加權指數之買賣策略之研究

股票雖具有高報酬率的特點，然而有買賣決策的問題。過去使用技術分析進行買賣，在參數敏感度調整會有很大的差異，以及研究使用的數據相互影響下，不容易窺探出技術指標的穩定性。所以針對某些特徵性的年份，選擇適當的技術指標、參數及研究這些指標的穩定性有其重要性。本研究以台灣加權指數的數值當作指標，蒐集近年來較具特徵性的幾個年份，選擇股票技術指標(布林通道、RSI、KD)，利用這些技術指標相互搭配，找出最佳的適用參數，並觀察這些參數在這些特徵性的年份間的變化。本文的研究結果顯示: 1. 在2011年加權指數跌幅21.46%的情況下，使用RSI絕對值的交易策略可達最高報酬率21.67%；在2017年加權指數漲幅為14.77%的情況下，使用三種技術指標策略並沒有明顯優於買進持有策略；在2018年加權指數跌幅9.18%的情況下，使用KD絕對值的交易策略可達到最高報酬率18.37%。 2. 在熊市下，單獨使用停損停利的絕對值方式當做賣出方式，其報酬率差於大盤；RSI或KD中使用絕對值的交易略策比較中，RSI更加穩定。在RSI或KD利用時間參數小的穿越時間參數大的策略比較中，RSI多數報酬率較大盤差，而KD有一半以上較大盤好的報酬率。在RSI下使用絕對值策略狀況下，較穿越策略好，然而在KD下使用絕對值策略和穿越策略兩者並沒有明顯優劣，並且在這兩種指標中可發現指標鈍化現象並不明顯。 3. 在這些特徵年份下，使用布林通道策略中，我們提出一種特徵式買賣方式，其報酬率穩定優於大盤，且比其他布林通道一般策略好。 4. 在這些特徵年份下，探討價量關係，傳統的價量八大定律並沒有成立，另外我們發現有一些簡單現象，在價格發生上漲或下跌較大波動時，成交量較原先增加現象並不明顯，僅在2017時特定情況下有較為明顯的漲跌關係；但成交量產生大波動時，價格沒有相對應變化(定律:量先價行是不成立)。

關鍵字

台灣加權指數；布林通道； RSI ； KD ；量價關係

並列摘要

Although stocks have the characteristics of high returns, there are problems with buying and selling decisions.In the past, technical analysis was used to buy and sell. Under the parameter sensitivity adjustment, there will be great differences, and the data used in research will affect the stability of technical indicators.Therefore, for certain characteristic years, it is important to choose appropriate technical indicators, parameters, and study the stability of these indicators. This study uses the value of the Taiwan Weighted Index as an indicator. Collecting several characteristic years in recent years,Select stock technical indicators (Boulin Channel, RSI, KD), use these technical indicators to match each other to find the best applicable parameters, and observe the changes of these parameters between these characteristic years. The findings of this article show: 1. In the case of a weighted index decline of 21.46% in 2011, the maximum return rate of 21.67% can be achieved by using the trading strategy of absolute RSI;In the case of a weighted index increase of 14.77% in 2017, the use of three technical indicator strategies is not significantly better than the buy and hold strategy;With the weighted index falling by 9.18% in 2018, the maximum return rate of 18.37% can be achieved using the trading strategy of KD absolute value. 2. Under a bear market, the absolute value method of stop loss and stop profit alone is used as the sell method, and the return rate is worse than the market;In the comparison of trading strategies using absolute values in RSI or KD, RSI is more stable.In the strategy comparison of RSI or KD using a small time parameter and a large time parameter, RSI mostly has a large return difference and KD has more than half of a large return.Under the condition of using the absolute value strategy under RSI, it is better than the crossing strategy. However, using the absolute value strategy and the crossing strategy under KD has no obvious advantages and disadvantages, and the index passivation phenomenon is not obvious in these two indicators. 3. Under these characteristic years, in using the Bollinger Band strategy, we propose a characteristic buying and selling method that has a stable return better than the broader market and better than other Bollinger Band general strategies. 4. In these characteristic years, to discuss the relationship between price and volume, the traditional eight laws of price and volume have not been established.In addition, we found that there are some simple phenomena. When the price fluctuates greatly, the increase in trading volume is not obvious than the original one. Only in 2017 under certain circumstances has a more obvious relationship of fluctuation.But when the trading volume fluctuates, the price does not change correspondingly (the law: the quantity first price line is not established).