金融理论与实践
金融理論與實踐
금융이론여실천
Financial Theory and Practice
2015年
10期
81-88
,共8页
股票市场%BDSS模型%极值理论%流动性调整%变现时间%风险测度
股票市場%BDSS模型%極值理論%流動性調整%變現時間%風險測度
고표시장%BDSS모형%겁치이론%류동성조정%변현시간%풍험측도
stock market%BDSS model%extreme value theory%liquidity adjusted%liquidation time%risk measurement
针对现有市场极值风险测度方法不能反映市场流动性风险的缺陷,采用价量结合的风险测度方法,引入由换手率(Turnover Rate,以下简称TR)求得的变现时间和极值理论(Extreme Value Theory,以下简称EVT)对Bangia等(1998)提出的经流动性调整VaR的La-VaR模型(BDSS模型)进行改进,进而运用改进后的La-VaR模型对上证综指(Shanghai Stock Exchange Composite Index,以下简称SSEC)经流动性调整的极值风险进行测度,并采用规范的返回测试检验方法(Back-testing)对模型的稳健性进行检验。实证结果表明:改进后的La-VaR模型比传统的BDSS模型更能准确测度中国股市经流动性调整的极值风险;在改进后的La-VaR模型中,基于极值理论的La-VaR模型比基于学生t分布的La-VaR模型不仅更能准确地测度风险,而且模型的溢出情况也更为随机,从而拥有更好的稳健性。
針對現有市場極值風險測度方法不能反映市場流動性風險的缺陷,採用價量結閤的風險測度方法,引入由換手率(Turnover Rate,以下簡稱TR)求得的變現時間和極值理論(Extreme Value Theory,以下簡稱EVT)對Bangia等(1998)提齣的經流動性調整VaR的La-VaR模型(BDSS模型)進行改進,進而運用改進後的La-VaR模型對上證綜指(Shanghai Stock Exchange Composite Index,以下簡稱SSEC)經流動性調整的極值風險進行測度,併採用規範的返迴測試檢驗方法(Back-testing)對模型的穩健性進行檢驗。實證結果錶明:改進後的La-VaR模型比傳統的BDSS模型更能準確測度中國股市經流動性調整的極值風險;在改進後的La-VaR模型中,基于極值理論的La-VaR模型比基于學生t分佈的La-VaR模型不僅更能準確地測度風險,而且模型的溢齣情況也更為隨機,從而擁有更好的穩健性。
침대현유시장겁치풍험측도방법불능반영시장류동성풍험적결함,채용개량결합적풍험측도방법,인입유환수솔(Turnover Rate,이하간칭TR)구득적변현시간화겁치이론(Extreme Value Theory,이하간칭EVT)대Bangia등(1998)제출적경류동성조정VaR적La-VaR모형(BDSS모형)진행개진,진이운용개진후적La-VaR모형대상증종지(Shanghai Stock Exchange Composite Index,이하간칭SSEC)경류동성조정적겁치풍험진행측도,병채용규범적반회측시검험방법(Back-testing)대모형적은건성진행검험。실증결과표명:개진후적La-VaR모형비전통적BDSS모형경능준학측도중국고시경류동성조정적겁치풍험;재개진후적La-VaR모형중,기우겁치이론적La-VaR모형비기우학생t분포적La-VaR모형불부경능준학지측도풍험,이차모형적일출정황야경위수궤,종이옹유경호적은건성。
Aiming at the issue that the existing market extreme risk measure methods can’t reflect the market liquidity risk, in this paper, we use liquidation time and extreme value theory (EVT) to modify the BDSS model, a kind of VaR model with liquidity adjusted which was put forward by Bangia et al in 1998.And then we use the improved La-VaR model to measure the extreme risk of shanghai stock exchange composite index (SSEC) with liquidity adjusted. Besides, we also use the normative back-testing method for the robustness testing of the La-VaR models. The empirical results show that the improved La-VaR models are better than the traditional BDSS model in the measurement of extreme risk in Chinese stock market with liquidity adjusted;and among the improved La-VaR models, the La-VaR model based on EVT not only has a better ability to measure risk accurately, but also can get a much more random situation for the failure prediction than the La-VaR model based on student t, so the La-VaR model based on EVT has a better performance.