| dc.contributor | 國貿系 | |
| dc.creator (作者) | 郭維裕 | |
| dc.date (日期) | 2020-03 | |
| dc.date.accessioned | 7-Apr-2026 13:29:06 (UTC+8) | - |
| dc.date.available | 7-Apr-2026 13:29:06 (UTC+8) | - |
| dc.date.issued (上傳時間) | 7-Apr-2026 13:29:06 (UTC+8) | - |
| dc.identifier.uri (URI) | https://ah.lib.nccu.edu.tw/item?item_id=181966 | - |
| dc.description.abstract (摘要) | 本研究計畫為三年期的計畫。研究的議題為探討經驗相似性（empirical similarity）在財務領域上的應用性。經驗相似性是由Gilboa, Lieberman, and Schmeidler (2006)所提出的一個嶄新的經濟決策原則。有別於財務經濟理論中的效用極大化決策原則，經濟主體（economic agent）或投資人進行其風險性決策時不再被假設為使用貝氏統計方法（Bayesian statistics）處理所參考的過去所有訊息。經驗相似性決策原則認為當經濟主體或投資人面對不確定性的決策時，會透過比對其過去經驗中與目前決策情境相類似的個案的方式，蒐集並過濾相關訊息，以做為其最終決策的參考，因此經驗相似性亦被稱為案例式決策原則（case-based decision rule）。
本研究計畫規劃由三個可能方向探討經驗相似性在財務研究議題上的應用。第一個方向為技術分析（technical analysis）。雖然技術分析常被投資人使用以協助進行投資決策，財務學界卻對技術分析的有效性抱持著質疑的態度。綜觀目前常用的技術分析指標，如移動平均（moving averages）、RSI指標（relative strength index）、交易區間（trading range，支撐線和壓力線）以及各式各樣的線型等，其決策原則皆相當直覺單純。以移動平均為例，投資人單純地依據短天期的移動平均是否往上或向下穿透長天期的移動平均，而進行相對應的買進或賣出（甚或放空）動作。如果投資人可以透過使用經驗相似性而參考過去的穿透情形進行案例式的比對的話，或許可以增進其預測的準確度。第二個方向則為風險溢酬（risk premium）的預測。Pesaran and Timmermann (1995, 2000)利用遞迴式以及滾動式迴歸模型進行美國與英國股市風險溢酬的預測。本研究嘗試應用經驗相似性的決策原則以增進風險溢酬的預測效能。第三個方向則為應用於探討基金經理人的團體交易（crowded trade）的特性上。這是一個相當新穎的重要研究課題。目前的文獻不多。Pojarliev and Levich (2011)研究外匯基金經理人的團體交易。本研究打算結合其實證模型與經驗相似性進行相關研究，希望能更加精準的測度基金經理人的團體交易傾向。以上的經驗相似性在財務領域相關重要研究課題上的應用乃為本研究計畫的主要貢獻。 | |
| dc.description.abstract (摘要) | This is a three-year research project. Its topic is the financial applications of empirical similarity. Empirical similarity is a brand new and interesting decision rule recently proposed by Gilboa, Lieberman, and Schmeidler (2006). It is quite different from the traditional utility maximizing decision rule. According to the utility maximizing decision rule, economic agents or investors are assumed to be Bayesian decision makers, who are able to fully and efficiently utilize past relevant information to make decisions under uncertainty in order to maximize their utility functions. Instead of imposing such stringent requirement on the ability of economic agents to make use of historical information, the empirical similarity decision rule posits that when making decisions under uncertainty, economic agents would tend to employ the empirical similarity decision rule that is case-based. Take housing market as an example, when a homeowner considers selling her house and wonders how much she is able to sell, she would probably base her assessment on the prices at which other houses in the vicinity of her house were selling. In other words, rather than based on past information, economic agents would choose to extract relevant information from previous similar scenarios or cases when making uncertain decisions.
This research project will explore from three different directions the potential applications of empirical similarity decision rule to important research topics in finance. Firstly, we would like to apply the empirical similarity to studying the profitability of technical analysis. There are many commonly used technical indicators in financial markets, such as moving averages, relative strength indicator, trading range, and various price patterns. These technical indicators provide investors with simple and intuitive trading rules for them to make investment decisions. We expect that the empirical similarity rule may help investors to make better use of these technical indicators to increase their profits. Secondly, we will apply the empirical similarity rule to forecasting future risk premiums in stock markets. Pesaran and Timmermann (1995, 2000) utilize the recursive and rolling regression models to predict the risk premiums for U.S. and U.K. stock markets. We plan to combine the empirical similarity rule with these regression models to enhance the prediction efficacy of risk premiums. Thirdly, we will investigate the crowded trading behavior of mutual fund managers based on the empirical similarity concept. The crowded trade is a very new and important topic in finance. Pojarliev and Levich (2011) examine the crowded trade phenomenon in currency funds. We will attempt to combine their model with the empirical similarity to study the crowded trades in equity funds. By implementing this research project, we hope to make significant contributions to finance area both academically and practically. | |
| dc.format.extent | 116 bytes | - |
| dc.format.mimetype | text/html | - |
| dc.relation (關聯) | 科技部, MOST105-2410-H004-033-MY3, 105.08-108.07 | |
| dc.subject (關鍵詞) | 經驗相似性; 技術分析; 風險溢酬; 團體交易 | |
| dc.subject (關鍵詞) | empirical similarity; technical analysis; risk premium; crowded trade | |
| dc.title (題名) | 經驗相似性在財務上的應用 | |
| dc.title (題名) | Financial Applications of Empirical Similarity | |
| dc.type (資料類型) | report | |
