技术指标的互补使用策略

技术指标是基于交易工具当前以及之前的价格或成交量情况的数学统计方法。当被用作技术交易策略的一部分，指标可以帮助交易者识别特殊的、容易被肉眼忽略的交易机会。然而，如果你使用多种衡量价格走势或成交量指标的话，你的分析将很容易受到迷惑。

这里，我们将会回顾多种不同的技术指标，展示如何使用互补的技术指标来提高交易策略，并解释选择不同种类指标以避免多重共线性（使用多种类似指标造成的一种情况）的重要性。

交易指标的类型
根据每组指标的作用，通常将其分为下面四种类型：
趋势指标衡量趋势的方向以及强度，通常结合某种价格平均线形成的基线使用。动能指标通过对比随着时间推移形成的价格变化来追踪价格的变化速度。波动性指标可以提供一个给定市场的交易范围、市场的加速或减速等信息。成交量指标表示已发生的交易活动的总量并分析价格波动的力量。

大多数情况下，指标本身不提供交易信号。交易者必须根据自己的交易逻辑，风格，风险承受能力，所选择的交易工具来解读交易指标的信息。将几种不同种类的指标综合到一个交易策略中（比如使用一个趋势指标以及一个动能指标）将会比使用同类指标带来更好的交易结果。

使用技术指标很重要的一点是试验将它们应用于市场交易的各种创新方法。虽然技术指标通常有推荐的使用参数，但是交易者不能让自己局限于这些规则。交易者可以依靠自己的研究以及对市场的观察来建立属于自己的交易指标。

综合使用指标
为了提高市场分析（或者是成功交易系统的可能性），交易者需要使用互补的指标——那些相互协作，但是可以对市场走势提供不同观点以及可以帮助你做出交易决定的有意义数据。因此互补指标可以提供特别的——而不是冗余的——信息，它们可以结合在一起使用来确认交易信号。

交易策略所使用的每一个指标都应该服务于不同的目地并可以就市场走势提供独特的视角——比如用来衡量动能或者评估总体趋势。比如所谓的用于简明扼要地介绍一个产品的电梯演讲。交易者需要准确的说出指标存在于交易图表中的目的。回答下面两个问题十分重要：是什么使得交易指标成为该交易策略的重要一部分？该指标可以提供什么样独特的信息？

我们可以比较两个交易系统来展示使用互补指标的优点。在第一个交易系统中，只有一个技术指标使用“止损反转”策略来提供交易信号。第二个交易系统，增加了一个趋势过滤器，如果条件不满足，交易就不会执行。

上面“简单的交易系统”显示了只使用随机摆动指标（一个动能指标）的“止损反向”交易策略，该指标周期为30，两个平滑参数为3。当满足下面两种情况时做多：

• K慢线穿过D慢线
• K慢线在超卖区域20下方

满足下面情况时做空：

• K慢线向下穿过D慢线
• K慢线在超买区域80上方

测试使用迷你标普尔500期货的十年历史数据（2002.6.17~2012.6.19），表现报告汇总在上述图中。报告显示，该指标的盈利系数为1.50，表明是可以盈利的。虽然盈利系数是可以接受的，但是如果添加一个互补交易指标的话，结果将会更好。
该指标在添加了一个可以帮助其判断趋势方向以及强度的指标之后表现更好。因此，一个简单的交叉易移动平均线，时段分别为50和60.这就为交易设置了一个额外的入场条件：
对于做多，50日移动平均线必须在60日均线上方。做空，50日移动平均线需要在60日移动均线下方。
下图显示了同时使用随机指标以及移动平均线指标的交易系统。在增加了交叉移动平均线指标作为过滤器之后，交易结果大大改善。随机指标的盈利概率相对稳定在73.17%，结合了移动平均线指标之后盈利概率达到77.42%，平均净盈利以及别的关键衡量指标都取得了显著进展。通过使用一个互补指标，我们可以改善交易策略标准，并得到一个可以取得更多盈利的系统。

