在使用White的真實性檢定和Stepwise Multiple Test消除資料勘誤的問題之後，有些技術分析確實可以擊敗大盤，在1989到2008，DJIA, NASDAQ, S&P 500, NIKKEI 225, TAIEX這五個指數中。但是在較不成熟的市場或較過去的時間內，我沒辦法找到任何強烈的關係在這些市場與超額報酬間。還有學習策略通常沒辦法獲得比簡單策略更好的表現，代表使用過去最好的策略來預測未來並不是個好主意。我同時還發現在熊市比穩定的牛市更有可能擊敗買進持有的策略。 / In five indices, DJIA, NASDAQ, S&P 500, NIKKEI 225, TAIEX, from 1989 to 2008, some technical trading rules indeed can defeat the broad market even after using the White reality check and stepwise multiple test to solve the data snooping problem. But in the markets like less mature ones or the one which was in the older period, I can’t find a strong relation between these markets and the excess return in my research. And the learning strategy usually can’t have a better performance than the simple one, means applying the rule which had a best record to forecast the future may not be a good idea. I also found that it is more likely to beat the buy and hold strategy when there is a bear market but not a steady bull market.

The price of stocks, futures, commodities and currency are for ever changing. Anyone interested in financial prices soon discovers that changes in prices are frequently substantial and are always difficult to forecast. This paper describes the behavior of prices from a statistical perspective. Specifically, employ several technical trading rules to uncover the trend of futures price movement and attempt to make profit out of the trend. In this paper, trading of seven technical trading systems is simulated for three futures contracts from September 1998 to March 2005 to test for market disequilibrium. The results differ by trading systems. Four systems produced positive mean net returns and five systems produced positive gross return when optimal parameters were used. These results indicate that, there exist opportunities to design profitable trading systems for futures markets.

效率市場理論認為資本市場上所進行的是一場公平遊戲，任何人在任一時點皆無法利 用他所能夠獲得的資訊，而持續地獲取超額利潤，因而市場也將總是處於均衡狀態。 本研究嘗試從經濟面及非經濟面，去探討目前台灣的證券市場，是否具備符合效率市 場的條件，其方法是運用產業分析的架構，來了解台灣的證券市場。 其次再以目前在台灣使用較廣泛的幾種技術分析指標做為研究工具，看看這些個別的 技術分析指標及各種指標的組合能否獲致超額報酬，以驗證市場是否符合弱式效率市 場假說。 本研究以運轉率為選取股票樣本的標準，研究期間為民國76牛蛙77年，使用的技術分 析方法包括：濾嘴法則、移動平均法、乘離率、強弱指標、人氣指標等，股價資料則 以證券所之日資料。 依據本研究所得的結果，當可明瞭目前台灣證券市場上經濟因素及非經濟因素對於市 場效率性的影響，並對於投資人所熱衷的技術分析是否真能獲利得到一些啟示，而這 些資料除可供主管機關，用以擬定提高證券市場功能之決策時的參考外，廣大的投資 人亦可因而了解其在投資時應採行何種分析工具，才能獲取較高的報酬。如此對於證 券市場效率性的提高將有若干的貢獻。

在這篇文章裡我們採用了White所提出的真實檢驗法(Reality Check)來解決探勘資料偏誤(Data-snooping bias)的問題，結果顯示從1980年到2008年間，技術分析法則的確可以幫助投資人在日圓兌美元及英鎊兌美元這兩個外匯市場獲利；我們也發現在外匯市場最普遍的技術分析方式─移動平均法(moving average)表現不如其他的技術分析法則，而通道突破法(channel break-out)的表現則明顯優於其他技術分析法則。 除了檢驗技術分析方法的獲利性之外，我們也嘗試著探討技術分析方法的獲利性與央行干預之間的關係，追隨Szacmary與Mathur在1997年所發表的論文，我們把技術分析法則擴充為在真實檢驗法中所使用到的所有法則，並且盡可能加長分析的期間。結果顯示技術分析法則的獲利與央行干預並不存在任何特定的關係。 / In this paper we construct a huge universe of simple trading rules and apply White’s Reality Check to mitigate data-snooping bias then detect the profitability of technical trading rules. We find that technical analysis is useful no matter in the full sample time or each subsample period. The channel break-out method outperforms the other methods in our finding while the profitability of the most commonly used moving average method is worse than the others. Furthermore, we inspect the relationships between the returns of technical trading rules and central bank intervention. The results suggest that there’s no evident relationship between the return series of trading rules and central bank intervention and are not consistent with the view of our following previous study.

