DSpace community: 財務金融研究所

COT 持倉、投資人情緒與能源期貨報酬預測：聰明錢假說之檢驗;COT 持倉、投資人情緒與能源期貨報酬預測：聰明錢假說之檢驗

Tue, 20 Oct 2026 07:49:25 GMT

title: COT 持倉、投資人情緒與能源期貨報酬預測：聰明錢假說之檢驗;COT 持倉、投資人情緒與能源期貨報酬預測：聰明錢假說之檢驗 abstract: 本研究檢驗交易者持倉、投資人情緒、金融風險、能源庫存意外 (inventory surprise)、與貨幣供給成長率是否對於能源期貨的報酬率具有預測能力，並檢驗能源期貨市場的聰明錢假說。研究標的包括輕質原油、RBOB 汽油、天然氣及 ULSD 柴油與取暖油等商品的期貨價格，資料期間為 2011 年 1 月 1 日至 2026 年 4 月 30 日。由於 COT 資料為星期二持倉統計值，因此本文以週二至週三的日報酬率為預測目標，其餘變數則為星期二以前之最近期資料；因此本文分析 COT 持倉部位與次日期貨報酬率之間的領先資訊關係。本文的分析模型控制庫存意外、M2 成長率、經濟政策不確定性、股票與原油價格的隱含波動、金融壓力、散戶情緒及消費者信心。持倉資訊分別以 COT 淨部位比例（NetOI）、Wang 式部位情緒指標（WangSI）及部位變化（Flow）衡量，並比較五類交易者的差異。本文發現，M2 成長率在多數基準模型中對於短期報酬率呈顯著負向影響，OVIX 則呈顯著正向影響；其餘情緒與風險指標未呈現一致的預測效果。在星期二持倉統計值方面，持倉部位變化（Flow）指標相對於 NetOI 與 WangSI 提供較多短期資訊；其中，生產者&商業避險者的持倉部位變化呈顯著正向影響，非可報告交易者的持倉部位變化呈顯著負向影響。GARCH-in-Mean 結果顯示，部分 COT 訊號與報酬或條件變異數具有顯著關係；但兩項條件均未在同一商品與同一訊號中同時成立，顯示五類交易者均不具有聰明錢效應。;This thesis examines whether trader positions， investor sentiment， financial risk， energy inventory surprise， and money growth are related to Tuesday-to-Wednesday returns in four NYMEX energy futures: Light Crude Oil Continuous， RBOB gasoline futures， natural gas futures， and ULSD diesel/heating-oil futures. The nominal sample period runs from January 1， 2011 to April 30， 2026. Except for the COT variables， which are measured using positions dated Tuesday， predictors are aligned to information available by Tuesday. Because Tuesday COT positions are normally released on Friday， the main COT estimates represent lead-lag information rather than a publicly executable Tuesday strategy. The models control for inventory surprise， M2 growth， economic policy uncertainty， equity- and oil-market implied volatility， financial stress， retail sentiment， and consumer confidence. NetOI， WangSI， and position change (Flow) are compared across five trader groups using fixed-effects panel regressions， commodity-specific regressions， and GARCH-in-Mean models. The results show that M2 growth is significantly negative in most baseline specifications， whereas OVIX is significantly positive; the remaining sentiment and risk indicators do not exhibit consistent effects. In the main analysis based on Tuesday position dates， Flow contains relatively more short-horizon information than NetOI and WangSI. Producer/merchant hedger Flow is significantly positive， while nonreportable Flow is significantly negative. Commodity-specific estimates indicate that COT relations vary across commodities， trader groups， and position measures. The GARCH-in-Mean results identify significant return or conditional-variance relations for some COT signals， but the two smart-money conditions never occur for the same signal in the same commodity. Therefore， none of the five trader groups supports the smart-money hypothesis.

