This thread is a continuation of my intermarket thread. In this thread we will discuss predicting the energy markets using both indexes and mutual funds in the energy sector. I have been looking for a source of data I can share with everyone. QuotePlus was nice enough to supply for me 7 different data series with permission to redistribute. I am going to upload one of them on ET. The complete set is available for free on TradersStudio.com. All it requires is for you to be a registered member of TradersStudio.com, registration is free. If you want to download the data just login to TradersStudio.com If you want to register the link is. http://www.tradersstudio.com/UserRegistration/tabid/56/Default.aspx The link to the data page is as follows http://www.tradersstudio.com/TradersResources/Data/tabid/161/Default.aspx I am including the symbol !EXV , "SELECT SPDR ENERGY INDEX" on ET for you to see how well this intermarket technology works. All seven data series are available on TradersStudio.com
Let's start this topic, we will use a simple intermarket analysis divergence between positively correlated markets. If the intermarket is in a uptrend and crude is in a downtrend we buy crude. If the intermarket is in a downtrend and crude is in a uptrend we sell crude. We defined then trend in crude based on if the close of crude was above or below a 40 day moving average of crude We defined the trend in SELECT SPDR ENERGY INDEX based on if the close of it was above or below it's 25 day moving average. This simple system was tested from 7/5/2000 to 5/22/2006. We did not deduct anything for slippage or commissions. We just wanted to see how predictive the intermarket relationship are using these this data series. During this period the system made $70,540 in net profits and over $20,000 on the short side. The average trade was about $1500.00 and the average short trade almost $900.00!. Most trend following systems lost money on the short side during this time. Just to give you a idea, channel breakout using a 20 day breakout only made $110.00 net profit both sides because it lost $14,000 on the short side. This simple intermarket system made over $20,000, just on the short side. Please register on TradersStudio.com and download the data. We will be running many optimization and trying different types on intermarket systems over the next week or two using the complete data series of energy indexes, and mutual funds. The complete list of the ones we are using are also included in the download.
Okay then Murray: First the obvious question; If you took the time to think through this setup, and then you tested it over a specific time period and you found it profitable, why would you publish it here rather than just trading it alone or as part of a portfolio?
That is a very good question and like always I will give you a honest answer. First I could publish tomorrow newspaper here and still most people would not trade it . The Crude market is very liquid so , ET can't create any problems for me. Second, I don't publish all my research only some of it. I am not giving my research or educational information away , I am trying to create educated traders who appreciate what I am trying to do and take a fair look at TradersStudio. I believe if they do , they will see that it is the best value as a backtesting and technical analysis platform.
Your results may appear to be good because you used a "biased-optimal" time interval, i.e. the period you tested was very "trendy and friendly" to your method. You'll probably have different results if you test the mid & late 1980's and 1990's.
I am not testing from 2000 forward because I wanted to be bias. I am testing from 2000 forward because that where this index data starts. Also you concept about the trend is not really true because of the profit on the short side. During this period Crude rallied about 29 point from first trade to last open trade. We made over 70 points!. This means you can't explain how well the system did based on trend bias. Also we win almost 74% of the short trades!. Here is the results workbook from TradersStudio.
i absolutely agree that optimizing a strategy to get it to show strong results for a backtest works. give me any 2 markets and i will produce an intermarket trading strategy that yields positive gains. i will also show how people who have the data can do this for free (except for time) with a "shotgun" approach.
The point you are all forgetting is that there is a reason why it works, oil stock lead crude. I have been building intermarket based systems for a decade and many of the ones I built 7-8 years ago have been profitable since I published them. Whenever you build a trading system you need to understand the premise. Using a ETF based on a basket of energy stocks make perfect sense. The results are very robust 46 of 64 combinations tested optimizing from 5-40 steps of 5 for both sets of parameters where profitable on both the long and short side. We also had 25 of them make more than $40,000. I know these number are amazing and you don't want to believe but I know several intermediate size CTA trading this type of stuff. I know because I taught them.
When you get to the intra-day timeframe you get a lot more noise. On a daily timeframe these relationships are more meaningful. In addition yes intermarket relationship can decouples. I wrote a lot about this in my orginal intermarket thread. http://elitetrader.com/vb/showthread.php?s=&threadid=55456&highlight=intermarket+analysis