Do individual judgment biases diminish in markets? [Elektronische Ressource] : the case of partition dependence in prediction markets / vorgelegt von: Ulrich Sonnemann

Do individual judgment biases diminish in markets? The case of partition-dependence in prediction markets Inauguraldissertation zur Erlangung des akademischen Grades eines Doktors der Wirtschaftswissenschaften durch die Wirtschaftswissenschaftliche Fakultät der Westfälischen Wilhelms-Universität Münster vorgelegt von: Ulrich Sonnemann aus Hannover 2008 Dekan: Prof. Dr. Stefan Klein Erster Berichterstatter: Prof. Dr. Thomas Langer Zweiter Berichterstatter: Prof. Dr. Dr. h.c. Klaus Backhaus Tag der mündlichen Prüfung: 18. November 2008 Table of Contents III Table of Contents List of Tables ..........................................................................................................VI List of Figures .....................................................................................................VIII List of Abbreviations ............................................................................................... X List of Symbols ....................................................................................................XIII 1 Introduction..............................................................................................................1 2 Fundamental concepts and literature review........................................................13 2.1 Prediction markets.............................................
Publié le : mardi 1 janvier 2008
Lecture(s) : 29
Source : MIAMI.UNI-MUENSTER.DE/SERVLETS/DERIVATESERVLET/DERIVATE-4871/DISS_SONNEMANN.PDF
Nombre de pages : 311
Voir plus Voir moins






Do individual judgment biases diminish in markets?
The case of partition-dependence
in prediction markets





Inauguraldissertation zur Erlangung des akademischen Grades
eines Doktors der Wirtschaftswissenschaften durch die
Wirtschaftswissenschaftliche Fakultät der
Westfälischen Wilhelms-Universität Münster









vorgelegt von:

Ulrich Sonnemann
aus Hannover

2008




























Dekan: Prof. Dr. Stefan Klein

Erster Berichterstatter: Prof. Dr. Thomas Langer
Zweiter Berichterstatter: Prof. Dr. Dr. h.c. Klaus Backhaus
Tag der mündlichen Prüfung: 18. November 2008 Table of Contents III

Table of Contents

List of Tables ..........................................................................................................VI
List of Figures .....................................................................................................VIII
List of Abbreviations ............................................................................................... X
List of Symbols ....................................................................................................XIII
1 Introduction..............................................................................................................1
2 Fundamental concepts and literature review........................................................13
2.1 Prediction markets........................................................................................13
2.1.1 Basic principles...........................................................................................13
2.1.1.1 Definition............................................................................................13
2.1.1.2 Applications14
2.1.1.3 Types of contracts ...............................................................................15
2.1.1.4 Selected market design issues .............................................................17
2.1.2 Forecast accuracy........................................................................................19
2.1.3 Arbitrage strategies.....................................................................................20
2.1.4 Theoretical background21
2.1.4.1 Hayek hypothesis................................................................................21
2.1.4.2 Information paradox and the “no-trade theorem” ...............................23
2.1.4.3 Interpretation of prices as market-derived probabilities .....................25
2.1.4.4 Individual judgment errors and market outcomes30
2.2 The description invariance principle and partition-dependence..............37
2.2.1 Preliminaries...............................................................................................37
2.2.2 Violations of the description invariance principle ......................................39
2.2.2.1 Violations in risky situations...............................................................39
2.2.2.2 Violations in uncertain situations........................................................40
2.2.3 Explanations for violations of description invariance.................................45
2.2.4 Support theory.............................................................................................50
2.2.4.1 The model............................................................................................50
2.2.4.2 Further empirical evidence..................................................................57
Table of Contents IV
2.3 Experimental asset markets .........................................................................60
2.3.1 Basic principles...........................................................................................60
2.3.2 Experimental tests of market efficiency......................................................66
2.3.2.1 Preliminaries.......................................................................................66
2.3.2.2 Information dissemination..................................................................67
2.3.2.3 Information aggregation and dissemination........................................70
2.3.2.4 Information production.......................................................................72
2.3.3 Experimental tests of individual judgment errors and market outcomes....74
3 Study 1: Laboratory study......................................................................................77
3.1 Basis treatment..............................................................................................77
3.1.1 Study design................................................................................................77
3.1.2 Descriptive analysis....................................................................................85
3.1.3 Main results.................................................................................................92
3.1.3.1 Judged probabilities............................................................................92
3.1.3.2 Equilibrium market prices...................................................................96
3.1.3.3 Comparison of market prices to individual probability judgments...103
3.1.3.4 1/N ignorance prior and consistency of prices with a theoretical
benchmark.........................................................................................109
3.1.4 Second order results..................................................................................113
3.1.4.1 Competence effects...........................................................................113
3.1.4.2 Trader-based analysis........................................................................121
3.2 Information treatment................................................................................133
3.2.1 Motivation and study design.....................................................................133
3.2.2 Main results and impact of information transmission...............................138
3.3 Interim conclusions .....................................................................................152
4 Study 2: An NBA/FIFA field experiment ...........................................................154
4.1 Experimental design....................................................................................154
4.2 Main results .................................................................................................161
4.2.1 Judged probabilities..................................................................................161
4.2.2 Market prices.............................................................................................164
4.2.3 Hypothetical “pseudo-arbitrage” strategies ..............................................168
4.2.3.1 Preliminaries.....................................................................................168 Table of Contents V
4.2.3.2 Bid-ask pseudo-arbitrage..................................................................168
4.2.3.3 Interpolated-price hypothetical arbitrage ..........................................173
4.3 Second order results....................................................................................179
4.4 Interim conclusions .....................................................................................191
5 Study 3: Naturally-occurring prediction markets for economic derivatives......193
5.1 Institutional structure of economic derivatives........................................193
5.2 Ignorance-prior analysis: The “mixture” model......................................198
6 Conclusion............................................................................................................205
7 Appendices209
8 References281

