2017年11月29日 星期三

Equity curve analysis: A fool’s path? 淨值曲線交易:是愚笨的嗎?


先講結論,避免有一些人說我在造謠,我的目的是要透過文章佐證並且釐清一些事實
結論:本文作者認為,簡單的使用技術指標來去針對淨值曲線進行交易系統管控是不好也具有反效果的,但是或許有其他的不同方式或是思考點可以達成做這個動作的目標->停止表現差勁的策略避免交易

Equity curve trading is simply a methodology where a trading strategy is turned on and off based on the gyrations of the equity curve. While there are many different approaches to employing equity curve trading, the concept is straightforward: trade strategies when they are making money, and temporarily turn them off when they are not. If only it were so simple.

淨值曲線交易是一種簡單的跟隨著淨值曲線變化來控制交易策略(系統)開啟或關閉的方法,雖然有許多不同的方法可以使用淨值曲線交易,但是主要的概念還是來自於當交易策略(系統)賺錢時進行交易,當交易策略(系統)虧損時停止交易.

With many of the basics of equity curve trading covered in the first installment, this article will look at two different methods of equity curve trading applied to three different real life trading strategies. Reviewing the results may allow some generalizations and conclusions about equity curve trading to be made.
Strategies scrutinized


前一篇文章中包含許多淨值曲線交易的基礎知識,因此本文將探討適用於三種不同實際運用在交易策略的兩種不同的淨值曲線交易方法。 並且檢驗其使用淨值曲線後的結果來跟原始策略進行比對

Three unique, real world trading strategies will be examined with and without equity curve trading approaches. The streategies are the following:

我們將針對以下三種市場進行淨值曲線交易方法的測試及評估

Gold: Swing strategy based on daily bars trading during “pit” hours. 
黃金:採用基於日K特定時間之下的擺盪交易策略

E-mini S&P 500: Intraday strategy based on one-minute bars.
S&P500:採用基於1分鐘線圖的日內交易策略

Japanese yen: Swing strategy based on 360-minute bars.
日圓:採用基於360分鐘線圖的擺盪交易策略

For each strategy, the most recent 450 trades will be studied, each of which trade only one contract per trade. This study represents three-to-nine years of trades, depending on the strategy. All trades are either from live trading or from out-of-sample walk-forward backtesting. No in-sample results are shown.

對於每個策略,最近的450筆交易將被用於學習,每筆交易只交易一份合約。 這項研究代表了三到九年的交易,取決於戰略。 所有的交易都是來自實際交易結果,或者是來自外部樣本的歷史回溯測試。 沒有採用樣本內交易結果。

The closed trade equity curves for each of the strategies are shown in “Trading it straight” (below).
每筆交易的已平倉曲線(採用實際交易或樣本外)都顯示在下方的圖片中

 
Evaluation process評估程序

Two different methods of equity curve trading will be examined in this study. The first method using a moving average of the equity curve as the criterion for turning the strategy on and off. When the “always on” equity is above the moving average line, the strategy will trade as normal. But, when the “always on” equity curve drops below the moving average, the strategy will be temporarily turned off until the equity curve again crosses above the moving average.
本研究將考察兩種不同的淨值曲線交易方法。 第一種方法是使用淨值曲線的移動平均值作為開啟和關閉策略的標準。 當“接受所有訊號”的淨值高於移動平均線時,策略將正常交易。 但是,當“接受所有訊號”的淨值曲線跌破移動平均線時,策略將暫時關閉,直到淨值曲線再次超過移動平均線。

Different length moving averages will be examined. We will use 10-, 25- and 40-period moving averages.
我們將確認不同長度的移動平均線。 我們將使用10,25和40週期均線。

The second method of equity curve trading will be using a new equity curve low as the cutoff criteria. In the same manner as the moving average approach, the strategy will be turned off when a new X trade, low in the equity curve, is reached, and turned on when the equity curve is above this value.
淨值曲線交易的第二種方法是將新的淨值曲線作為停止標準。 與移動平均法相同,當心的交易結果所產生的淨值曲線在達到更低時,該策略將被關閉,當淨值曲線高於該值時,該策略將被打開。

To keep consistency between the two approaches, the 10-, 25- and 40-period lows of the equity curve will be used here too.
為了保持兩種方法的一致性,淨值曲線的10,25和40期最低值也將在這裡被使用。

These two filtering approaches are common in equity curve trading, but are by no means the only methods. Practically any indicator can be applied to an equity curve, in the same way these indicators can be applied to actual price data. This, of course, opens up a whole Pandora’s Box of possible approaches, with an infinite number of possibilities when indicator parameters are taken into consideration. Such a situation inevitably leads to over optimization and curve fitting, as the trader attempts to “tune” an equity curve trading approach to his particular equity curve.
這兩種過濾方法在淨值曲線交易中很常見,但決不是唯一的方法。 實際上,任何指標都可以應用於淨值曲線,就像這些指標可以應用於實際價格數據一樣。 當然,這屬於其中一種開啟潘多拉盒子的方法,因為當指標參數被考慮時,就有無限多種的可能性。 這樣的情況不可避免地將導致過度優化和曲線擬合,因為交易者試圖將淨值曲線交易方法“調整”到他特定的淨值曲線上。

