چه کسی برنده است؟ مطالعه بقا بلند مدت معامله گر در یک بازار سهام مصنوعی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|16183||2003||7 صفحه PDF||سفارش دهید||2388 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Physica A: Statistical Mechanics and its Applications, Volume 324, Issues 1–2, 1 June 2003, Pages 227–233
We introduce a multi-asset artificial financial market with finite amount of cash and number of stocks. The background trading is characterized by a random trading strategy constrained by the finiteness of resources and by market volatility. Stock price processes exhibit volatility clustering, fat-tailed distribution of returns and reversion to the mean. Three active trading strategies have been introduced and studied in two different market conditions: steady market and growing market with asset inflation. We show that the profitability of each strategy depends both on the periodicity of portfolio reallocation and on the market condition. The best performing strategy is the one that exploits the mean reversion characteristic of asset price processes.
Agent-based simulation of financial markets is a rapidly growing field . In previous works  and , we introduced the Genoa Artificial Stock Market (GASM), a computer simulator of financial markets which is able to reproduce main stylized facts present in real market, i.e., volatility clustering and fat-tailed distribution of returns with very simple assumptions about the behaviour of agents and a realistic market microstructure. The system is able to keep track of every agent portfolio and the agents are endowed with limited financial resources. In the first version of GASM, only one risky asset was traded in exchange for cash. In this paper, we present a multi-asset artificial market and we study the long-run performance of different trading strategies in a competitive market with finite wealth.
نتیجه گیری انگلیسی
Our study has shown that in a multi-asset artificial financial market, with finite amount of resources and a random trading strategy, asset price processes exhibit stable reversion to the mean. We have studied the long-run behaviour of three active trading strategy showing that their performance depends critically on the market condition (steady or growing) and on the periodicity of portfolio reallocation. Only a strategy fully exploiting the reversion to the mean of the price process gives satisfactory results in all cases.