An Early Agent-Based Stock Market: Replication & Participation


Download 445 b.
Sana25.08.2018
Hajmi445 b.


An Early Agent-Based Stock Market: Replication & Participation

  • László Gulyás (gulyas@sztaki.hu) Computer and Automation Research Institute Hungarian Academy of Sciences

  • Balázs Adamcsek AITIA, Inc. Budapest, Hungary

  • Árpád Kiss AITIA, Inc. Budapest, Hungary Loránd Eötvös University, Budapest


Overview

  • Motivation for…

    • Agent-Based Modeling
    • Experimental Economics
    • Participatory Simulation
  • The Early Santa Fe Artificial Stock Market

  • Results:



Agent-Based Modeling (ABM)

  • A form of computational modeling.

  • Aiming at creating complex (social) behavior “from the bottom up”.

    • Complex interactions of
    • A high number of
    • (Complex) individuals.
  • A generative and mostly theoretical approach:

    • Computational “thought experiments”,
    • Existence proofs, etc.


Experimental Economics

  • Controlled laboratory experiments with human subjects.

    • The effect of human cognition on economic behavior.
    • Learning and adaptation.
    • Social traps, etc.
  • Typical tools:

    • Observation (Videotaping)
    • Questionnaires, etc.
  • An experimental approach.



Participatory Simulation (PS)

  • A computer simulation, in which human subjects also take part.

  • Agent-based simulations are well suited:

    • Individuals are explicitly modeled, with
    • Strict Agent-Environment and Agent-Agent boundaries.
  • Bridges the theoretical and experimental approaches. Can help both of them:

    • Testing assumptions and results of an ABM.
    • Generating specific scenarios (e.g., crowd behavior) for laboratory experiments.


Summary of the project

  • Replication of a famous ABM in finance.

    • Replication of results is a most important step in science!
  • Conversion to a PS.

    • Partly as a demonstration of our General-Purpose Participatory Architecture for RePast (GPPAR).
  • Initial Experiments, testing:

    • Original results’ sensitivity to human trading strategies.
    • Human versus computational economic performance.
    • The effect of human learning between runs.


The Santa Fe Artificial Stock Market 1/2

  • A prominent model of agent-based finance (Arthur, Holland, LeBaron, Palmer and Tayler, 1994.)

  • A minimalist model of two assets:

    • “Money”: fixed, risk-free, infinite supply, fixed interest.
    • “Stock”: unknown, risky behavior, finite supply, varying dividend.
  • Artificial traders

    • Developing trading strategies.
    • In an attempt to maximize their wealth.


The Santa Fe Artificial Stock Market 2/2

  • Two distinct behavioral regimes:

    • One:
      • Consistent with Rational Expectations Equilibrium.
      • Price follows fundamental value of stock.
      • Trading volume is low.
    • The other:
      • “Chaotic” market behavior.
      • “Crashes” and “bubbles”: price oscillates around fundamental value.
      • Trading volume shows wild oscillations.
      • Appears to be “in accordance” with actual market behavior.


The Early SFI-ASM 1/4

  • The most known version of the SFI-ASM was published in 1997, after several years of work.

  • However, a first, simpler design was published in 1994. It has

    • Less realistic market mechanisms.
    • Simpler trading strategies for agents.
  • We were working with the early version.



The Early SFI-ASM 2/4

  • Dividend:

    • A stochastic (Ohrnstein-Uhlenbeck) process.
  • Possible Actions:

    • Selling/Buying one share,
    • Or holding.
  • Market Clearing:

    • A rationing scheme (agents may only get a fraction of their bids).
    • May yield fractional shares.


The Early SFI-ASM 3/4

  • Agents:

    • 60 ‘trading rules’:
      • Specifying actions (buy, sell, hold) based on market indicators:
        • Price > Fundamental Value, or
        • Price < 100-period Moving Average, etc.
      • Reinforced if their ‘advice’ would have yielded profit.
    • A Genetic Algorithm
      • Activated in Poisson-distributed intervals (individually for each agent).
      • Replaces 10-20% of weakest the rules.


The Early SFI-ASM 4/4

  • Trading rules:(condition, action, strength)

  • Action:

    • Buy, Sell, Hold
  • Condition:

    • Ternary string:
    • 110*1***0
    • Matching the binary (true/false) string of market indicators.
  • A classifier system.



Replication Results 1/4

  • Our implementation confirms those reported in the original publication.

  • The interesting case is that of a complex system, which yields

    • Market volatility and high volume.
    • Agents’ strategies grow diverse.


Replication Results 2/4



Replication Results 3/4

  • Wealth is only a sign of the agents’ heterogeneity.

  • What is the underlying reason?

    • Different trading strategies.
  • Measure:

    • The average number of “used” (non-*) bits in the rule set.


Replication Results 4/4

  • Concluding remarks:

    • The agent community learned to ‘manipulate’ price in such a way that it follows FV. (Subject to a certain range of error.)
    • Agents “self-organize” (i.e., mutually adapt) to achieve this.
    • However, heterogeneity suggests that some learned to be smart, while others learned to “sacrifice” their wealth.


Participatory ASM: Questions

  • Can agents adapt to external trading strategies, just as well as they did to those developed by fellow agents?

  • Would the apparently complex market behavior appear so to human players? Or would they easily learn to control the market?

  • Will computational agents outperform humans, particularly in a fast game?

  • What effect would human learning between sessions play on the outcome?



Participatory ASM: Implementation, Design

  • The illusion of a ‘real market’:

    • A fast, ‘real-time’ game.
  • Based on the General-Purpose Participatory Architecture for RePast (GPPAR).

    • Can be used to transform arbitrary ABMs to participatory simulation.
    • Networked execution.
    • Extensive logging: the possibility of “replay”.


Participatory ASM: Experimental Settings

  • Inexperienced subjects (CS students and office workers).

  • Not allowed to communicate.

  • “Open-ended” runs, stopped by the experimenter without prior notice.

  • 3-4 runs per person.

  • Questionnaire after the session.



Participatory ASM: Experimental Results 1/4

  • The presence of human traders increased market volatility.

  • The higher percentage of the population was human, the higher the difference was w.r.t. the performance of the fully computational population.

  • However, this may also be an effect of human learning.



Participatory ASM: Experimental Results 2/4

  • Despite the increased level of market deviations, price followed fundamental value.

  • This suggests that computational agents are able to adapt and ‘keep’ the market in balance.

  • However, their ability has its limitations…

  • The lesson of the initial runs:



Participatory ASM: Experimental Results 3/4

  • This initial mishap also demonstrates the effect of human learning.



Participatory ASM: Experimental Results 4/4

  • Human learning is also obvious in the relative performance of human participants and computational agents.

  • Notes:



Participatory ASM: Trading Strategies

  • Humans initially applied technical trading strategies, but gradually discovered fundamental strategies.

    • The winning human’s strategy was:
      • Buy if price < FV, sell otherwise.
  • The experiments confirmed that technical trading leads to market deviations.



Conclusions

  • We have introduced Agent-Based Modeling and Participatory Simulation.

  • We have argued for the use of PS to test ABMs and to help setting up laboratory experiments.

  • We have demonstrated the applicability of the concept.




Do'stlaringiz bilan baham:


Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2017
ma'muriyatiga murojaat qiling