

95.13%

+$34.74

-$85.18

-$19.72

Buy

10% Up

68.72%

+$2.74

-$3.28

-$1.07

You Pick

65

%

+1.43

-2.04

+0.34

Saudi Arabian

Economic Health

No Trade

62.47%

$0.00

$0.00

$0.00

125

30

15

10% Dn

56.79%

-$2.61

+$2.74

-$1.12

65

70

Sell

Exit Issue

48.54%

-$15.34

+$26.02

-$6.31

35

40

100

94

100

Max Dn

22.98%

-$120.74

+$96.61

-$22.22

<

85

25

35

35

30

10

10

75

1

2

3

4

1

2

>

03

03

03

03

04

04

?

Return to Form

Execute a Trade

If US military involvement in Saudi Arabia in 3rd Quarter 2003 is not between 105 and 125, this trade is null and void. Otherwise, if Iraq civil stability in 4th Quarter 2003 is below 85, then I will receive $1.43, but if it is not below 85, I will pay $2.04.

Abort trade if price has changed?

Execute

Trade Scenario


Best of both l.jpg

Simple Info Markets

Market Scoring Rules

Scoring

Rules

opinion

pool

problem

thin market

problem

100

.001

.01

.1

1

10

Best of Both

Accuracy

Estimates per trader


Laboratory tests l.jpg

Laboratory Tests

  • Joint work with John Ledyard (Caltech), Takashi Ishida (Net Exchange)

  • Caltech students, get ~$30/session

  • 6 periods/session, 12-15 minutes each

  • Trained in 3var session, return for 8var

  • Metric: Kulback-Leibleri qilog(pi /qi) distance from market prices to Bayesian beliefs given all group info


Mechanisms compared l.jpg

Mechanisms Compared

  • Survey Mechanisms (# cases: 3var, 8var)

    • Individual Scoring Rule (72,144)

    • Log Opinion Pool (384,144)

  • Market Mechanisms

    • Simple Double Auction (24,18)

    • Combined Value Call Market (24,18)

    • MSR Market Maker (36,17)


Environments goals training l.jpg

Environments: Goals, Training

(Actually: X Z Y )

Case A B C

1 1 - 1

2 1 - 0

3 1 - 0

4 1 - 0

5 1 - 0

6 1 - 1

7 1 - 1

8 1 - 0

9 1 - 0

10 0 - 0

Sum: 9 - 3

Same A B C

A -- -- 4

B -- -- --

C -- -- --

  • Want in Environment:

    • Many variables, few directly related

    • Few people, each not see all variables

    • Can compute rational group estimates

    • Explainable, fast, neutral

  • Training Environment:

    • 3 binary variables X,Y,Z, 23 = 8 combos

    • P(X=0) = .3, P(X=Y) = .2, P(Z=1)= .5

    • 3 people, see 10 cases of: AB, BC, AC

    • Random map XYZ to ABC


Msr info vs time 7 prices l.jpg

KL(prices,group)

1-

KL(uniform,group)

MSR Info vs. Time – 7 prices

1

% Info Agg. =

0

0 5 10 15

Minutes

-1


Experiment environment l.jpg

Experiment Environment

(Really: W V X S U Z Y T )

Case A B C D E F G H

1 0 1 0 1 - - - -

2 1 0 0 1 - - - -

3 0 0 1 1 - - - -

4 1 0 1 1 - - - -

5 0 1 1 1 - - - -

6 1 0 0 1 - - - -

7 0 1 1 1 - - - -

8 1 0 0 1 - - - -

9 1 0 0 1 - - - -

10 1 0 0 1 - - - -

Sum 6 3 4 10 - - - -

Same A B C D E F G H

A -- 1 2 6 -- -- -- --

B -- -- 7 3 -- -- -- --

C -- -- -- 4 -- -- -- --

D -- -- -- -- -- -- -- --

  • 8 binary vars: STUVWXYZ

  • 28 = 256 combinations

  • 20% = P(S=0) = P(S=T) = P(T=U) = P(U=V) = … = P(X=Y) = P(Y=Z)

  • 6 people, each see 10 cases: ABCD, EFGH, ABEF, CDGH, ACEG, BDFH

  • random map STUVWXYZ to ABCDEFGH


Msr info vs time 255 prices l.jpg

KL(prices,group)

1-

KL(uniform,group)

MSR Info vs. Time – 255 prices

1

% Info Agg. =

0

0 5 10 15

Minutes

-1


Combo market maker best of 5 mechs l.jpg

Combo Market Maker Best of 5 Mechs


Overlapping patches l.jpg

D

A

C

G

F

B

E

H

Overlapping Patches

  • Simple all-combo MSR per patch

  • Allow trade only if vars in same patch

  • Is Markov network, much research re update

    • Exact update rule possible, but only if net is tree

    • But any predictable error allows money pump

      • Monte Carlo sampling update solves this?

    • Arbitrage neighbors is simple, robust, not Bayes

  • Adapt tech to share assets across patches

  • How change or add patches?


