臺灣股市籌碼資料的資訊內涵-以籌碼指標建構投資組合之實證研究

本研究以2010至2019年共十年之上市公司為樣本，探討臺灣股市籌碼資料的資訊內涵。我們將歸屬於「機構法人」與「散戶」的籌碼資料計算成15種不同的籌碼指標，同時分成大、中、小三種市值股票池進行探討，依據每個籌碼指標由大至小排序，分別建立5個投資組合，並以日頻率動態調整投資組合之持股。透過考量交易成本後的投資組合績效，本研究分析哪些籌碼資料具有有效的資訊內涵，同時檢視籌碼資料與投資組合報酬的關係，期望能作為投資者於實務操作時的參考依據。　　研究結果發現，首先，在屬於機構法人的籌碼資料中，投信買賣超張數指標最高的組別，於三種市值股票池中，不論是Sharpe ratio或是累積報酬率，皆有優於大盤的表現，意謂追隨投信買超較多的股票將能獲取較佳的報酬。而借券賣出張數指標最小的組別，皆具有Sharpe ratio優於大盤的表現，於大市值股票池中的累積報酬率更是優於大盤，代表借券賣出張數較少的股票具有較佳的股票報酬，為「消極看多」的操作邏輯。其次，在屬於散戶的籌碼資料中，融資使用率指標最低的組別，其Sharpe ratio於大、小市值股票池中具有優於大盤的表現，但累積報酬則未優於大盤。在大、小市值股票池中，融資使用率指標越高的組別，反而具有越差的績效表現，顯示融資使用率與股票報酬呈反向關係。

關鍵字

法人買賣超；融資；融券；投資策略

並列摘要

This research investigates the information contents of the trading data of listed companies in Taiwan stock market from year 2010 to 2019. We calculate two types of investing indicators from the trading data, the institutional and individual investors, respectively. We sort stocks into five portfolios according to each investing indicator, and rebalance the portfolios dynamically on a daily basis. Meanwhile, we also create three kinds of stock pools – large, medium, and small according to market size, and investigate whether there is any difference between market size. By measuring after-fee performance of these portfolios, we analyze which indicator contains informative contents, and check the relationship between trading data and portfolio returns, with the aim to create an investing guide for practical use. Empirical result indicates that according to the trading data of institutional investors, portfolios with the highest net bought by security investment trust companies (SITC) have better Sharpe ratio and cumulative return than market portfolio in three kinds of stock pools. It indicates that investors can get better returns if they buy the stocks which are most bought by SITC. Portfolios sold short the least by institutional investors have better Sharpe ratio than market portfolio in three kinds of stock pools, and the portfolio has better cumulative return in large-size stock pool. It implies that stocks sold short the least have better returns, indicating a “passive long signal”. In the trading data of individual investors, portfolios with the least margin purchase balance have better Sharpe ratio than market portfolio in large and small-size stock pools, but both of their cumulative returns can’t beat the market. We also find that these portfolios’ cumulative returns decrease while margin purchase balance increase, indicating a negative relationship between stock returns and margin purchase balance.