以日交易頻率執行多空動能交易是否能在美國商品期貨市場中獲利?

Abstract

本文使用日頻率數據探討 Narayan, Ahmed, and Narayan (2015)的「多空動能交易」策略在操作 22 種美國商品期貨之下的獲利表現,以 2002 年到 2020 年作為研究期間所得出的實證結果顯示:(1)策略唯有在使用「當日開收盤價報酬率」衡量交易訊號時,方可創造顯著的獲利,若是以「隔日收盤價報酬率」衡量交易訊號反而可能會蒙受損失;而且,(2)當進一步考慮原始保證金之機會成本時,因為以「當日開收盤價」操作屬當沖交易,故所需支付的保證金只有以「隔日收盤價報」操作的一半,因而可使策略在交易兩倍部位之下創造出更高的獲利;另外,(3)透過網路券商下單的交易成本比起委由實體券商下單更加低廉,且此一交易成本的節省亦可顯著提升策略的報酬表現;不過,(4)由於策略所參考的交易訊號集合是由多種商品期貨品項與多種長短期移動平均天數的組合所構成,導致設置年化帶寬的效益相當有限。

In this article, we employ daily data of 22 U.S. commodity futures to investigate the profitability of the “long-short momentum trading” strategy as introduced in Narayan, Ahmed, and Narayan (2015). Using data spanning from 2002 to 2020, our empirical result shows that: (1) the strategy earns significantly positive capital gains only when measuring momentum by “opened-to- closed” returns, whereas employing “closed-to-closed” returns to measure momentum may even result in capital loss; and (2) after further accounting for the opportunity cost associated with initial margins, the strategy based on “opened-to-closed” returns tends to make significantly more profits since it takes advantage from the saving of a half initial margins due to daily trading, which therefore enables it to trade more positions; in addition, (3) by contrast to placing orders via real brokers, cheaper trading cost by making transactions via online brokers also leads the strategy to be significantly more profitable; but however, (4) because the information set of the strategy is constructed by the number of examined commodity futures as well as the number of combinations that we use to calculate short-term and long-term moving averages, introducing bandwidth might be not as beneficial.