台灣資產成長效應之非線性結構：供應鏈地位的調節;台灣資產成長效應之非線性結構：供應鏈地位的調節

Tue, 20 Oct 2026 07:49:22 GMT

title: 台灣資產成長效應之非線性結構：供應鏈地位的調節;台灣資產成長效應之非線性結構：供應鏈地位的調節 abstract: 本研究採用 1987 年 7 月至 2025 年 6 月台灣上市櫃公司為樣本，探討資產成長效應（Asset Growth Effect）在台灣股市長期不顯著之謎。相較於美國市場顯著的負向線性關係（Cooper et al.， 2008），本研究主張此一台灣悖論根源於傳統模型忽略了台灣企業在全球供應鏈中的地位異質性。據此，本研究創新引入財報中的合約負債比率作為關係專屬性投資（RSI）程度與供應鏈訂單保障能力的核心代理變數，並結合 Hou， Xue， and Zhang (2015) 的 q 理論框架，推導出資產擴張由受罰轉為溢酬的定價反轉機制。實證結果顯示，全樣本 Fama-MacBeth 迴歸中資產成長率（AG）無顯著線性預測能力。然而，依合約負債比率進行 50%／40%／10% 的非對稱子樣本分組後，隱藏的定價結構完整浮現：缺乏訂單保障的弱勢供應商（Low RSI）其 AG 係數顯著為負，呈現理性擴張受罰；而具備高額預付款保障的優勢供應商（High RSI）其 AG 係數則明顯反轉為正，形成實質的擴張溢酬。此一正負兩股定價力量的平均化抵銷，正是傳統線性模型失效的底層原因。進一步分析顯示，擴張溢酬現象高度集中於電子科技業（High RSI 群體市值佔比達 75.1%），顯示該機制根植於台灣電子代工供應鏈的特殊結構。切分點穩健性測試亦證實，正向擴張溢酬僅在極端前 10%～5% 的門檻下顯著，與合約負債極端右偏的分佈特徵高度吻合。本研究為 Goo and Wang (2024) 所觀察到的非線性 U 型特徵提供了完整的微觀解釋，並確立合約負債比率為解構台灣股市資產成長效應非線性動態的關鍵調節變數。;This study investigates the "Taiwan Paradox"—the long-standing statistical insignificance of the asset growth (AG) effect in the Taiwanese stock market—using a sample of listed and OTC companies from July 1987 to June 2025. In contrast to the significant negative linear relationship documented in the U.S. market (Cooper et al.， 2008)， we contend that this paradox stems from conventional models neglecting supply chain heterogeneity. Accordingly， this study introduces the contract liability ratio as a novel proxy variable for the degree of relationship-specific investment (RSI) and supply chain order security. Integrating this proxy into the investment-based asset pricing framework (q-theory) of Hou， Xue， and Zhang (2015)， we derive a pricing reversal mechanism where asset expansion shifts from a valuation penalty to a premium. The empirical results show that while full-sample Fama-MacBeth regressions yield no significant linear predictive power for AG， an asymmetric 50%/40%/10% sorting based on the contract liability ratio reveals a hidden pricing bifurcation. Weak suppliers lacking order security (Low RSI) exhibit a significantly negative AG coefficient (-0.0095， t = -2.04)， indicating a rational expansion penalty. Conversely， dominant suppliers backed by high prepayment guarantees (High RSI) display an AG coefficient that strongly reverses to positive (0.0138， t = 2.65)， forming a substantial expansion premium. The mutual cancellation of these opposing cross-sectional forces causes standard pooled linear models to fail. Further analysis demonstrates that this expansion premium is highly concentrated in the electronic technology sector， where the High RSI group commands a 75.1% market value share， indicating that this mechanism is rooted in the unique structure of Taiwan′s electronics supply chain. Breakpoint robustness tests verify that the positive premium is confined to the extreme top 10% to 5% thresholds， matching the highly right-skewed distribution characteristic of contract liabilities. In conclusion， this study resolves the puzzle of Taiwan′s invalid AG effect， providing a comprehensive supply-chain and q-theoretic foundation for the distinctive non-linear U-shaped pattern observed by Goo and Wang (2024).

經理人文化背景認同如何影響公司風險之承擔-以 S&P1500 指數上市公司為例;經理人文化背景認同如何影響公司風險之承擔-以 S&P1500 指數上市公司為例

Tue, 20 Oct 2026 07:49:12 GMT

title: 經理人文化背景認同如何影響公司風險之承擔-以 S&P1500 指數上市公司為例;經理人文化背景認同如何影響公司風險之承擔-以 S&P1500 指數上市公司為例 abstract: 本研究旨在探討經理人文化背景認同對公司風險承擔行為之影響。不同於傳統僅以國籍作為文化代理變數之做法，本文將經理人的生涯分為三個關鍵階段— —出生地、大學教育及工作經歷，透過Hofstede（ 2010）與Schwartz（ 2008）所提出之文化價值指標，量化各階段的文化屬性，並進一步計算該文化與公司所在地文化之差距，作為衡量經理人文化認同差異的基礎。研究樣本涵蓋1992 年至2022 年間S&P 1500 指數所屬公司之高階經理人，並以非系統風險作為公司風險承擔的衡量指標，藉由多元回歸模型，控制公司與經理人層級之變數、產業與年度固定效果後，探討文化價值差異對公司風險承擔行為的實證影響。實證結果顯示，出生階段與大學階段所對應之文化背景呈現較多顯著文化構面，而工作階段之顯著結果相對有限。部分達顯著水準之文化價值，其影響方向與主要理論預期不一致，顯示文化價值可能透過權力集中、不確定性控制、聲譽與職涯風險等不同機制影響經理人決策。;This study examines the effect of managers’ cultural background and cultural identification on corporate risk-taking. Unlike conventional approaches that rely solely on nationality as a proxy for culture， this study divides managers’ life trajectories into three key stages—place of birth， university education， and work experience. The cultural attributes associated with each stage are quantified using the cultural value indices developed by Hofstede (2010) and Schwartz (2008). The cultural distance between each manager’s background and the culture of the company’s home country is then calculated as the basis for measuring differences in managers’ cultural backgrounds. The sample consists of senior executives from firms included in the S&P 1500 Index between 1992 and 2022. Corporate idiosyncratic risk is employed as the proxy for corporate risk-taking. Using multivariate regression models that control for firm-level and manager-level characteristics， as well as industry and year fixed effects， this study investigates the empirical relationship between cultural value differences and corporate risk-taking. The empirical results show that the cultural backgrounds corresponding to the birth and university stages exhibit significant associations across a greater number of cultural dimensions， whereas significant results are relatively limited for the work stage. For some cultural values that reach statistical significance， the estimated directions are inconsistent with the primary theoretical predictions. These findings suggest that cultural values may influence managerial decision-making through various mechanisms， including the concentration of decision-making authority， the desire to control uncertainty， reputational concerns， and career risk.