技術分析一直是許多研究的熱門主題，也被眾多市場參與者廣泛運用在市場交易，而最普遍且最受歡迎的技術分析工具即為移動平均法。 本研究設計三種移動平均交易方法（一種只考慮收盤價，一種考慮收盤價及交易量，而另一種則將交易量作為收盤價的權重），每種交易方法皆使用五天為短期移動平均天數，十天、五十天、一百天、一百五十天、兩百天為長期移動平均天數，總計十五種移動平均交易規則，運用在臺灣中型100成分股以產生買進與賣出訊號，並依訊號進行交易動作，進而在未考慮交易成本的假設下計算出單次交易的平均報酬、平均持有天數，以及Hit ratio（正報酬的交易次數占總交易次數的比例），藉以探討移動平均法在此種股票的獲利性。而以交易量為價格權重來產生移動平均交易方法是基於相信帶有較高交易量的價格較有意義，盼藉以測試此種方法是否正如預期，相較於一般傳統的價格移動平均法有更好的績效。 本研究雖然未考慮交易成本，但呈現的單次交易平均報酬可以提供讀者與實際臺灣股票市場交易成本作比較，藉以了解考慮交易成本後的情況。而本研究除了呈現所有成分股單次交易的平均報酬、平均持有天數及Hit ratio的平均值，也將成分股依照ICB行業分類指標分成幾個主要產業，並呈現各產業內成分股的平均值，企圖了解特定交易方法是否在特定產業有較好的績效。 結果顯示，產生最好績效的移動平均交易方法也僅能有一半的交易次數得到正報酬，而就整體而言，將交易量作為價格權重的移動平均方法，也沒有產生相較於傳統價格移動平均法更好的績效，因此可以說，這類的技術分析對於這些股票無法有較好的績效。 / Technical analysis has been widely studied and used by many researchers and market participants. The most common and popular technical trading rule is moving average since it is mathematically well defined and used by many analysts. This article examines the profitability of technical analysis for FTSE TWSE Mid-Cap Taiwan 100 Index constituents under the hypothesis of no transaction costs. It uses three strategies (Price Strategy, Price and Volume Strategy, and PV Strategy) and fifteen moving average rules to generate buy and sell signals, and then compute average returns per trading, average holding days per trading, and hit ratios to see the profitability. It is believed that prices come with high volumes are more meaningful than those with low volumes. All of these strategies and trading rules are not only used for all constituents of FTSE TWSE Mid-Cap Taiwan 100 Index without consid-ering industry classifications but also for each major industry classifications of these constituents. Therefore, we can understand whether specific trading rules have better performances for specific industries of these stocks. The results are not that optimistic. Overall Price and Volume Strategy has the best results of hit ratio, however, the highest value is barely 50%, which means it can only have a half trading times positive returns. As for PV Strategy which uses weighted price moving average to trade, the performance has no significantly better than using simple price moving average rule. It can say that Technical Analysis like moving average can hardly have good performances on these stocks.

