影響臺指期貨價格走勢之分析

Abstract

在電腦硬體技術快速發展及軟體演算法漸漸趨於成熟的現今，有別於以往使用手動下單的交易型態，愈來愈多的機構投資人及散戶，使用電腦軟體來進行自動化的「程式交易」，更甚者有同時具備資訊技術與財經知識的投資人，利用電腦來進行所謂的「高頻交易」。 因此，如何在極短的時間內，找出影響未來數秒之價格走勢影響因子，讓投資人可以即時的做出反應，賺取利潤並規避損失，是一重要議題。 本研究由文獻中找出可能影響金融商品價格短期內突然改變的構面，做出推論、建立假設、設計量化指標，並以2014年12月10日至2015年1月21日期間臺灣指數期貨共30個交易日的日內資料（含逐筆成交資料與委託簿報價資料）為實證標的，進行假設驗證，以找出可以顯著影響臺指期價格突然改變的因素。 實驗結果顯示，本研究透過供需原則假設，並搭合理的公式設計，以及適合的參數做搭配，可以在極短線內有效的預測臺指期貨走勢。

Today, in the rapid development of computer hardware technology and software algorithms gradually maturing. Unlike conventional trading patterns, More and more institutional investors and retail investors, using computer software to implement the "Program Trading". Furthermore, there are people who along with investment in information technology and financial knowledge, using the computers to carry out the "High Frequency Trading". So, how to find the impact of future price movements factors in a very short period of time are very important. So that investors can react instantly to earn profits and avoid losses, it is an important issue. This study was to find out from the literature that may affect a sudden change in commodity prices in the short term financial facets, make inferences, build assumptions, and design quantitative indicators. Using data between 10 December 2014 and 21 January 2015 period Taiwan Index Futures for the empirical subject, verify assumptions, to identify factors that can significantly affect Taiwan Index Futures sudden price changes. Experimental results show that according to supply and demand principle through assumptions and formulas designed to take reasonable and appropriate parameters to do with, can effectively predict TAIEX goods movements within a very short-term.