注意多重共线性
虽然使用多种指标的交易策略可以改善交易结果，但是错误的使用指标组合却会产生危害。一些交易者可能会故意选择相似的而不是互补的指标——为了相互印证一个交易信号。使用类似指标的问题是——比如两种动能指标——它们实际上不会相互验证，因为它们的只会产生重复的交易结果。这两种指标貌似是在互相验证，但是这仅仅是因为它们都是基于同一个数学计算公式。
多重共线性是一个统计系统，意思是对于同样的信息进行多重统计。当类似的指标同时使用这种情况即发生。如果关联程度非常高的指标用来量化市场活动性，比起那些没有使用不同种类指标的交易策略，它的意义性要逊色很多。如果所有的指标都基于同样的收盘价，比如，没有产生新的信息，因此，不会产生真正意义上的相互确认。这就类似于在一个部门相似的市场条件下错误地试图使自己的投资组合多样化。

确认指标在市场中是否具有同样的作用的一个简单的方法就是在一张图中回顾它们的交易结果。如果每个指标提供相似的信号并且以大概相同的方式移动，那么它们就很有可能是共线性的，不应该在一起使用。
迷你标普500指数显示了“重复努力”：随机指标与顺势指标CCI同时使用。它们两者都是动能指标，并且同时，例如，周期设置为30.我们可以从图表中看到，这两个指标几乎相同的结果。交易策略中使用这样的指标只会带来害处，因为它们产生冗余的交易信号，并且没有真正的彼此验证。第二个指标不应该产生同样的交易信号，并且它应该减少第一个指标产生的信号并且确认更成功的信号——提高第一个信号的成功几率或者数量。

交易者和投资者使用技术指标评估过去以及当前的市场行情并预测未来的价格走势。使一个交易策略的成功率最大化，很重要的一点是，每一个技术指标都提供不同的观点，并且对过去以及当前的价格活动协同合作提供全面的观点，这样交易者可以建造一个适合未来价格波动的交易模型。

 
A complementary approach to trading technical indicators

Technical indicators are mathematical calculations based on a trading instrument’s past and current price or volume activity. When used as part of a technical trading strategy, indicators can help traders identify unique opportunities in the markets that could be overlooked by simply viewing a price chart. However, you also can cloud your analysis by tracking too many indicators that measure the same qualities of price and volume.

Here, we will review the different types of technical indicators, demonstrate how to apply complementary technical indicators to enhance a trading strategy and explain the importance of selecting dissimilar indicators to avoid multicollinearity, a condition that results from employing multiple similar component techniques.
Indicator types

Indicators often are divided into four categories based on what each group measures:

Trend indicators measure the direction and strength of a trend and typically use some form of price averaging to establish a baseline. Momentum indicators track the speed at which prices change by comparing prices over time. Volatility indicators provide information about the trading range in a given market and its acceleration and deceleration. Volume indicators represent the amount of trading activity that has occurred and analyze the force behind a price movement.

For the most part, indicators by themselves do not provide trading signals. Instead, each trader must interpret the information to determine trade entries and exits based on his or her own trading logic, style, risk tolerance and even chosen trading instrument. Incorporating different types of indicators into a trading strategy (such as by applying one trend and one momentum indicator) will provide better results than using multiple indicators of the same type.

An important part of using technical indicators is to experiment with innovative ways of applying them to the markets. While technical indicators often have a recommended application, traders are by no means limited to those rules. Traders can rely on their own research and market observations to build a better mousetrap.

Combining indicators
To improve market analysis (and the odds of a successful trading system), traders should use complementary indicators — those that present different views of the market while working collectively to provide meaningful data on which to base trading decisions. Because complementary indicators provide unique — rather than redundant — information, they can be used together to provide confirmation for trading signals.

