Meese–Rogoff 之謎—即匯率預測難以超越隨機漫步（Random walk）—仍然是國際金融領域的核心挑戰。本研究利用Dukascopy提供的2024 年 7 月 1 日至 2025 年 6 月 30 日十組 G10 貨幣對的逐筆報價數據，重新審視這一難題，並探討市場基本面與政策利率是否包含可供匯率預測的訊息。我們比較無漂移隨機漫步、固定與滾動視窗之普通最小平方法結構模型，以及帶有漂移的布朗運動（Brownian motion），並使用均方根誤差（Root mean square error, RMSE）衡量各模型預測性能。此外，我們借用統計力學大正則系綜（Grand canonical ensemble, GCE）描述，以愛因斯坦-玻色統計（Bose–Einstein statistics）建立報價規模與價格的關係，引入交易溫度及成交量累積震盪指標。
實證主要結果有三。第一，Meese–Rogoff 之謎於當代高頻資料完整重現，即使根據慷慨式設計使用與預測同期基本面的結構模型，其短時間預測之均方根誤差仍高於隨機漫步一至兩個數量級，且十組貨幣對無一例外。第二，布朗運動描述顯示，市場波動比政策利率差異的驅動力大幾十倍，這給出謎題之定量物理表述。第三、GCE框架成功地刻畫匯率市場狀態，並能捕捉政策利率變化對買賣壓力的影響，儘管目前尚未能建立點預測功能。這些結果凸顯市場噪聲在短期內對基本面和政策利率訊號的主導作用，並展示統計力學方法在理解匯率動態的潛力。

The Meese–Rogoff puzzle—the difficulty of outperforming a random walk in exchange-rate forecasting—remains a central challenge in international finance. This study revisits the puzzle using tick-by-tick foreign-exchange data from Dukascopy for ten G10 currency pairs from July 1, 2024, to June 30, 2025, and investigates whether market fundamentals and policy interest rates contain exploitable information for exchange-rate prediction. We compare a random walk without drift, Ordinary-Least-Squares (OLS) structural models with fixed and rolling windows, and Brownian motion with drift, evaluating predictive performance using root mean square error (RMSE). We further examine a statistical-mechanical description of the foreign-exchange market based on the Grand Canonical Ensemble (GCE), using Bose–Einstein statistics to characterize the relationship between quote size and price and introducing measures of trading temperature and the Volume Accumulation Oscillator.
Three main findings emerge. First, the Meese–Rogoff puzzle is fully reproduced in contemporary high-frequency data: even structural models with contemporaneous fundamentals available for prediction produce RMSEs one to two orders of magnitude larger than those of a random walk across all ten currency pairs. Second, a Brownian-motion description reveals that market fluctuations are tens of times larger than the driving force associated with policy-rate differentials, providing a quantitative physical interpretation of the puzzle. Third, the GCE framework successfully characterizes market states and captures the influence of policy-rate changes on buying and selling pressure, although it does not yet provide point forecasts. These results highlight the dominance of market noise over fundamental and policy-rate signals at short horizons and demonstrate the potential of statistical-mechanical approaches for understanding exchange-rate dynamics.

中文文獻
陳旭昇（2013）。《時間序列分析：總體經濟與財務金融之應用》（二版）。臺北：東華書局。
英文文獻
Bacchetta, P., van Wincoop, E., & Beutler, T. (2010). Can parameter instability explain the Meese–Rogoff puzzle? In NBER International Seminar on Macroeconomics 2009 (pp. 125–173). University of Chicago Press.
Bank for International Settlements. (2025). OTC foreign exchange turnover in April 2025: Triennial Central Bank Survey. https://www.bis.org/statistics/rpfx25_fx.htm
Bicci, A. (2016). Limit order book and its modelling in terms of Gibbs Grand-Canonical Ensemble. Physica A: Statistical Mechanics and its Applications, 463, 516–524.
Clark, T. E., & West, K. D. (2007). Approximately normal tests for equal predictive accuracy in nested models. Journal of Econometrics, 138(1), 291–311.
Diebold, F. X., & Mariano, R. S. (1995). Comparing predictive accuracy. Journal of Business & Economic Statistics, 13(3), 253–263.
Engel, C., & West, K. D. (2005). Exchange rates and fundamentals. Journal of Political Economy, 113(3), 485–517.
Fama, E. F. (1984). Forward and spot exchange rates. Journal of Monetary Economics, 14(3), 319–338.
Giacomini, R., & White, H. (2006). Tests of conditional predictive ability. Econometrica, 74(6), 1545–1578.
Karth, M., & Peinke, J. (2002). Stochastic modeling of fat-tailed probabilities of foreign exchange rates. Complexity, 8(2), 34–42. https://doi.org/10.1002/cplx.10068
Meese, R. A., & Rogoff, K. (1983a). Empirical exchange rate models of the seventies: Do they fit out of sample? Journal of International Economics, 14(1–2), 3–24.
Meese, R. A., and Rogoff, K. (1983b). The Out-of-Sample Failure of Empirical Exchange Rate Models: Sampling Error or Misspecification? In Exchange Rates and International Macroeconomics, edited by Jacob A. Frenkel, 67–112. Chicago and London: University of Chicago Press.
Renner, C., Peinke, J., & Friedrich, R. (2001). Evidence of Markov properties of high frequency exchange rate data. Physica A: Statistical Mechanics and its Applications, 298(3–4), 499–520.
Rossi, B. (2013). Exchange rate predictability. Journal of Economic Literature, 51(4), 1063–1119.
Tusset, G. (2026). The rise of econophysics: A connected history of two disciplines. Cambridge University Press.
Viaggiu, S., Lionetto, A., Bargigli, L., & Longo, M. (2012). Statistical ensembles for money and debt. Physica A: Statistical Mechanics and its Applications, 391(20), 4839–4849.
Lemons, D. S., & Gythiel, A. (1997). Paul Langevin's 1908 'On the Theory of Brownian Motion' ('Sur la théorie du mouvement brownien'). American Journal of Physics, 65(11), 1079–1081.