Back-Tested Results of the Dow Double Diamond Plan

The Dow Double Diamond Plan was started as a new service on October 28, 2004, and we we believe it has tremendous wealth-building potential.

     Of course, the problem with a new service is there is no track record on which to assess performance. And while we have all heard the familiar warning, Past performance is not an indication of future results, sometimes that’s all we have to go on.

Because we do not have a tremendous amount of live trading data for you to make a proper assessment of the Dow Double Diamond Plan we have decided to show how the plan performed over a historical period. In this case we are showing the back-testing results generated as the Dow Double Diamond Plan was developed. Here's how I proceeded:

The first step was the hardest and that was coming up with the "system concept." Fortunately I did have a starting point. The Dow Double Diamond Concept is centered around one of my favorite and reliable analytical tools, which I constantly use to assess the market's potential on a daily basis. So the first real step was to take an analytical tool and morph it into a workable trading system. Once I had reached this point I was able to test and further refine the concept.

At some point during that process the Dow Double Diamond Plan evolved into a set of "hard and fast rules." There are no guidelines or generalities in the Dow Double Diamond Plan. When the market, namely the Dow, does specific things, a trade is triggered and a position change is made. This is sometimes called a "mechanical trading plan." The important point to grasp is the mechanical nature of this plan eliminates judgment calls. And that, of course, eliminates losses due to human emotion and error.

Now, the nice thing about having hard and fast rules is they can be programmed into a computer and then the computer can sift through a mountain of data while you do other things (in my case, reading a bedtime story to my girls, for instance). Thanks to the wonders of current technology we can quickly see what would have happened if we had used a given set of trading rules over a specific time period. The benefits of being able to back test a trading system are obvious but there are also pitfalls that must be avoided.

One of the problems with back testing is the desire to optimize a trading system - that is, make it better and better and better… often until it doesn't work anymore. For starters, it's not too difficult to construct a system that works great on a small segment of historical data. Then you can tweak it a little here, and a little there, and the next thing you know you're thinking you have unearthed the Holy Grail of trading systems. But then you discover it doesn't work on many other segments of data and or different time periods. Often an adjustment made to solve one problem inevitably ends up creating two new ones. In other words, it's easy to over-engineer and get something you never wanted.

So, how did I develop and refine the Dow Double Diamond Plan while avoiding the inherent problems of over-optimization? In other words, how did I keep it both simple and effective? I started with the guidelines of the proven 3-day trend concept and developed it a further. But, I focused the bulk of my energy towards overcoming the most obvious limitations. The raw 3-day trend concept, which I often discuss in my nightly commentary in Tactical Trading Outlook, boils down to this: The market has a natural 2.5 day cycle, but when the natural cycle extends out for a full three days or more, a newly established or established trend is much more likely to extend in its current direction. That concept along with the standard rules for the 3-day trends forms the foundation of the Dow Double Diamond Plan. And with that I was ready for the initial round of tests.

The results of the first tests were impressive (see Chart A), but the standard rules for the 3- day trends were still more like guidelines for short- term traders than hard and fast rules. There is that word again, guidelines, and I just told you the Dow Double Diamond Plan is not based on guidelines. That was something that had to be fixed. You see, a guideline is something that can be crossed at times, but crossing it is generally not recommended. Yet, due to human nature, the temptation is always there. Understanding how the potentially damaging effects human emotions and judgment can interfere with making the right trading decisions, I took the steps necessary to remove this aspect and make the system completely mechanical and thus more reliable and, yes, safer too.

To this end, I used the initial back testing run to determine the best course of action relative to several of the guidelines. That is, I had to make decisions on what actions, if any, would be taken when certain things occur. The goal was to settle on an alternative that would handle events consistently and effectively. But, since we are dealing with guidelines, there is no best way. Sometimes one way works better than another way, and then the next time it is the other way around. Several compromises had to be made. Like many things in life, you just have to find a good balance between two similar sides. Generally speaking I gravitated towards the more conservative alternative. At this point I was able to perform a proof of concept test run. If this one works then I know I have a solid foundation on which to build a complete system.

The chart below shows the initial proof of concept test run of the raw Dow Double Diamond Plan concept. It was constructed using eight years of daily data from May 1996 to May 2004. The chart below shows the performance of the initial hard and fast rules concept before adding protective considerations and refinements.

I can not stress how important this chart is. This chart proves that the Dow Double Diamond Plan on a cash-only basis (blue line) beats the Dow's raw performance. Each line represents the compounded return of $20,000 following each strategy: Cash invested in the Dow, a Cash DDD account, a 2:1 Margin DDD account, and a 5:1 Futures DDD account. Keep in mind that despite these impressive returns, they do not reflect any adjustments or considerations regarding certain known limitations of the 3-day trend concept. The main point of this first and very long eight year test run was to prove or disprove the validity of the founding concept. When ever you have results like this you know the underlying logic of the approach is very good.

Over a period of eight years the first test run on a cash basis beat the Dow by more than a three to one margin. And, we must consider the extreme variety of conditions that were experienced over the past eight years. We saw a huge bull market, the turn of the Century, a massive bear market the 911 tragedy, and a renewed bull market. With this chart I have confirming evidence that the Dow Diamond Plan has a very solid foundation on which to build. So, it only gets better from here.