學術上不接受技術分析等方法，認為股價已經在市場上充分反應，過去的歷史股價不能對未來進行預測。然而，業界或一般的投資人，卻往往把技術分析拿來做為買賣的依據。實際上以歷史資料做模擬交易，卻可以發現許多技術分析的法則在某些市場、股票、期間之中，可以獲得相對於買進賣出更好的報酬。有趣的是，任何一種操作法則或是特定一組參數選擇，在樣本外的操作則無法完全發現同樣的結果。故以技術分析所獲得的超額報酬，究竟是此機制有效還是單純運氣成分，許多技術分析的文獻以及著作往往著墨甚少。 本論文利用狀態轉換模型(Regime Switching Model)捕捉台灣加權股價指數，將股價的動態分為上漲以及下跌兩種狀態，並估計其市場參數—漲跌速度、漲跌速度標準差、轉換機率。其次將所估計的市場參數做為模擬的依據，可發現在單純隨機的環境下，某些市場參數組合存在移動平均線的交易策略明顯優於買進持有策略。研究中以敏感度分析的方法，呈現各個單一市場參數的改變情形，對於操作績效影響的方向。 最後將2001~2010的的台灣加權股價指數，估計市場參數並找尋當下最適的移動平均組合，允許每季重新調整參數，並實際以收盤價做為買賣模擬。結果發現移動平均線操作，確實能提供比買進持有更好的報酬，並減低每年報酬率變異。

為了讓一般的投資大眾能享有智慧型、系統化的策略交易環境，本研究計畫發展一個可供大量使用者共用、並且容易上手的策略交易平台。為了達到這個目的，此平台必須擁有快速且大量的運算能力，雲端運算所提供之大量且可擴充的運算能力，使之成為最適宜的平台。為滿足不同使用者不同的投資偏好，此平台提供多項常用之技術指標與K線型態辨識功能讓使用者利用基因演算法產生符合其偏好的交易策略。在策略產生之後，使用者可以在平台上檢視交易策略在不同商品、不同時間區間上的表現，並從最後的策略回測報告中加以評估，挑選出獲利能力、波動程度與交易頻率都符合需求的交易策略。

由於技術分析之有效性一直為學者們所探討且備具爭議的議題，有部分學者認為技術分析無效，但實務上，技術分析在金融市場上的應用卻相當廣泛；此外，由於外匯市場為交易量龐大且眾多學者研究之重要金融市場，故本文回顧過去多項國內、外研究，並經由實地探訪台灣外匯市場上之投資人，針對仍未被研究，但實務上所採用之技術指標進行多項交易策略之模擬並探討其獲利性。 本文主要針對新臺幣兌換美元之匯市，探究將商品通道指標(Commodity Channel Indexes)及威廉指標(Williams Overbought/ Oversold Index)兩種技術分析指標應用於此外匯市場上之獲利情形，採用1993年1月1日至2012年12月28日，共計二十年，5279筆銀行間交易之新台幣兌美元之匯率日間資料，擷取其最高價、最低價及收盤價並透過程式交易進行回溯測試，並針對此兩種技術指標建構多種交易策略，歸納並分析其中可獲得超額報酬之技術策略，期能找出獲得最佳投資報酬以及提高交易的成功機率與獲利能力之法則。 藉由Matlab運算處理後，將此兩種技術指標應用於過去20年之歷史價格進行回溯測試，本實證研究發現：1. 採用威廉指標之策略普遍績效都較採用商品通道指標來的好，不僅在總報酬率的表現上比較好，採用威廉指標所執行的交易獲利的機率也必較高。2. 由於此兩技術指標應用在新台幣兌美元之外匯市場上可得的績效高於買賣策略應用於此市場之平均年化報酬率，故證實此兩種技術指標應用在新台幣兌美元的外匯市場上，均可獲得超額報酬，此兩技術指標在新台幣兌美元市場是有效的、可獲利的。 / It has long been a controversial question to scholars whether or not technical analysis is efficient. Although some scholars believe technical analysis is useless, it has been broadly used in the financial markets for a long time. As foreign exchange markets are one of the most important financial markets with huge trading volume in the world, this paper reviews many past literature and extracts trading strategies from some real investors in Taiwan’s foreign exchange markets. Additionally, this paper focuses on testing the trading performance of applying the technical indexes which have not been researched in the academic field but have often been utilized in the real exchange markets. This thesis mainly concentrates on the exchange market of New Taiwan Dollar against US dollar and examines the trading performance of utilizing two technical indexes which have been used but not been researched in foreign exchange markets : Commodity Channel Index (CCI) and Williams Overbought/Oversold Index (WMS). The dataset of this paper is from January 1st, 1993 to December 28th, 2012, an overall of 20 years and 5279 times of daily NTD/USD exchange rates between banks. To complete the back-testing, this research utilizes the highest, lowest and close price from those materials and analyzes the technical strategies which obtain excess profits. By generalizing the results of those trading strategies, investors can find the best trading rules and increase the returns from applying these two technical indexes to foreign exchange market. The results of this research are as follows: 1. WMS’s performances are chiefly superior to the CCI’s performances. Not only the total profit rates of technical strategies from WMS are higher than the total profit rates of trading rules from CCI, but also the rates of profits on WMS always demonstrate a better result than the rates of profits on CCI. 2. Both of these two technical indicators can produce excess profits. Compared to the average annual return of buy-and-hold strategy in this market, both of the two technical indexes conduct a better performance. As a result, these two technical indexes are effective in NTD/USD market.

