系统工程理论与实践
繫統工程理論與實踐
계통공정이론여실천
Systems Engineering—Theory & Practice
2013年
10期
2645~2653
,共null页
指数追踪 投资组合 启发式遗传算法 沪深300
指數追蹤 投資組閤 啟髮式遺傳算法 滬深300
지수추종 투자조합 계발식유전산법 호심300
index tracking; investment portfolio; heuristic search; CIS300
消极组合管理方法已由国内外众多基金的表现证明是一种有效的资产组合投资方式.指数基金作为采取消极管理策略的典型代表,其业绩超越多数采取积极管理模式的基金.指数基金管理者的主要目标是使其基金的收益尽可能接近其标的股指,如我国的沪深300,美国的标普500的收益.本文提出了一种基于启发式遗传算法的寻优方案,通过最大化效用函数来寻找一个最为经济的指数复制组合.该组合同时应该满足拥有最少的资产数量、尽可能少的权重调整次数、最小的收益波动性等限制条件以减少基金开销,并使其收益尽量接近或者超越标的指数的收益.为使该策略具有更强的实用性,文章考虑了股票具有最小交易规模、投资权重分布不平均等实际限制.实验所得策略通过构造追踪组合来匹配沪深300指数,其综合效果超过了使用二次规划、等权或者是先验经验构筑的投资组合.
消極組閤管理方法已由國內外衆多基金的錶現證明是一種有效的資產組閤投資方式.指數基金作為採取消極管理策略的典型代錶,其業績超越多數採取積極管理模式的基金.指數基金管理者的主要目標是使其基金的收益儘可能接近其標的股指,如我國的滬深300,美國的標普500的收益.本文提齣瞭一種基于啟髮式遺傳算法的尋優方案,通過最大化效用函數來尋找一箇最為經濟的指數複製組閤.該組閤同時應該滿足擁有最少的資產數量、儘可能少的權重調整次數、最小的收益波動性等限製條件以減少基金開銷,併使其收益儘量接近或者超越標的指數的收益.為使該策略具有更彊的實用性,文章攷慮瞭股票具有最小交易規模、投資權重分佈不平均等實際限製.實驗所得策略通過構造追蹤組閤來匹配滬深300指數,其綜閤效果超過瞭使用二次規劃、等權或者是先驗經驗構築的投資組閤.
소겁조합관리방법이유국내외음다기금적표현증명시일충유효적자산조합투자방식.지수기금작위채취소겁관리책략적전형대표,기업적초월다수채취적겁관리모식적기금.지수기금관리자적주요목표시사기기금적수익진가능접근기표적고지,여아국적호심300,미국적표보500적수익.본문제출료일충기우계발식유전산법적심우방안,통과최대화효용함수래심조일개최위경제적지수복제조합.해조합동시응해만족옹유최소적자산수량、진가능소적권중조정차수、최소적수익파동성등한제조건이감소기금개소,병사기수익진량접근혹자초월표적지수적수익.위사해책략구유경강적실용성,문장고필료고표구유최소교역규모、투자권중분포불평균등실제한제.실험소득책략통과구조추종조합래필배호심300지수,기종합효과초과료사용이차규화、등권혹자시선험경험구축적투자조합.
Passive portfolio management has been proved as an efficient method to get the average mar- ket return and can outperform most of funds which are managed actively. Index fund is one of typical passively managed funds. Its major objective is to mimic the performance of a benchmark stock index, i.e. S&P500 in US and CIS300 in China. In order to reduce the cost of transaction by limiting the number of rebalancing and unnecessary spending on less influential stock, this paper aims at proposing a heuristic genetic algorithm that could help to build a tracking portfolio with the help of a tailored utility flmction. The portfolio contains fewer stocks than the tracked index, while has adequate accuracy and lower costs. The technique is applied on building a stock portfolio which tracked the performance of CIS300, with a positive comprehensive output which outperforms tracking portfolio built by quadratic programming, equally weighted or experience guided models.