U-Mart2002

Many students including those from Osaka Sangyo University and Chuo University participated. Students of Osaka Sangyo University and Chuo University had experienced trading using U-Mart many times in classes. So those universities regarded that this open experiment was an opportunity to understand the results of their educations. After the experiment, a joint seminar was organized by participant universities for academic exchange among students.

¥Date
  November 5, 2002
¥Place
  Department of Economics,
Osaka City University Co-hosted by SICE and Japan@Society of System@Engineering
¥Participants
  7 teams with machine agents and
12 teams@with human agents
¥Agent type
  Machine and Human

  1. # Sessions per day: 8 times
  2. Session interval: 10 seconds
  3. Trading period: 24
  4. Total # Sessions: 192 times

yBreakdown of human agentsz
¡Osaka Sangyo University: @8 teams, total 24 member
yAmong thosez: Undergraduates who had used the U-Mart System in class for half a year, and Several "speculators"(maybe)
¡Kyoto University: @1 team, total 3 member
yAmong thosez:@Graduates and undergraduates who were not familiar with the U-Mart system and futures market trading.
¡Chuo University: 1 team, total 3 member
yAmong thosez:@Undergraduates who had experienced the U-Mart System and developed machine agents

yMachinez

Team University Member ID Description
Ishiyama Chuo University m1`m3 EStrategic trading posing as an amateurEin the market: Kawasaki
EDecisions based on George Sorosfs Theory of Reflectivity and Williams %R: Kim
EDecisions based on the Oscillation Trend: Ishiyama
Kobayashi Chuo University m4`m6 EStochastic Theory (revised): Harada
E Day-to-Day Price Movements Psychological Line: Kobayashi
EStochastic Theory: Nakata
Nakajima Osaka City University m7`m9 E Kaubakka
E Price Maker
E Selling/buying according to price range
Sawa Osaka City University m10`m12 ESophisticated clients with three agents working in coordination
EEach agent had 7 strategies:
ECountermeasure against losing at the last count, semi-simple regular siege, RSI Analysis, series method, short-to-medium-term average method, regular siege, Williams %R
E Duration of search: 14 days
E Each participant implemented every strategy --> Results were reported to the server.
EDuration of using optimum strategy: 8 days
E After trials of the 7 strategies, decided the best one and used it.
E Duration of taking countermeasure against losing at the last count: 2 days
E Tried to secure position near zero
Kobayashi Tokyo University m13`m15 EDay trade type
E Trend type
E Pseudo-arbitrage Type
Kanai Osaka City University m16`m18 E Arbitrage trading between spot and futures
ESimple averaging sell
E Averaging buy employing dollar cost averaging method
ETeam working: When the market was on its upward course, (3) would make profit and (2) would hedge loss, and in opposite condition, those would take opposite roles respectively.(1) would make a profit under any condition only if an arbitrage opportunity given.
Ariyama Osaka Prefecture University m19`m21 EOn-line fuzzy learning A
E On-line fuzzy learning B
E Neural network

yHuman agentsz

@
Team University Member ID Member name
Taniguchi A Osaka Sangyo University m31`m33 Ogasawara, Tabuchi, Inoue
Fudaikeihokonsei Osaka Prefecture University &
Osaka University of Economics and Law
m36`m38 Ariyama, Fukase, Kitano
Taniguchi G Osaka Sangyo University
m39`m40 IharaCTakachi
Chuo 2002 Chuo University m34Em41Em42 Nakata, Kobayashi, Harada
Chuo ‚P Chuo University m44`m46 Kim, Kawasaki, Ishiyama
Taniguchi F Osaka Sangyo University m47`m49 Goto, Yokoyama, Kato
The Sai Osaka Sangyo University m50Em60Em61 Sai, Ichikawa, Fujii
Kyotodai Kyoto University m51`m53 Shinagawa, Endo, Lee
Taniguchi B Osaka Sangyo University m55`m57 Irifune, Sugihashi, Matsuo
Taniguchi B Osaka Sangyo University m63`m65 Kubosaki, Tanaka, Sen
SUPER M Osaka Sangyo University m67`m69 Emura, Sakamoto, Okoshi
TaniguchiE Osaka Sangyo University m71`m73 Ota, Maekawa, Hayashi

œAgent

Rank Agent name Member ID Team University CASHi\j
No.1 Ariyama02 m20 Ariyama Osaka Prefecture University
7,412,344,000
No.2 Ariyama01 m19 Ariyama Osaka Prefecture University
5,226,312,000
No.3 Kato m49 Taniguchi F Osaka Sangyo University
No.4 Tabuchi m32 Taniguchi A Osaka Sangyo University

œTeam

Rank Agent name MemberID Team University
No.1 Ogasawara, Tabuchi, Inoue m31`m33 Taniguchi A Osaka Sangyo University
No.2 IharaETakachi m36`m40 Taniguchi G Osaka Sangyo University
No.3 Team_TK01`03 m13`m15 Team_TK Tokyo University
No.4 Isiyama01 m 1`m 3 Ishiyama Chuo University
No.5 ItaEyoseEcom01`03 ‚10`m12 ItaEyoseEcom Tokyo Institute of Technology

yU-Mart2002z
Ÿm 19, 20, 21 (Team Ariyama, Osaka Prefecture University)were @prominent (including a bankrupted agent in the count).
yAgents in the blackz@Machine 14/23 (60%)@Human 16/35 (46%)
yBankrupted Agentsz@Machine 2/23 (9%)@Human 3/35 (9%)
Total 5/58 (7%)

 

yResult of U-Mart 2002z