Results測試結果

Results for the gold  strategy are shown in “Gold: Equity analysis” (below). As mentioned in the first installment, although net profit and maximum drawdown are important numbers, the ratio of the two (net profit/maximum drawdown) is of primary concern. A higher ratio is preferred because it tells us that we are receiving more profit for each dollar of drawdown.
黃金策略的結果顯示在“黃金:權益分析”(下圖)中。 正如第一批中所提到的,雖然淨利潤和最大虧損是重要的數字,但兩者(淨利潤/最大虧損)的比率才是首要關注的。 較高的比率才會是首選,因為這告訴我們,每損失一美元的風險,我們都會獲得更多的利潤。


 As seen in the table, equity curve trading for the gold strategy is almost always worse when compared to the “always on” approach. An example of the different equity curves for the moving average technique is shown in “Feedback effect” (below).  
如表中所示,與“接受所有訊號”的方式相比,黃金策略的淨值曲線交易幾乎總是比較糟糕。 “反饋效果呈現”(見下圖)顯示了使用移動平均技術的不同淨值曲線結果。



 In this case, it is clear that the equity curve trading approach is harmful. While steep drawdowns are stopped many times by the equity curve trading, significant upside trades also are missed, leading to a net effect of poorer overall performance. (Results for the E-mini S&P 500 strategy are shown in “E-mini S&P 500: Equity analysis,” below.) 
在這種情況下,淨值曲線交易方法顯然是有害的。 儘管淨值曲線交易產生多次陡峭的跌幅,但是在重大獲利的交易也被忽略,導致總體表現變得較差。 (E-mini S&P 500策略的結果顯示在下文的“E-mini S&P 500:權益曲線分析”中。)



 Equity curve trading for the E-mini S&P 500 strategy is better for the moving average approach, but worse for the new low approach. The degree of improvement is highly dependent on the choice for the look back period. There is no “one size fits all” approach for this particular strategy. (Results for the Japanese yen strategy are shown in “Japanese yen: Equity analysis,” below.)  
E-mini S&P 500策略的淨值曲線交易對於移動平均法來說更好,但是對於新的淨值曲線新低策略則更糟糕。 改善的程度高度依賴於回溯期間(週期性)的選擇。 對於這個特定的策略,沒有“以一概全”的辦法。 (日元策略的結果顯示在下面的“日元:權益曲線分析”中。)



As shown in “Japanese yen: Equity analysis,” equity curve trading for the yen strategy is always worse than just letting the strategy trade continuously, sometimes significantly worse. The new low technique is better than the moving average technique. Both methods underperform the “always on” case, sometimes by quite a large margin.
如“日元:權益曲線分析”所示,日元策略的淨值曲線交易總是比讓策略持續交易更糟,有時甚至更慘烈。採用淨值曲線新低技術則比一般的移動平均技術更好。 但是,這兩種方法的表現狀況都比“接受所有交易訊號”的情況差一些,有時候還有很大的差距。

Base is better 基本就是最好的

For these three real-life strategies, the equity curve trading approaches do not improve performance, except in a limited number of cases. This highlights three important issues with equity curve trading. 
對於這三種實際使用的策略,在這三種有限的情況之外,淨值曲線交易方法並沒有改善績效。 這突顯出了淨值曲線交易的三個重要問題。

First, the results depend on the strategy being traded. Not every strategy will act the same for each method of equity curve trading. This can be seen in the contrast between the E-Mini S&P 500 strategy, where equity curve trading may improve performance, and the Japanese yen strategy, where equity curve trading never improves performance. 
首先,結果取決於交易策略。並非每種策略對於每種淨值曲線交易方法都是一樣的。這可以從E-Mini S&P 500戰略(淨值曲線交易可能會提升業績)和日圓淨值曲線交易永遠無法改善績效的日元策略之間形成對比。

Second, the choice of trading technique is also critical. It may be that one particular approach improves performance, but a different approach harms it. This is evident in the E-Mini S&P 500 strategy, where only the moving average approach improves the strategy.
其次,交易技巧的選擇也很關鍵。這可能是一個特定提高績效的方法,但另一個不同的方法就會造成反效果。這在E-Mini S&P 500策略中是明顯的,只有移動平均策略才能改善策略。

Finally, even within one approach, the choice of parameter (the look back period in this case) for the equity curve trading has a large, but inconsistent, impact. This is clearly evident in the gold strategy, where a wide range of profit to drawdown ratios are possible, simply by changing the look back period.
最後,即使在一種方法中,淨值曲線交易的參數(回溯期間)的選擇具有很大但不一致的影響。這一點在黃金策略中顯而易見,通過簡單地改變參數(回溯期間),可以實現大範圍的利潤下降。

Equity curve trading has its vocal proponents in the industry. But the overall efficacy of the approach is certainly not crystal clear. In fact it is relatively easy to take a good strategy and significantly degrade its performance by employing equity curve trading. 
淨值曲線交易在業內擁有少數支持者的聲音。但是這種方法的總體功效當然不是很清楚。事實上,通過採用淨值曲線交易來開啟好的策略並顯著降低其績效反而是相對容易的。

While the overall objective of equity curve trading is unquestionable—to cease trading with poor performing strategies—it is probable that there are better ways of accomplishing that goal. This study indicates that equity curve trading with simple indicators has more downside than upside.    
雖然淨值曲線交易的目標是不容置疑的 - 停止表現差勁的策略避免交易 - 很可能有更好的方法來實現這一目標。這項研究表明,用簡單的指標進行淨值曲線交易將會造成整體績效有更多的下滑空間。

原文網址:http://www.futuresmag.com/2015/04/15/equity-curve-analysis-fool%E2%80%99s-path