Meta-Trader - Coatl Review - Part 3

Welcome back Meta-Traders.

Here is part 3 of my review of the COATL trading system. If you haven’t already, go back and read Part 1 and Part 2 before tackling part 3.

One thing to appreciate about Coatl is that it is very flexible trading system. To call it a system at all is a major oversimplification given its vast (2.41*10^16) decision space. Daniel addresses this by calling it a “framework” out of which profitable systems are derived. As we will see shortly, there are many clever ways in which it can be implemented in a portfolio fashion.

Before we get to the portfolios, take a look at the user interface. All of Daniels’s EA’s sport an attractive user interface and Coat is no exception. It shows the ATR value, profit earned up until now, the underlying logic being used by the 3 embedded systems, and “Internal Closing Threshold” or take profit/stop loss in pips. Looking at the screen, it’s easy to see the underlying logic and pip values for exit used by each of the 3 embedded systems for a given pair.

One interesting way that Coatl is deployed is in “currency centered” portfolios. My favorite is the “Euro-Centered” portfolio which trades EUR against USD, GBP, CAD, JPY, CHF, AUD and NZD. Each pair has a recommended set of custom parameters which saves you the time of performing a 5-hour optimization for each pair, followed by the non-trivial task of picking the best parameters to trade. Each pair also has a recommended risk value per trade when traded in a portfolio based on back-testing and analysis in Daniel’s portfolio analyzer program.

An important point is that these portfolios were back-tested on 9 years of historical data (2000-2009) then tested on "out of sample" data for 2010 and beyond. This avoids the pitfall of hindsight and optimization in system design.

As you can see on the right, I’ve been trading the EUR centered portfolio in a demo account for less than 3 months and its up nearly 50% in that time. I inadvertently omitted EUR/NZD from the portfolio. These are pretty impressive results to be sure. But with risk comes reward, and we need to understand the worst case scenario before trading it live.

To understand the risk, let’s examine how the system trades. Each of the 3 embedded sub-systems will open only one trade at a time. Once opened, a trade remains open indefinitely until either:

• The system closing condition occurs or
  
• The “Internal closing threshold” is hit which is essentially a take-profit or stop loss, always less than 300% of the ATR. These values take affect at the close of the current candle - in this case once a day since we are trading on the daily chart.
  
• The Emergency Stop Loss occurs which is 600% of the ATR. These values take affect intra-candle.
  

The Internal Closing threshold is optimized, and in the case of EUR/USD, its 225% of the ATR. Using the above screen shot - we see the ATR is 142 pips. That means the stop-loss in pips in this case is 319 pips (142 * 2.25) for one of the 3 positions. Given the recommended position size settings of less than 1% each on a $2000 account, positions range between 0.01 and 0.02 per . That is 10 cents a pip meaning a stop loss on one trade would come to nearly $32 or about 1.6% of the account. Multiply that times 3 positions and the max stop loss for 1 pair is $96. Multiply that times 7 (for 7 pairs) and you can see that comes to $672 or 33% of the account! That's the normal stops loss case.

In an emergency stop loss case, it would be 2.6 times that or 85%, effectively wiping the account. Now before you dismiss this system as too risky, consider the fact that not all 3 positions have to be in the same direction. For example, for the EUR/GPB pair, 2 of the 3 systems are long EUR/GBP, and one is short. That means this system is not NFA compliant, so for US-based traders, you need to use FXDD which has a back-office solution to address this situation.

Also, consider the fact that not all pairs are likely to move in the same direction at the same time. The EUR portfolio is a bit of an exception since EUR comes first every pair, so a strengthening EUR means a rising pair in nearly all cases. In fact if you look back at the performance of the Coatl EUR-Centered portfolio, it corresponds very strongly with the performance of EUR/USD itself, peaking about 3 weeks ago, and going into drawdown ever since. For other pairs and portfolios, that might not be the case. Consider a portfolio containing AUD/USD and USD/CHF. Since USD is on different sides of the pair, it’s more likely that these pairs would move in opposite directions.

It is important to note, however that this portfolio (and other Coatl portfolios) have a higher worst-case number -48% in this case. That means that if the portfolio lost 48% of its equity, we would stop trading it. This is a higher value that for most other portfolios on Asirikuy. But we get more potential return as a result as we will see below.

And on the topic of portfolios, Asirikuy has portfolios for USD-centered, JPY-centered, CHF-centered portfolios. The table below shows all the different pairs traded in those systems.

When you consider that each pair takes a 5-hour optimization run, followed by wading through the 18-sided hypercube to pick parameters, it’s clear how much work as gone into construction of these portfolios.

Also, the JPY-centered portfolio trades GBP/JPY which was an old favorite of the Zulu-Traders due to its high level of volatility. A 200-pip move in GBP/JPY was pretty common, and I recall seeing days of 500-pips or more movement on a big day.

And as if all that weren’t enough, there are also Coatl portfolios that run against the 1-hour and 4-hour charts. These portfolios stick mainly with the major pairs.

There’s also a “20-year” portfolio that trades the daily charts for the handful of major pairs for which there is 20-years or more of history. The standout thing about this portfolio is that it was optimized with historical data from 1990-2000, then tested out of sample on 2000-2010. Prior to 1999, the EUR was tested using DEM.

So as you can see, there is no shortage of ways to deploy Coatl in a portfolio setting. And as you know I’m always interested in performance. So how have these portfolios performed since they went live earlier this year?

Well I can’t share the exact performance numbers, but I can say that every portfolio is up and many of them are up better than 10%. Every portfolio also has between 10 and 20 trades open, and in some cases the open draw downs are equal to 30%-50% of the closed profits. So this system opens a lot of trades, and keeps them open for days, weeks or months. So be sure you are comfortable with that before going live with this system.

Returning to my favorite - the EUR-centered portfolio - it’s interesting to see that the live account on Asirikuy hasn’t done anywhere near as well as my own demo account. Strange since it was opened just a few days after my demo account was started.

Finally, 3 month’s demo trading history is not much to go on. So let’s wrap up with an analysis of the EUR-centered portfolio on 9-years of back-test history using Asirikuy’s profit and drawdown analyzer tool. Here are the highlights:

• 100K grew to 14 Million dollars!
  
• The “Arithmetic Mean Gain” was nearly 60% per year
  
• Maximum Drawdown was just over 21%
  
• The maximum drawdown period was just about 6 months which is about half as long as some of the other systems on Asirikuy
  
• On a monthly basis, the system was profitable about 60% of the time with average profit clearly exceeding the average loss.
  
• The largest losing month was just over 14% while the largest winning month was just over 22%
  
• System was profitable in every one of the last 10 years - with 2010 being the biggest winning year. Interesting that it was optimized with 2000-2009 and 2010 was the best year, up over 110% for the year! The best year prior that was 2003 which was up about 93%

Based on these numbers, this is clearly one of the most impressive portfolios on Asirikuy. But keep in mind the higher worst case scenario and remember to use risk capital only.

To summarize, Coatl is one of the impressive accomplishments to date on Asirikuy. Nothing is guaranteed in life, let alone in the world of forex trading. But I’m confident that if you pick a portfolio, start with an adequate amount of risk capital (at least $2000 USD), follow all the recommendations, and take a long-term perspective, there’s an excellent chance you will make some good returns trading these portfolios.