The Book that Launched 1000 Systems: Whats Working Now

A framework of system ideas for traders of the S&P 500, stocks, gold and precious metals, bonds and oil

By Dan Murphy Million Dollar Target

The results listed herein are based on hypothetical trades. Plainly speaking, these trades were not actually executed. Hypothetical or simulated performance results have certain inherent limitations. Unlike an actual performance record, simulated results do not represent actual trading. Also, since the trades have not actually been executed, the results may have under (or over) compensated for the impact, if any, of certain market factors such as lack of liquidity. You may have done better or worse than the results portrayed. This newsletter has been compiled from sources believed to be reliable, but we do not guarantee its accuracy or completeness. It is to be used solely for information purposes and is not to be a solicitation for any stocks mentioned herein.

Im going to talk a little bit about opening range breakouts and breakdowns. Opening range breakout (or ORB) was coined by Tony Crabel (his extremely expensive book if you can get your hands on it. Here it is on Amazon.com)

It is pretty outdated for reasons Ill explain in a bit. I think he was running a hedge fund for a while; maybe he still is. Havent really heard much from him in a while. Lets go ahead and see why- Theres something that happened today in the S&P 500 that just really reminded me of what a lot of people think. What do I do on this show if you want to call it thatI actually show you whats going on, what the odds are and were only using systems that have tremendous edges. Of course no system is ever going to get it right all the time. Lots of systems go into drawdowns- they all do at one point or another. I just want to share with you some good ol calculations. It wasnt all that hard to calculate but its something I think you should know about. Really its the difference between trending markets and non-trending markets. What works best on trending markets? Trending systems of course. Things like breakouts- like an end of day breakout system which is what the famous trader Richard Dennis used. Which doesnt necessarily work as great as it did back then. There are better methods that you can use that are much, much better. You see, markets are always changing. Thats kind of the hint that Im going to be talking about as we get into this. Since markets do change then you have to adapt to those changing environments.

Since there are different types of markets out there lets go ahead and talk about one of Tonys little edges in the market- he coined the NR7. All this means is narrow range 7 periods. Lets see if today would qualify as an NR7. See if today is the narrowest range over the past seven bars, and lets just check that for the past seven periods. The idea was that if you have the narrowest range in the past seven periods maybe you want to buy into a breakout. What he did was calculate the opening price and then buy N Tics above or below the open (this stuff is kind of old, weve gone way past that. If youve ever purchased any of my systems such as Futures Boss, which is for day trading, were using one-minute bars to really understand exactly where the market is going). The gist of it is is that we buy on a stop just above the opening price, and then hopefully because it was a narrow range - narrow ranges give way to higher ranges. Its like Ive talked about before you have periods of high volatility and then it leads to periods of low volatility, and then you have high volatility and it leads to a period of low volatility. Thats like the market breathing in and out. Its always going to cycle that way, right? Lets really look at how things actually work in the real world. Lets design a system that would look for the narrowest range, lets make it real simple. Youll quickly see if there is an edge in simple. Lets put a buy stop in the previous day high, and were going to do that for every instance we found an NR7. Thats exactly what weve got here on the S&P 500 - we found every time theres a narrow range:

Lets make sure its doing everything its supposed to do. Oh yeah, right here this is a great example. You can see that this is a very small range, all these ranges over here for the past six periods total were small than the range over here, and we put our buy stop just above the high, and certainly thats exactly where it was executed. Since its a day trading model we go ahead and sell on the close.

Lets go ahead and see what the results are. Oh boy, not so great.

Lets go ahead and look at the performance graph:

I dont see anything there buy noise folks, noise. But wait a second, Tony liked to use lots of different markets and different asset classes which is what Ive been teaching you. We want to use multiple system ideas multiple asset classes- we have our little Ross Perot pie chart that we put our money into. Hopefully when one system zigs the other system zags. Sometimes they can all be correlated, but over time this is what makes us have smaller drawdowns and higher percentages gains. He was onto something a long time ago thats really good for him. Now unfortunately it obviously didnt work, but lets check out other markets that you might agree are trending.

Lets look at gold. Put that in there real quick, and you can see this is a loser over here:

On the average how is this doing?

That looks pretty damn good folks.

So that shows you that the gold market likes to do exactly what Ive talked about. It goes from periods of low volatility, which is what we were looking for in this little system idea with the NR7, and then it breathes out. Thats when you see the volatility expansion. All were doing is putting our buy stop just above the high and selling on the close. Over here you can see the narrow range it goes ahead and goes up above that high right there and sells on the close:

Lets go ahead and look at several other markets here instead of just gold.