參考Fang 2003年研究方法架構，我們檢驗了結合策略(結合技術分析法則和時間序列模型)應用在六個亞洲股票市場。由於技術分析法則和時間序列模型皆可利用過去歷史資訊來預測報酬，所以結合策略的實證結果優於技術分析法則和時間序列模型。此篇中超額報酬的計算是與買進持有相比較下未考慮交易成本的超額報酬。實證結果顯示，結合策略在完整樣本中可以成功的預測資產報酬，在六個國家的平均上，結合策略的超額報酬為0.19%優於技術交易法則下的0.13%和時間序列模型下的0.17%。並且，發現在新興國家如台灣、泰國、馬來西亞和南韓的預測能力比在已開發國家市場如香港和日本還要來的好。預測能力可被低階的自我相關係數解釋。除此之外，發現我們的預測能力受到非同步交易的影響。非同步交易所造成的衡量誤差使得超額報酬下降，但是我們的預測能力還是存在的。 / Following Fang and Xu (2003), we examine trading strategies combining technical trading rules and times series forecasts on six Asian stock markets. Since both technical trading rules and time series models can exploit predictable components as function of past prices or returns, the combined strategies outperform both technical trading rules and time series forecasts. The excess returns before transaction costs for each rule and country are compared to a passive buy-and-hold strategy. The combined strategies are quite successful in predicting asset returns in full samples. On average the buy-sell returns for combined strategies are 0.19% much higher than 0.13% for technical trading rules and 0.17% for time series models. Besides, we also find that all three rules have more explanatory power in emerging markets such as Taiwan, Thailand, Malaysia and Korea than more developed markets such as Japan and Hong Kong. The predictability can be explained by significant low-order autocorrelations in returns. Moreover, excess returns (pre-trading costs) for both time series models and combined strategies can be partially attributed to the measurement errors arising from non-synchronous trading. The non-synchronous trading bias reduces but does not eliminate the predictive power of combined strategies.

本論文主要在探討以移動平均法則為基礎的簡單技術分析指標，以及時間序列模型在台灣股票市場是否具有獲利能力，研究期間為1987/01/01-2006/12/31共20年的樣本期間。我們發現只有使用(1,50,0)和(1,50,0.01) 這兩個移動平均交易法則時才有顯著的報酬；並以AR(1)-GARCH(1,1)-M作為時間序列的預測模型。研究發現在股價上漲的時候，技術分析指標的確有較好的預測能力；而在股價下跌時，利用時間序列模型有較佳的獲利能力。因為技術分析指標與時間序列模型分別捕捉到不同的資訊，將兩預測工具結合在一起應該可以得到一個更好的組合預測指標。本文的實證研究發現此一組合預測指標，不管是在多頭或空頭期間時，都可以比使用單一分析工具獲得更高報酬。