List of Tables VI

List of Tables

Table 2.1: Contract types: Estimating uncertain quantities or probabilities
(Wolfers and Zitzewitz (2004, p. 110, Table 1)). ......................................16
Table 3.1: Sets of asset partitions for the different treatments in Study 1. .................79
Table 3.2: Breakdown of the total number of trades...................................................86
Table 3.3: Trading volume statistics. ..........................................................................86
Table 3.4: Temporal distribution of trades in a 10-minute trading period..................87
Table 3.5: Mean standard deviation of trade prices (based on all trades)...................88
Table 3.6: Mean standard deviation of trade prices (based on last three trades). .......88
Table 3.7: Summary statistics on bid arbitrage opportunities.....................................89
Table 3.8: Summask arbitrage opportunities. ...................................90
Table 3.9: Time-weighted average of bid and ask sums.............................................91
Table 3.10: Pre-trading individual probability judgments............................................94
Table 3.11: Post-trading individual probability judgments. .........................................95
Table 3.12: Measures of price convergence over time. ................................................99
Table 3.13: Mean equilibrium market prices (2nd trading round) and
individual judgments (pre-trading and post-trading). ..............................102
Table 3.14: trade prices (actual/standardized),
1/N ignorance prior, and theoretical benchmark......................................112
Table 3.15: Interaction effects between partition type and competence level
in pre- and post-trading probability judgments........................................116
Table 3.16:
in equilibrium market prices for the sports event domain. ......................120
Table 3.17: Mean expected profits per unit of traded asset. .......................................127
Table 3.18: Differences in the average number of trades (Panel A) and
the average trade volume (Panel B) per trader by competence level.......128
Table 3.19: Differences in sure (minimum) payoffs, maximum possible payoffs,
and payoff ratio by competence level of traders......................................131
Table 3.20: Subjectively expected payoff based on individual judgments by
competence level of traders. ....................................................................132 List of Tables VII
Table 3.21: Sets of asset partitions for the different treatments in Study 1
(information treatment)............................................................................134
Table 3.22: Mean equilibrium market prices (2nd trading round) and
individual judgments (pre-trading and post-trading) in the
information treatment of Study 1.............................................................139
Table 3.23: Interaction effects between partition type and treatment
(basis vs. information treatment) in pre- and post-trading
probability judgments. .............................................................................140
Table 3.24: Measures of price convergence over time (info treatment). ....................141
Table 3.25: Interaction effects between partition type and treatment
(basis vs. information treatment) in mean equilibrium market prices. ....143
Table 3.26: Results of hypotheses tests (3.7.(i)–3.7.(iv)). ..........................................151
Table 4.1: Partition-dependence in before-trading judgments for
NBA Playoffs events. ..............................................................................162
Table 4.2: ents for
FIFA World Cup events...........................................................................163
Table 4.3: Per-day profitability of hypothetical bid/ask pseudo-arbitrage
strategies. .................................................................................................173
Table 4.4: Per-day profitability of hypothetical interpolated-price
pseudo-arbitrage strategies. .....................................................................179
Table 4.5: Interaction effects between partition type and recruitment channel
in before-trading probability judgments for each NBA team. .................182
Table 4.6: Interaction effent channel ility judgments, pooled over four-team
markets (market clones)...........................................................................186
Table 4.7: Interaction effects between partition type and self-rated competence
in before-trading probability judgments for each FIFA World Cup
team..........................................................................................................188
Table 4.8: Interaction eff-rated competence
in before-trading probability judgments, pooled over four-team
markets (market clones)...........................................................................190
Table 5.1: Results of regressions of forecast errors on the difference between
observed forecast and 1/N forecast. .........................................................200 List of Figures VIII

