巨量雲端運算高頻交易平台

The Massive Cloud Computing High-Frequency Trading Platform

10.6245/JLIS.2016.422/652

陳奕光(Yi-Kuang Chen)；林承翰(Lin-Cheng Han)

雲端運算 ； 高頻交易 ； 服務導向架構 ； 技術分析 ； 交易策略評估 ； Cloud Computing ； High-Frequency Trading ； Technical Analysis

圖書館學與資訊科學

42卷2期（2016 / 10 / 01）

5 - 24

繁體中文

本研究應用雲端分散式的架構來建置一處理大量使用者交易需求的高頻交易投資策略服務平台，此平台有以下特色：（一）系統後端採用SOA 架構，將龐大的交易系統切割佈署到雲端叢集之上，並提供單一的Façade 介面供外部使用者呼叫。（二）不斷接收外部的即時報價訊息，並產生海量的即時市場狀態資訊，包含多種技術分析指標、買賣規則…等，以供高頻交易的策略作為買賣的依據。（三）利用Java Message Service 將大量的即時市場狀態資訊快速、非同步的派送給分佈在雲端叢集各節點的系統模組，並採取Publisher-Subscriber 的模式來維持分散後各系統模組之間的鬆散關係。（四）多樣化的統計演算法模型可供使用者作為產生個人化投資策略之依據。產生的新策略可馬上投入即時的模擬交易環境下監控與評估其策略績效。

The study employs a service platform which is based on Cloud Distributed Architecture to deal with a high volume of users trading needs of high‐frequency trading strategy investment services. This platform is designed to equip four characteristics: (a) The back‐end system uses the SOA architecture to deploy huge trading systems onto the cloud cluster, and provides a single Façade interface for external users call. (b) The external real‐time quotes and messages are received constantly, and a large amount of real‐time market information is generated, including a variety of technical analysis indicator, trading rules, etc. All forms the basis for high‐frequency trading strategies. (c) Through Java Message Service, a large number of real‐time market information is allowed to be quickly and asynchronously transmitted to each node in the distributed system module of the cloud cluster. Publisher‐Subscriber mode is adopted to maintain loose relationship between each module of the system after dispersion. And (d) A variety of statistical algorithm models can be used as basis for users to generate personalized investment strategies. This new strategy is expected to be immediately put into real‐time simulation trading environment to monitor and evaluate its strategic performance.

1. Shen, H.,Pétrot, F.(2011).Using amdahl's law for performance analysis of many-core soc architectures based on functionally asymmetric processors.International Conference on Architecture of Computing Systems
2. 林威辰(2011)。新竹市=Hsinchu City，國立交通大學應用數學系=Department of Applied Mathematics, National Chiao Tung University。
3. 徐宇薇、林俞君、潘長華、吳現任、戴天時、王釧茹(2014)。使用CUDA 平行計算機制運用到高頻套利交易─以台指期權市場為例。第三十一屆組合數學與計算理論研討會
4. 陳進忠(2005)。股票聖經：技術分析寶典。臺北縣=Taipei County:中經社文化有限公司=CENS - Taiwan sourcing service provider。
5. 黃嘉斌譯、Edwards, Robert D.、Magee, John(1998)。股價趨勢技術分析（上）。臺北市=Taipei City:寰宇出版股份有限公司=IPC International Publishing Co., Ltd. (IPCI)。