Well look at silver futures using the same exact rules. Not the greatest:

Platinum:

Thats even better than Gold was. Again these are all related asset classes. Lets find something totally non-related.

Crude Oil:

You can see that there might be something there. Finally lets go to T-Bonds:

- Which for the most part have actually been, you can find mean reversion systems that work great with T-Bonds. I ran this earlier and we actually have a pretty decent system for day trading T-Bonds I wouldnt use just a very simple system like that. When you start to use really really really simple rules such as the narrow range over the past seven bars and then put a buy stop over the previous days high and then sell on the close for a day trade. If you see something like that and its such simple rules then you know youre onto something. You can make it a little more complex but not too complex. Occams Razor says the simple solution is likely the best solution.

Unfortunately this doesnt include commission or slippage, which would make this system pretty much untradeable AS IS unfortunately, but youre onto something with this kind of idea it just needs to be refined a little bit more. But this is a great starting point for you. I thought that I would share that with you and one more idea because you know what we just didnt find anything that worked with the S&P 500 did we? It seems to work on a whole lot of other markets, this whole narrow range idea. Thats because most other markets are trending, they inhale they exhale. Whereas the S&P 500 is a different animal for many different reasons. You can look at it that we have news that filters out almost instantly, the reaction takes place very quickly and thats a terminal move. So instead of starting things the news tends to end things. Kind of a dichotomy between different asset classes.

Lets try a different idea here and think of the exact inverse. Lets make sure todays range is the largest bar in the past seven days and well short the damn thing. So if this thing is up well go ahead and short at a limit at yesterdays high. Were actually looking for a really big range dayits actually up on the day and instead of going long up here were actually going to go short and cover on the close. The results:

Here we go, lets go ahead and look at the end results again this is just looking for the largest range over seven periods and shorting at the previous days high covering at the close. Very simple system rules and thats what weve got- right around 200 trades on that. Not including commission and slippage:

Those definitely eat into the profit factor 1.84 in there. I should actually show you win percentage- I normally dont show win percentage a whole lot because I dont really care about it. Im looking more at the profit factor, which is looking at the win size versus loss size, and the win percentage and combining that into one number instead of just looking at win percentage which is only going to be half the equation. Pretty good system rules overall, its interesting though that its exactly the opposite of what works in trending markets. Now lets go ahead and look at some trades to make sure that youve got it, and again that was looking for the largest range on an up bar had to be an up day. So big move up over the past seven periods, which I didnt optimize. If you wanted to optimize itd be twelve, but that number will change over time, which Ill get into in a bit. You can count back seven:

This is by far the largest range over the past seven periods. You short at the previous days high, and then you just cover on the close. Not a bad system overall, I would definitely add a whole lot more to this. Again, when you start with a simple platform its kind of like when youre painting and then you use a pencil to outline to where youre going to go with the entire painting. Then takes the hard work of filling everything in. So far this is a rough sketch, if you add a few things to it including maybe shrinking down maybe to the one-minute bar on the charts then this could be a much better system than it already is. I have to say for such simple rules this is a fantastic idea.

Now heres a chart with that same large range over the last twelve periods, this was completely optimized over the past. This wont occur into the future.

Thats what I wanted to talk about: is that markets do drift and that you do have to change parameters slightly over time. Its like if you took lets say one of the most popular programs out there for trend following which would use daily charts and stop orders which would be really easy to trade. Thats what the Turtle Systems did, which buy N day breakouts. That number N has changed over time. Many things have changed over time because they were invented during a time with very high inflation so you have a lot of trending moves during that period. You must constantly adapt to an ever-changing marketplace. Its just like life. The quickest to adapt not only survives, but thrives. Adapt or die off like the dinosaurs. I remember way back I was reading an interview with the famous hedge fund manager Bill Eckhardt and he was saying the exact same thing and thats what led me to adapting my systems with out of sample data using a rolling window. Think about it: do you really need ALL the data in your tests, or do you need to focus more on recent data? Put it this way: evolution says your ancestors had gills to swim around in the ocean. Do you need gills today to survive in your environment? Michael Phelps might say yes, but for most of us, we dont need gills to function effectively. If youre looking to increase your odds of success, look into walkforward testing. That my friends is a topic for another discussion.
2012 Joshua Tree Financial, Inc. www.milliondollartarget.com