This thread was broken off of my opening range breakout thread. I started a tutorial on developing a trading system there and it was recommended to me that I start a new thread. The original thread's link was. http://elitetrader.com/vb/showthread.php?s=&threadid=54765&perpage=6&pagenumber=26 I started this discussion on page 26 of this thread and will move over these early links here. In this tutorial I am building a trend following system combined with opening range breakout. If you want more information on opening range breakout I have a free tutorial on this topic on TradersStudio.com. This tutorial is part of my video on opening range breakout. The link to the free part of this video is : http://www.tradersstudio.com/Default.aspx?tabid=129 I have two chapter of the video on TradersStudio.com for free. The first one is totally free and the second one requires you to register on the site which is also free.
Let's start our series on developing a trading system. We will test the premise that many trends are noisy and we get a market breakout, pullback and then a real move. We will develop a trend following system which trades on a pullback in a trend after a trigger to indicate the trend has continued. The case study will walk you though the steps including things that don't work. I want you to be able to follow along with the process. This system is composed of three main components. A trend detector. A pullback detector and then a entry method which confirms the continuation of the trend. We will start testing our model as a stop and reverse system, later we will see if we can improve it by developing exits. We will start our next installment with testing several different trend components on this basket.
Here are the markets we will be using in this case study Corn , Euro Currency, Live Hogs, Australian Dollar Copper , Municipal Bonds Cocoa , Heating oil Natural Gas, Canadian dollar Unleaded Gas, Palladium Crude oil , Orange Juice Soybeans , Cotton Japanese Yen, Sugar Dollar Index, Coffee Swiss Franc, Eurodollar Lumber, Ten-Year Note Live Cattle, soybean oil We will be using back adjusted data for this analysis. We will start our analysis on the trend detection module in our next installment.
Let's now discuss developing our Trend following system. We need to develop the trend detection component. The first one I will discuss is the triple moving average crossover. The triple moving average crossover works different than the dual moving average crossover. The dual moving average crossover has many problems. One example is if I use a short term moving average which is 1/2 the dominant cycle and a long term equal to the dominant cycle , the system is 180% out of phase with the market and I will be buying tops and selling bottoms, except in cases on major trends. When we use three or more moving averages we don't have this problem. I have optimized a triple moving average system on this basket. The results are attached. I deducted $75.00 slippage and commission. Please view attached report and system code for the optimization run. In the next installment we will discuss finding good values for the optimization.
In tomorrow installment we will discuss how to pick the best set of parameters from the optimization runs. Next we will analyze these results. This simple 3 MA crossover is not the system, it is a piece of it. We will be adding a few more layers on top of this over the next few days. My goal here was to walk though the process both good and bad. I am actually developing this system on the fly as part of this case study. If this system does not work we will be changing direction. I though it was a good case study to do this for Elite Trader members. This way they could see how a system developer works though his development.
The whole topic of picking the best set of parameters from a optimization is a bit of an art. You almost never want to pick the best performing set of parameters. You want to pick a set of parameters which does well, and has similar sets of parameters with similar performance. Curve fitting is a big problem in optimization, but when we are picking the best set of parameters across a basket of markets that helps reduce this problem. Here is how I am going to do this. Please download the optimization report in Excel. I would like you to post your analysis on what sets of parameters you would pick. Yes, this is pretty robust so there is more than one good answer. What I want to see are your reasons why you picked them. My goal is to force you to follow though the analysis and explain your answers. I will give you feedback as part of the learning process. Tomorrow night or Tuesday morning I will post my critics and my analysis. I will then discuss the next step in this system development process.
You have includes some of the most illiquid futures in the USA and, simultaneously, you have omitted ALL futures outside the USA. Inexcusable. Thursday's volume on March06 Palladium was less than 500 contracts, March06 Lumber was less than 600 contracts, and March06 MuniBond was five (5) contracts. Way to go mister big shot Educator. Among the spectacularly trending non-USA markets that you didn't include, Thursday's volume on March06 Japanese Govt Bonds was 24,000 contracts, March06 EuroBund was 1.2 MILLION contracts, March06 EuroSTOXX was 591,000 contracts, and March06 London Robusta Coffee was 6,800 contracts. Maybe you're not aware that futures are traded outside the USA, or that some of the most beautifully trending markets are in the Eastern Hemisphere. It's true, even if you choose to ignore it.