At this point we know we are building on a solid foundation, so it is time for some improvements. Again, there are certain known limitations to the 3-day trends. You can't just follow them straight up, which is the reason they are called guidelines. Specifically, there are times when there is no safety net. On certain days there is no way to get out of a trade no matter what happens, and that is just not acceptable. So I added an exit strategy based on a combination of a longer trending period and a hard stop. My goal was not so much to enhance the return (although it did do that) but to provide protection on the downside. Even though we have a solid system concept, we know the market can and will do what ever it wants on any given day. Because of this possibility we need protection or a safety net. The safety net adds security to the system and security means confidence for you.

The safety net two fold: First there is a hard stop at 3%. This is like a mountain climber's safety line. If a climber slips and falls he always has a main safety line that ensures he will only fall so far and won't get hurt. The 3% hard stop means that no matter what happens we have strong safety line that will prevent a significant fall or injury.

There is also a second more dynamic stop that can and often does move inside the 3% stop level (be a tighter stop.) As a climber climbs higher he adjusts the anchor point of his safety line higher too. So, in the event of a mishap, a fall from a higher level won't be as dramatic as it would otherwise be. These safety enhancements provide protection and add to the overall robustness of the system with an added benefit of improving overall system performance.

As far as constructing these enhancements, I tried many combinations. Is it better to reverse a long position straight to the short side if the 3% stop is reached? Or, is it better to exit to cash and simply wait for the next trade signal? And, if we go to cash, is there a way to get back in if a signal is not given for a long time. There are many combinations and I tried them all (my wife and children may have suffered during this period). I finally settled on a set of hard and fast rules. The set of rules is not necessarily the very best in terms of optimizing every situation. I chose instead to go with what makes the most sense from a practical real world forward looking standpoint. There is always give-and-take with these decisions, and this is where one has to be extremely careful regarding the dreaded problem of over-optimizing.

I avoided the problem of over- optimizing in two ways. First, I spent an absurd amount of time poring over the data in an effort to understand what was happening when things went wrong as opposed to just trying to find a way to make it work out better. During this process I noticed that most of the situations in question were not significant. None of the really big gains or losses surrounded these events. That fact alone adds tremendous confidence to the final result. Second, the shorter the time frame you work with the more likely you are to have a less dynamic plan. We started out looking at eight years, which is a very long time. The fact we looked at such a long test period means we have to look at general market action. This is the best way to avoid optimization issues. Finally, I avoided over-optimizing because; I was over-conscious of it during each and every step of the developmental process. If anything, I erred on the side of under-optimizing. In the end, the only thing I added to the core 3-day concept was a sound and logical risk-management strategy, which provides us with greater security and confidence. This corrected the obvious limitations of the founding concept. I kept it simple, and as you'll see, improved its performance.

With the additions in place, I back tested the finished Dow Double Diamond Plan on nearly 4 years of 10- minute data. This is where the rubber hits the road. Testing a system with a lot of historical daily data is a great way to start and gives a good indication of a system's overall performance potential. But it's less than ideal. You have to wonder, what really happened on each and every day the system said to "buy." Did it say to "sell" first? Was it a clean and clear signal? Could you have really gotten that entry price? And what about that exit price? By using 10- minute intraday data I was able to clearly and accurately answer all of these questions and many more.

Nonetheless, this nearly four year test period is an excellent one because it shows how the system handled a variety of market conditions - advances, declines, chop, drift, etc. During the test period from July 2002 to May 2006, we had a little of everything, including war, a change in interest-rate bias, mixed economic reports, and a presidential election. The system handily beat the Dow on a straight cash basis, and with 2:1 and 5:1 leverage the results were nothing short of spectacular. For the near 4-year period, compound annual results came in at 34%, 52%, and 86% for the 2x, 3x, and 5x trading respectively.

From July 2002 to May 2006 there were 107 trades, which is about two or three trades per month. Of those 107, 52 are greater than zero and 55 are less than zero. Not exactly 50%, but very close. Over the long run the win loss ratio is expected to be above 50%, but from time to time it will slide below and then it will recover. If you flip a coin 100 times you should get 50 heads and 50 tails, but if you actually do it in the real world, you won’t get exactly 50 of each very often because things just don’t work out that way. The average gain is about 2.25% while the average loss is about 1.15%. Make more than you lose and you have a winning system.

Drawdown and Safety Data

Every Dow Double Diamond trade is not a winner. While 52 trades made on average 2.25% each, there were 55 trades that lost on average 1.15% each. This is just a quick reality check. The goal is to allow lots of small gains to compound into a large gain by year's end, and the big annual gains to compound into the wealth you want over time. But from time to time there will always be steps backwards. During the near four year period there were setbacks. One was a 6.85% in late August to mid-September of 2002. But that setback was followed by a gain of 21.7% over the following six weeks. The point is that even though we know there will be setbacks, we also know the Dow Double Diamond Plan has the ability to quickly recover from them. And thus over time it should substantially increase your wealth.

In the second half of 2003 the was a relatively flat period. Although the total pullback during this time was just 4.62%, it seemed as though nothing was happening. I am not going to mislead you. Anything can happen when it comes to trading the market and it is possible that starting today the Dow Double Diamond Plan will embark on a period of poor performance. This is a risk when dealing with any mechanical based system. But here is something very comforting to know. Even if you had started out at the worst possible point in 2003, even with this initial drawdown, after 12 months your total return was positive, actually up 18.57%.

As the months go by and the small gains pile up and compound, you'll begin to see the extraordinary value of the Dow Double Diamond Plan. It is a robust system that has the potential to build wealth when the DIAs are traded with full margin (2x leverage.) Of course if you have more time to build wealth, are retired and desire conservative growth or income, or just prefer less volatility, the plan can outperform the Dow on a straight cash basis (1x). Try it and see for yourself.

Jim Patterson
Editor
Dow Double Diamond Plan