Each indicator that is used in a trading strategy should serve a distinct purpose and provide a unique perspective of the markets — such as to measure momentum or evaluate the overall trend. Like a so-called elevator speech used to explain a product’s purpose in a succinct, pointed delivery, traders should be able to explain concisely why an indicator is on a chart. Answering two questions is paramount: What is it that makes this indicator an important part of the strategy? What unique information does this indicator provide?

We can compare two trading systems to demonstrate the advantages of using complementary indicators. In the first system, only one technical indicator is used to generate trading signals for a simple stop-and-reverse strategy. For the second system, a complementary indicator is added that provides a filter, or confirmation, for the trades. If the conditions are not met, the trade will not be initiated.

“Simple system” (above) shows the basic stop-and-reverse strategy that uses only a stochastic oscillator (a momentum indicator) with the settings 30 for the length, 3 for the first smoothing input and 3 for the second smoothing input. The strategy enters a long trade when two conditions are met:
• SlowK line crosses over SlowD line
• SlowK line is below the oversold territory of 20

A short trade is entered when:

• SlowK line crosses under SlowD line
• SlowK line is above the overbought zone of 80

Ten years of historical E-mini S&P 500 futures data were included in the test (June 17, 2002 through June 19, 2012). The summary performance report is displayed on the chart. This tells us that the strategy shows a profit factor of 1.50, indicating that this strategy would be profitable. While the results are acceptable, adding a complementary indicator can improve performance.

The system is enhanced using a complementary trend indicator to help determine the direction and strength of the trend. For this, a simple moving average crossover is used with lengths of 50 and 60. This adds one additional condition to trade entries:

• For long trades, the 50-period average must be greater than the 60-period average.
• For short trades, the 50-period average must be less than the 60-period average.
“Trend filter” (below) shows the strategy that uses both the stochastic and moving average indicators. With the addition of the moving average crossover filter, the results are improved significantly. While the percent profitable remains relatively steady, 73.17% for stochastic only vs. 77.42% with the moving average filter, the average trade net profit and other key metrics show noteworthy improvement. By adding a complementary indicator, we are able to refine the strategy’s criteria to create a more profitable system.

Beware multicollinearity
While using multiple indicators in a trading strategy can improve results, using the wrong combination of indicators can hurt a system. Some traders may intentionally apply similar — rather than complementary — indicators with the objective of finding confirmation for a trading signal. The problem with using similar indicators — such as two momentum indicators — is that it really provides no confirmation at all because it produces only duplicate results. The two indicators may appear to confirm one another, but this is only because each is based on the same math.

Multicollinearity is a statistical term that refers to the multiple counting of the same information. It occurs when similar indicators are used at the same time. If highly correlated indicators are used to quantify market activity, the results are less meaningful than if different types of indicators are used as part of a strategy. If all indicators are based on the same closing prices, for example, no new information is produced, and therefore, no true confirmation is provided (see “Replicating efforts”). It is akin to trading similar markets in one sector in a mistaken attempt to diversify your portfolio.

A simple method of determining if indicators measure the same thing about the markets is to review their results on a single chart. If each indicator provides similar signals and they move in the same general manner, they likely are collinear and should not be used together. “Replicating efforts” shows a daily chart of the E-mini S&P 500 with a stochastic oscillator and the Commodity Channel Index (CCI) indicator applied. Both the stochastic and CCI are momentum indicators, and both, for this example, have been set to the same length of 30. We can see from the chart that the indicators produce nearly identical results. Using these two indicators together would be detrimental to a strategy because they produce redundant signals with no true confirmation. A second indicator should not catch the same signals, instead it should reduce the number of signals and confirm the more successful — either by percentage or magnitude — signals of the first indicator.

Traders and investors use technical indicators to evaluate past and current market conditions to predict future price movements. To maximize the odds of success for a trading strategy, it is important that each individual indicator provides a different view, and collectively a comprehensive view, of past and current activity so that traders can build a trading model better equipped to forecast the future ups and downs of market prices.

本文翻译由兄弟财经提供

文章来源：http://www.fxstreet.com/education/technical/a-complementary-approach-to-trading/2012/10/11/