本文選題：Kelly策略　切入點：投資組合構(gòu)建　出處：《華南理工大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

【摘要】：高頻交易近兩年在國內(nèi)興起，勢頭強勁，然而經(jīng)歷兩年后的盈利狀況卻不容樂觀，隨著參與者的蜂擁而入以及國內(nèi)證券市場微觀結(jié)構(gòu)的變化，高頻交易難度增大并且收益率下降。本文針對這一問題，提出通過Kelly策略優(yōu)化單一高頻策略模型的收益，并且建立基于高頻交易的投資組合策略，這也是本文的創(chuàng)新點。 現(xiàn)代的投資組合理論大多都是基于Markowitz在1952年提出的均值-方差模型，本文則是從另外一個視角出發(fā)，基于美國貝爾實驗室工程師Kelly提出的“資金增長最快的投資比例”理論進(jìn)行投資組合構(gòu)建策略研究。 本文首先介紹投資組合構(gòu)建的主要研究方法以及發(fā)展創(chuàng)新，提出了基于最優(yōu)增長路徑的Kelly投資策略的必要性，并且對有關(guān)Kelly策略的國內(nèi)外研究現(xiàn)狀進(jìn)行了綜述，初步論述Kelly策略。 接下來，具體介紹Kelly策略。首先從投資單一資產(chǎn)出發(fā)，利用收益率服從二項分布的賭局介紹Kelly策略的基本原理，并驗證其長期最優(yōu)性，分析其理論依據(jù)與風(fēng)險。接著，研究了資產(chǎn)組合的Kelly策略，給出計算Kelly策略最優(yōu)投資比例的一般化方程表達(dá)式。進(jìn)一步給出資產(chǎn)收益率獨立且服從于同一二項分布的資產(chǎn)組合下的Kelly投資組合策略，求解并分析其性質(zhì)。接著引出多期投資組合問題，并建立基于Kelly策略的多期最優(yōu)投資組合構(gòu)建策略。 對于高頻交易的相關(guān)知識，本文詳細(xì)介紹高頻交易在我國近兩年的發(fā)展現(xiàn)狀、高頻交易的主要特性、高頻交易系統(tǒng)以及常用的高頻交易模型：均線模型、布林線模型和RSI模型。創(chuàng)新的提出高頻交易組合策略，并給出相互獨立的兩個高頻交易策略的最優(yōu)資金分配模型與求解方法。 實證分析環(huán)節(jié)，選取5分鐘的滬深300股指期貨數(shù)據(jù)，通過不同的高頻交易策略模型，首先驗證Kelly策略在單一高頻交易模型應(yīng)用上的優(yōu)勢，再驗證了兩個高頻交易模型組合策略的可行性。 最后，對全文的工作進(jìn)行了總結(jié)，，并對未來研究的方向進(jìn)行了展望。
[Abstract]:High frequency trading has been rising in China for nearly two years. However, after two years, the profit situation is not optimistic. With the influx of participants and the changes of the microstructure of the domestic securities market, In order to solve this problem, this paper proposes to optimize the returns of a single high-frequency strategy model by using Kelly strategy, and to establish a portfolio strategy based on high-frequency trading, which is also the innovation of this paper. Most of the modern portfolio theories are based on the mean-variance model proposed by Markowitz in 1952. Based on the theory of "the fastest-growing investment ratio" proposed by Bell Labs engineer Kelly, the strategy of portfolio construction is studied. This paper first introduces the main research methods and development innovation of portfolio construction, puts forward the necessity of Kelly investment strategy based on optimal growth path, and summarizes the current research situation of Kelly strategy at home and abroad. The Kelly strategy is discussed preliminarily. Then, the Kelly strategy is introduced in detail. Firstly, the basic principle of Kelly strategy is introduced by using the bet with the binomial distribution of yield service, and its long-term optimality is verified, and its theoretical basis and risk are analyzed. In this paper, the Kelly strategy of portfolio is studied, and the expression of general equation for calculating the optimal investment proportion of Kelly strategy is given. Furthermore, the Kelly portfolio strategy with independent return on assets and under the same binomial distribution is given. Then the multi-period portfolio problem is introduced, and the multi-period optimal portfolio construction strategy based on Kelly strategy is established. For the related knowledge of high-frequency trading, this paper introduces the current situation of high-frequency trading in China in recent two years, the main characteristics of high-frequency trading, the high-frequency trading system and the commonly used high-frequency trading model: the average line model. Brinling line model and RSI model. The combination strategy of high frequency trading is proposed, and the optimal fund allocation model and solution method of two independent high frequency trading strategies are given. In the empirical analysis, we select the Shanghai and Shenzhen 300 stock index futures data for 5 minutes, and verify the advantages of Kelly strategy in the application of the single high-frequency trading model through different high-frequency trading strategy models. The feasibility of the combination strategy of two high frequency trading models is verified. Finally, the work of the paper is summarized, and the future research direction is prospected.
【學(xué)位授予單位】：華南理工大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：F724.5;F224

【參考文獻(xiàn)】

相關(guān)期刊論文 前1條

1 陸士杰;楊朝軍;;基于股價服從對數(shù)正態(tài)分布的凱利投資策略[J];經(jīng)濟(jì)數(shù)學(xué);2013年03期

本文編號：1579678