本研究旨在探討管理期貨基金 (Commodity trading advisor or managed futures funds , CTA )風險變動。本研究利用巴克萊銀行避險基金資料庫篩選出CTA,共計1241支CTA,共計27,672個觀察值。研究樣本期間為2000年1月至2012年5月,頻率為月資料。本研究以過去3、6、9個月移動標準差作為CTA風險,另外,大部分基金均以S&P 500指數作為追蹤基準,並衡量其3、6、9個月追蹤誤差風險,探討CTA風險變動現象是否來自蓄意變動 (intentional) 抑或是均值復歸 (mean reverting)現象。我們的方法使用實際的基金交易,並且主動改變風險。CTA追蹤誤差和移動風險現象均來自於均值復歸。 此外,本研究探討過去績效與風險變動關聯性,結果顯示,短期過去績效與風險變動關係為正,而中、長期過去績效與風險變動關係為負。另外對個別CTA而言,大部份CTA風險變動和追蹤誤差風險現象,具有均值復歸,少數具有人為變動。本研究結果,可以提供相關CTA文獻上風險變動實證意涵。
This study aimed to explore risk change of the commodity trading advisor (or managed futures funds, CTA). In this study, monthly CTA is obtained from Barclays hedge fund database, with a total of 1,241 samples and 27,672 observations. In this study we use both total risk and tracking error risk of CTA as risk measures with rolling window of 3, 6, and 9 months to examine whether changes in total risks (tracking error risks) result from being either intentional or mean reverting. The sample period is from January 2000 to May 2012. Our findings show that changes in both total risks and tracking error risks result from mean reversion. In addition, this study investigates the effects of the performance on changes in total risks and tracking error risk respectively. Our findings show that shot-term past performance are positively related to changes in total risks and tracking error risks while medium and long-term past performance negatively relates to changes in total risks and tracking error risks.In other words, it suggests of intention for short-term period but mean reverting for medium and long-term periods. Finally, we further discuss changes in total risks (tracking error risk) individually, and find mean reverting in most of them but intentional in several of them.