3 manipulation l.jpg

3. Manipulation

  • Lie (WorldCom) – hits all forecast institutions

  • Mix-it-up – slow reveal trade to max $ of info

  • Mislead – Terrorists lose $ to raise price error

    • Even if works, still useful if can calibrate accuracy

    • 0.1% prob. gain 0.1% of $500B/yr pays $1M PAM

    • Field, Lab agree: on ave manip. fails

    • Any trade not to info is “noise” trade

    • More noise trading => more accuracy!

    • Less error, more harm? Use price/harm linear


Simple manipulation model l.jpg

Simple Manipulation Model

Kyle Style Market Microstructure Game Theory

Market maker

Manipulator

Informed trader

Noise trader

Equilibrium


Lab experiments confirm l.jpg

Lab Experiments Confirm

Joint with David Porter, Ryan Oprea

  • 12 subjects, V=0,40,100

  • Each clue like “Not 100”.

  • Manipulator = price bonus

  • Manipulators bid higher

  • Others accept lower

  • Prices no less accurate


4 moral hazard l.jpg

4. Moral Hazard

  • Ever trading profits from sabotage?

    • Not: 9/11, ’82 Tylenol, ’02 PaineWebber

    • Yes: hacker extortion, life insurance kill

    • Hard to match willing capital & skilled labor

  • Terrorism futures differ:

    • Good: amounts of money very small

      • We’d pay $10 to know where robber hits next

    • Bad: individuals influence events more?


Solutions for high stakes l.jpg

Solutions For High Stakes

  • Only play money, or winnings to charity

  • Limit who can play to trusted traders

    • Certified by reduce privacy?

  • Let investigators see trading records

  • Limit trading positions

    • E.g., gain no more than $20 if attack


5 hiding prices l.jpg

5. Hiding Prices

  • Prices may alarm public, inform bad guys

  • Traders not now know prices when order

    • Deal via: Limit orders, market orders, …

  • Can extend this to show less about price

    • Might hide prices re attack a week before

    • Orders specify how respond to price history

    • Trusted traders might see actual prices


6 decision selection bias l.jpg

6. Decision Selection Bias

  • If traders think deciders will use info that traders do not have

  • Then market advice can contradict trader info

  • Solution:

    • Some w/ decider info trade (or price decide)

    • Make decision time clear to traders


Slide54 l.jpg

Change

Chance

attack

if no

policy

change

Apparent

center

Do not

change

True

center

Chance attack if policy change


Problem seems uncommon l.jpg

Problem Seems Uncommon


7 lies damn lies and statistics l.jpg

7. Lies, Damn Lies (and Statistics)

  • E.g., WorldCom, other accounting scandals

  • Participants in, advisors out

    • Participants contribute directly to forecast

    • Advisors try to persuade participants

  • Advise via shared interest, track record

    • Forecasting institution seems irrelevant

  • But if advisors can bet, more reason to lie?

    • Let advisors show bet stake

    • Expect advisors to use info via bet, not talk?


8 embezzle l.jpg

8. Embezzle

  • Play money: Why employees bother?

  • Real money: Shirk to trade? Topics to pay off favorites? Withhold info from team?

  • Try: new “color of money”

    • Employees, teams give initial accounts

    • Subsidize each topic at estimated info $ value

    • Consistent increase credited as $ to org.

    • Team account trades on team info first

    • Special accounts trade standard info streams


Terrorism futures seem feasible technically if not politically l.jpg

Terrorism Futures Seem Feasible, Technically If Not Politically

0) Not about experts vs. self-chosen amateurs

  • How avoid self-defeating prophecies?

  • How handle billions of possible combos?

  • What if terrorists lose $ to mislead us?

  • What if terrorists gain $ by give us info?

  • How not alarm public, inform terrorists?

  • Price can mislead if deciders know more.


Timeline policy analysis market l.jpg

Timeline - Policy Analysis Market

2000

DARPA conceives info market project

Bush pres., RFP, begin Phase I ($100K)

Poindexter heads IAO, with pet TIA

Spring – start 2yr. Phase II ($750K)

May – Congress report ex: Israel Bio Terror

Plan: register 8/1, few test 9/1, open 1/1

7/28 – Senators complain: “repugnant"

7/29 – DARPA FutureMAP Killed

7/30 – Poindexter quits

2001

2002

2003


Terror betting is unthinkable l.jpg

Terror Betting Is “Unthinkable”

  • We deny trade sacred for secular (P. Tetlock)

    • E.g., money vs. risk of death

    • Outraged at those we see do (really, we all do)

    • More outraged if they think a long time first

    • If catch ourselves, seek moral cleansing

  • PAM painted as crossing a moral boundary

    • “None of us should intend to benefit when they hurt some of us.”

    • So politicians must not think long before rejecting


Pam concerns l.jpg

PAM Concerns

Terrorists themselves could drive up the market for an event they are planning and profit from an attack, or even make false bets to mislead intelligence authorities. U.S. Senators Wyden and Dorgan, Press Release, July 28, 2003.

Would-be assassins and terrorists could easily use disinformation and clever trading strategies to profit from their planned misdeeds while distracting attention from their real target. Steven Pearlstein, Washington Post, July 30, 2003.

Trading . . . could be subject to manipulation, particularly if the market has few participants – providing a false sense of security or . . . alarm. Joseph Stiglitz, Los Angeles Times, July 31, 2003.


  • Login