本文選題：期貨交易維持保證金 + SV-M模型��； 參考：《湖南大學》2014年碩士論文

【摘要】：伴隨著金融全球化進程,期貨市場間的暴漲暴跌聯(lián)動效應(yīng)加大。作為期貨市場防范風險的第一道門檻,交易維持保證金對期貨市場乃至整個資本市場而言都至關(guān)重要。期貨交易維持保證金貫穿于整個期貨交易過程,且其效果關(guān)系到期貨市場的穩(wěn)定性和效率。本文分別基于單個期貨合約和期貨組合兩個視角,沿著期貨交易維持保證金設(shè)定模型的構(gòu)建、期貨交易維持保證金的設(shè)定以及期貨交易維持保證金設(shè)定效果的評價這一路徑展開。 就單個期貨合約的交易維持保證金設(shè)定而言,本文考慮到常用的收益率序列遺漏盤中價格變動訊息,以滬深300股指期貨5分鐘高頻數(shù)據(jù)為基礎(chǔ),結(jié)合期貨操作實務(wù)特點,構(gòu)造多頭和空頭的日內(nèi)最大損失率序列,提出一個新的波動率測度即日內(nèi)最大損失率方差,并將隨機波動均值內(nèi)(SV-M)模型、極值理論的超閾值(POT)方法和冪譜風險測度(PSRM)方法相結(jié)合,構(gòu)建一個新的維持保證金水平設(shè)定模型(即SV-M-POT-PSRM模型),據(jù)此分別進行多頭和空頭維持保證金水平的設(shè)置。實證研究證實,相較于GARCH-M-VaR, GARCH-M-POT-VaR, SV-M-VaR和SV-M-POT-VaR模型,無論是樣本內(nèi)還是樣本外,在謹慎性指數(shù)(PI)以及平均過高要價(OCI)方面,SV-M-POT-PSRM模型的優(yōu)越性明顯,說明基于該模型設(shè)定的交易維持保證金既能有效防控風險,又能減少被占用資金,從而使資金使用效率得以顯著提高。 考慮到在期貨操作實務(wù)中投資者傾向于采取組合投資策略,往往持有期貨合約組合以分散和對沖風險,因此僅僅孤立地對單個期貨合約的交易維持保證金進行研究顯然是片面的,而需要基于期貨組合的層面考慮交易維持保證金的設(shè)定。鑒于此,本文將多元t-Copula模型、POT方法以及PSRM方法相結(jié)合,構(gòu)建一個新的期貨組合交易維持保證金設(shè)定模型(即多元t-Copula-POT-PSRM模型)。從實證結(jié)果來看,在PI值和OCI值方面,多元t-Copula-POT-PSRM模型均優(yōu)于現(xiàn)行的線性相加模型,尤其是在OCI方面,該模型使樣本內(nèi)和樣本外OCI值大幅降低,表明其既能有效地防范風險,又能降低機會成本。
[Abstract]:With the development of financial globalization, the linkage effect between futures markets is increasing. As the first threshold for the futures market to guard against risk, it is very important for the futures market and even the whole capital market to maintain the margin. Futures trading maintains margin throughout the futures trading process, and its effect relates to the stability and efficiency of futures market. Based on the perspective of single futures contract and futures combination, this paper constructs the model of margin setting for futures trading. The path of setting up the margin for futures trading and evaluating the effect of the margin setting for futures trading is carried out. With regard to the setting of the trading margin for a single futures contract, this paper takes into account the fact that the usual yield series omits intraday price change information, based on the 5-minute high frequency data of Shanghai and Shenzhen 300 stock index futures, and combining with the practical characteristics of futures operation. In this paper, we construct a series of intraday maximum loss rates for long and short positions, and propose a new volatility measure for the variance of the maximum loss rate in the day. Combined with the power spectrum risk measurement (PSRM) method, a new model of maintaining margin level (SV-M-POT-PSRM model) is constructed. According to this model, the long and short margin levels are set up respectively. The empirical study proves that compared with GARCH-M-VaR, GARCH-M-POT-VaR, SV-M-VaR and SV-M-POT-VaR models, the SV-M-POT-PSRM model is superior to the GARCH-M-POT-VaR, SV-M-VaR and SV-M-POT-VaR models in terms of the cautious index (Pi) and the average asking price (OCII). It is shown that the margin set up based on the model can prevent and control risks effectively and reduce the occupied funds, thus improving the efficiency of the use of funds significantly. Considering that in the practice of futures operations, investors tend to adopt a portfolio investment strategy, often holding futures contract portfolios to diversify and hedge risks, Therefore, it is obvious that it is one-sided to study the trading maintenance margin of a single futures contract in isolation, and it is necessary to consider the setting of the trading maintenance margin based on the level of futures portfolio. In view of this, this paper combines the multivariate t-Copula model pot method with the PSRM method to construct a new futures portfolio trading margin setting model (i.e. multivariate t-Copula-POT-PSRM model). From the empirical results, the multivariate t-Copula-POT-PSRM model is superior to the existing linear additive model in Pi value and OCI value, especially in the aspect of OCI. The model can significantly reduce the OCI value within and outside the sample, which shows that it can effectively prevent the risk. It can also reduce the opportunity cost.
【學位授予單位】：湖南大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：F724.5;F224

【參考文獻】

相關(guān)期刊論文 前10條

1 侯雋;;有偏型EWMA的股指期貨保證金設(shè)定的實證研究[J];財經(jīng)科學;2009年04期

2 徐國祥,吳澤智;我國指數(shù)期貨保證金水平設(shè)定方法及其實證研究——極值理論的應(yīng)用[J];財經(jīng)研究;2004年11期

3 徐正國,張世英;調(diào)整"已實現(xiàn)"波動率與GARCH及SV模型對波動的預(yù)測能力的比較研究[J];系統(tǒng)工程;2004年08期

4 花擁軍;張宗益;;滬深股市極端風險的實證研究與比較分析——基于GPD分布的極值POT模型[J];系統(tǒng)工程;2009年02期

5 林宇;譚斌;魏宇;;基于多元GARCH與極值理論的資產(chǎn)組合風險測度研究[J];管理學報;2010年04期

6 韓德宗;王興鋒;楊敏敏;樓迎軍;;基于極值譜風險測度的動態(tài)保證金水平設(shè)定[J];管理科學;2009年01期

7 蘇衛(wèi)東,張世英;多元長記憶SV模型及其在滬深股市的應(yīng)用[J];管理科學學報;2004年01期

8 魏宇;;股票市場的極值風險測度及后驗分析研究[J];管理科學學報;2008年01期

9 魏宇;;滬深300股指期貨的波動率預(yù)測模型研究[J];管理科學學報;2010年02期

10 黃波;顧孟迪;李湛;;偏正態(tài)隨機波動模型及其實證檢驗[J];管理科學學報;2010年02期

，

本文編號：1792327