List of Figures

Figure 2.1: (Unpruned) fault tree with possible reasons why a car might fail
to start as used by Fischhoff, Slovic, and Lichtenstein (1978);
taken from Fox and Clemen (2005, p. 1418, Figure 1). ............................42
Figure 3.1: Construction of assets for two DAX partitions..........................................81
Figure 3.2: Screenshot of the trading application (practice market). ...........................83
Figure 3.3: Example of the time course of an experimental session.85
Figure 3.4: Development of price differences over time for the finance assets...........97
Figure 3.5: ent of e for the sports assets. ............97
Figure 3.6: Development of e for the weather assets. .........98
Figure 3.7: (Summed) equilibrium market prices for the two partition markets
(intervals I and I )...................................................................................101 1 2
Figure 3.8: rium ma partition markets
(intervals I and I ).101 3 4
Figure 3.9: Frequency distributions for the difference (after-before)
in judgments for packed events. ..............................................................107
Figure 3.10: Frequency distribution of subjects’ self-rated competence. ....................114
Figure 3.11: Plots of cell means for judgments in the finance event domain
(before- and after-trading). ......................................................................117
Figure 3.12: Frequency distribution of the number of trades per trader (buy & sell). .122
Figure 3.13: Frequency distribution of the proportion of trades per trader
(trade concentration metric).....................................................................123
Figure 3.14: Message window in Study 1 (information treatment). ............................135
Figure 3.15: Overall time-series price chart (info treatment) by slot
(domain: finance, low intervals). .............................................................144
Figure 3.16: Overall time-series prient) by slot
(domain: finance, high intervals). ............................................................144
Figure 3.17: Overall time-series price chart (info treatment) by slot
(domain: sports, low intervals). ...............................................................145
Figure 3.18: Overall time-series prient) by slot
(domain: sports, high intervals). ..............................................................145 List of Figures IX
Figure 3.19: Overall time-series price chart (info treatment) by slot
(domain: weather, low intervals). ............................................................146
Figure 3.20: Overall time-series prient) by slot
(domain: weather, high intervals). ...........................................................146
Figure 3.21: Measures of partition-dependence in the information treatment. ............148
Figure 4.1: Construction of asset partitions (NBA Playoffs victory totals). ..............156
Figure 4.2: Construction of asset partitions (FIFA World Cup goal totals).156
Figure 4.3: Study homepage.......................................................................................158
Figure 4.4: Screenshot of the trading application used in Study 2 (practice market).159
Figure 4.5: Illustration of the experimental setup of Study 2.....................................160
Figure 4.6: Price chart (Dallas Mavericks, DAL). .....................................................165
Figure 4.7: Price chart (Miami Heat, MIA)................................................................166
Figure 4.8: Price chart (Germany, GER)....................................................................167
Figure 4.9: t (Italy, ITA).............................................................................167
Figure 4.10: Hypothetical cross-market pseudo-arbitrage (Dallas Mavericks, DAL). 171
Figure 4.11: arket pseudo-arbitrage (Italy, ITA)........................172
Figure 4.12: Interpolated-price hypothetical arbitrage (Dallas Mavericks, DAL).......176
Figure 4.13: Interpolated-price hypothetical arbitrage (Italy, ITA). ............................177
Figure 5.1: Digital option prices on ranges of retail trade statistics
(taken from Gürkaynak and Wolfers (2006)). .........................................197
Figure 5.2: Illustration of a negative forecast error (M – X) < 0 if M > M . ....200 obs obs 1/N
Figure 5.3: Error Minimization (MAE vs. Lambda). .................................................202
List of Abbreviations X

List of Abbreviations

ANOVA Analysis of Variance
ARG Argentina
AUS Australia
BRA Brazil
Caltech California Institute of Technology
CASSEL California Social Science Experimental Laboratory
CBoT Chicago Board of Trade
CDA Continuous Double Auction
CHI Chicago Bulls
CIV Côte d’Ivoire
CLE Cleveland Cavaliers
CME Chicago Mercantile Exchange
CPI Consumer price index
CRC Costa Rica
CRO Croatia
CRRA Constant Relative Risk Aversion
CZE Czech Republic
DAL Dallas Mavericks
DAS Decision Analysis Society
DEN Denver Nuggets
DET Detroit Pistons
DJIA Dow Jones Industrial Average
ECU Ecuador
ED Economic derivatives
EMH Efficient Market Hypothesis
EUT Expected utility theory
FAQs Frequently Asked Questions
FIFA Fédération Internationale de Football Association
GBM Geometric Brownian Motion
GDP Gross domestic product

Soyez le premier à déposer un commentaire !

17/1000 caractères maximum.