貨幣政策衝擊對比特幣現貨與期貨市場之價格發現的影響;貨幣政策衝擊對比特幣現貨與期貨市場之價格發現的影響

Tue, 20 Oct 2026 07:48:58 GMT

title: 貨幣政策衝擊對比特幣現貨與期貨市場之價格發現的影響;貨幣政策衝擊對比特幣現貨與期貨市場之價格發現的影響 abstract: 本研究探討美國聯邦公開市場委員會（Federal Open Market Committee， FOMC）貨幣政策衝擊是否影響比特幣現貨與期貨市場之價格發現。隨著比特幣期貨與相關金融商品發展，比特幣逐漸由去中心化支付工具轉變為與傳統金融市場連結更深的風險性資產，因此貨幣政策資訊是否改變其市場間資訊傳遞機制，成為值得研究的重要議題。本文以比特幣現貨價格與比特幣期貨價格之五分鐘高頻資料為研究對象，並結合 FOMC 宣告日、貨幣政策衝擊的非預期成分及正負向貨幣政策衝擊的非預期成分，運用誤差修正模型檢驗現貨與期貨市場偏離長期均衡後之調整關係，再以共同因子成分方法（Component Share， CS）衡量兩市場對共同有效價格之相對貢獻。實證結果顯示，比特幣現貨與期貨價格之間存在長期均衡關係，且兩市場皆具有短期誤差修正效果。進一步而言，單純的 FOMC 宣告日效果並非影響價格發現的主要因素，真正具有影響力的是公告內容中超出市場預期的貨幣政策衝擊。當貨幣政策衝擊納入模型後，貨幣政策衝擊不僅會影響比特幣現貨與期貨報酬，也會改變兩市場對價格偏離的調整速度。共同因子成分方法結果顯示，在一般交易期間，比特幣現貨市場具有較高價格發現貢獻；然而，在 FOMC 貨幣政策衝擊進入市場時，比特幣期貨市場之價格發現貢獻明顯提高。此結果表示，比特幣雖具有全天候交易與去中心化特性，但其金融化與機構化發展已使其價格形成過程更容易受到傳統貨幣政策與總體金融資訊影響。;This study investigates whether U.S. monetary policy shocks affect price discovery between Bitcoin spot and futures markets. With the development of Bitcoin futures and related financial products， Bitcoin has gradually transformed from a decentralized payment instrument into a risky asset more closely connected to traditional financial markets. Therefore， whether monetary policy information changes the information transmission mechanism between Bitcoin spot and futures markets has become an important research issue. This study uses five-minute high-frequency data on Bitcoin spot prices and CME (Chicago Mercantile Exchange) Bitcoin futures prices， together with FOMC (Federal Open Market Committee) announcement dates， monetary policy surprise measures， and positive and negative monetary policy shocks. An error correction model is employed to examine the adjustment process between the spot and futures markets when prices deviate from their long-run equilibrium. In addition， the component share measure is used to evaluate the relative contribution of each market to the common efficient price. The empirical results show that a long-run equilibrium relationship exists between Bitcoin spot and futures prices， and both markets exhibit short-run error correction effects. Furthermore， the results suggest that the FOMC announcement day effect alone is not the main factor influencing price discovery. Instead， the unexpected component of monetary policy announcements plays a more important role. After incorporating monetary policy surprises into the model， monetary policy shocks are found to affect not only Bitcoin spot and futures returns， but also the speed at which the two markets adjust to deviations from equilibrium. The Component Share results indicate that， during ordinary trading periods， the Bitcoin spot market contributes more to price discovery. However， when FOMC monetary policy surprises enter the market， the contribution of the CME Bitcoin futures market to price discovery increases significantly. These findings suggest that although Bitcoin is characterized by around-the-clock trading and decentralization， its increasing financialization and institutionalization have made its price formation process more sensitive to traditional monetary policy and macro-financial information.