System Results and Why can’t I get close to this? – Part 1.

Tuesday, 18 June 2013

I often refer to my system in my posts and the aim of my trading is to follow the system rules as closely as I can and so get results as close to the system results as possible. Let me preface this by stating emphatically that there are dozens, if not hundreds of methods that will work successfully in forex markets. Even a simple moving average crossover system will produce a profit of 15-30% per year if followed rigidly. The problem is always in following the method rigorously.

My system rules are the result of the insights I have gained over many years of trading culminating in extensive research I did back in late 2011 to mid 2012. Many of you must have wondered what my system is. It consists of time-tested, well proven technical analysis basics – trendlines, Fibos, Support/Resistance and chart patterns. The magic happens when these are combined with insights of time-cycles I gained when I did my research. I am reluctant to disclose on this blog the timing insights that are the key to my method. The results of following the system rules mechanically are fantastic. Below are the system trade logs I have kept in parallel with my actual trade logs. Before that, I need to explain my money management strategy for the system.

Financial markets are the finest example of crowd psychology in action that we can find in the modern world. The bigger the crowd, i.e. the greater the number of participants in a market, the more applicable  insights into crowd psychology will be. In statistics this is referred to as the sample size: the larger the sample size, the more reliable the results of the statistical analysis. Therefore the most reliable markets to trade will be those that have the most participants, i.e. the most liquid markets. Therefore, I decided to trade only the most liquid markets. In forex these are the EURUSD, USDJPY, EURJPY and in the second tier GBPUSD, AUDUSD; i.e. 4 or 5 pairs at most.

With this settled, I divide my trading capital into 4 equal parts to start with, and assign 1/4 capital to each pair. Each trade in the system risks 10% of the capital available to that pair. This might seem like a lot, but 10% out of the 25% allocated to the pair works out to 2.5% of total capital risked in each trade. This happens to be the sweet spot for capital to risk as advocated by many successful traders over the years.

As the trades accumulate profits and losses, this method of allocating capital automatically allocates more to the pairs which are more successful, and less to the less successful. This also fits well with trading advice from experienced traders throughout history – Do more of what works and cut back on what doesn’t work.

One key aspect of my system is that it is always in the market. When one long trade is closed, it is immediately flipped to short. Similarly when a short trade is closed, it is immediately flipped to long. The criteria for closing a long trade is the same as the criteria for going short, and vice-versa. This is emotionally difficult because usually the close is after a good profit or a stop loss. But I’ve found that the best trades are those flipped just after a loss. Makes sense – being stopped out is the markets way of telling us that we were wrong. If being bullish is wrong then one must go short!

Here are the system results for January to June (up to 18 June):  one sheet each for EURUSD, USDJPY, EURJPY, GBPUSD, AUDUSD.

EURUSD System v2 live test Jan-Jun 2013

At the moment the system is short EURUSD and almost -100 pips in the red, as this pair is bumping up against tough resistance at 1.34 which is also the stop level. This is a situation where in live trading I would have closed this trade now, but the system stays in until it is stopped out.USDJPY System v2 live test Jan-Jun 2013

The system is long USDJPY since last Friday, and I closed my live trade near the bottom at a big loss, although the stop loss had not been exceeded. The pair has now recovered all the loss and looks like it will go higher. 
EURJPY System v2 live test Jan-Jun 2013

The EURJPY trade is tracking USDJPY and like UJ it came close to the stop level in late Friday NY session. I got out at a loss – the pair has now recovered all the loss and looks strong.GBPUSD System v2 live test Jan-Jun 2013

GBPUSD is short since this morning, after having been whipsawed by the tiny break of 1.5750 and reversal down. In hindsight the stop should have been at least 10 pips from the key 5750 level. Still, the policy of immediately flipping trades is at work here, the short from 1.5680 helping recover some of the earlier loss.AUDUSD System v2 live test Jan-Jun 2013

System is short AUDUSD from yesterday, the trade is approx 100 pips in profit.

Some readers may question the results logged here – they seem too fantastic. One has to remember that these are trades without emotion, and emotion is the key problem that blocks us from succeeding. Sure, the results will not be identical in real life unless executed by a machine. It is too hard for a person to take trades at all hours of the day, not get affected by news events, etc. But if I could get a mechanical system to trade this, the results will be close to what the mechanical system achieves.

The question many of you will ask is why not program this system and automate it? The short answer is that the rules are not indicator based, and all the ‘programmable’ platforms (mainly MT4 and NinjaTrader) are only good for developing systems using indicators. I don’t use any of those indicators, except Fibos. From what I’ve seen of MT4 and Ninja, it is not possible to program trendline breaks and recognizing chart patterns like support, resistance, etc. I’m sure it is possible to program this using advanced AI techniques but that is beyond my resources at the moment.

In the next post I’ll explore the lessons I’ve learned about myself, why I am unable to execute my system rules and what I can do about it.

By the way, if any of you are thinking that my results are bulls#^t and are tempted to make any derogatory comment, save your energy. I don’t care if you don’t believe me. I am not selling anything, and if you think this is nonsense, move on.

Advertisements

6 thoughts on “System Results and Why can’t I get close to this? – Part 1.

  1. Daniel

    Can you tweak your system to produce 4 signals instead of 2? So you will have strong buy, strong well, weak buy, and weak sell?

    Like

    Reply
    1. ZenFXTrader

      Thanks for the feedback Daniel. It’s not really possible without introducing a fuzziness around entries and exits. If entries are ambiguous, the exits also become ambiguous and that breaks the very foundations of the system. For example, if I am in a long trade and get a signal that says weak sell, what do I do? Sell half and add half short? How will that reconcile with the other half that is still long? Or sell half but don’t go short? Then when it becomes a strong sell, do I sell the remaining half and go full short or half short? I may not get another weak sell signal to go short the other half. You can see how it complicates matters. My aim when I developed my rules was to keep the rules CLEAR and SIMPLE. One fundamental fact in nature is that simple systems are best. A favourite principle of mine which is attributed to Albert Einstein says “Make things as simple as possible, but not simpler.” Although my system does suffer losses in reversals every now and then, the overall result is pretty damn good and reason enough to keep it as it is.

      Like

      Reply
  2. PJ

    Any person willing to bear his soul to the trading world the way you do on a daily basis is to be respected and admired. I’ve been following your trading journey as a non participating observer for several months. I find myself celebrating your wins and mourning your losses. Even though my trading methodology is different then yours, I am a fan and appreciate your commentary and above all, your ability to be honest with yourself.

    Like

    Reply
    1. ZenFXTrader

      Thanks for the feedback Perry. My blog is meant for serious people and I’m glad that there are at least a handful who understand and appreciate what I’m trying to do here.

      Like

      Reply
  3. fxgai

    One thing that strikes me is that (if I understand correctly) the system is always either long or short, but never flat. How does this system square with the concept of the risk / reward trade-off, which is fundamental in various fields in which risk management is applied?

    Limiting capital-at-risk per trade is important, but I believe only a part of it. If being say, long, has proven to be wrong, then it is not necessarily the case that there is a decent risk/reward opportunity in going short. (I agree that sometimes it is a good decision, just not always… Sometimes one’s direction is right, but timing or stop loss level was wrong.)

    After being stopped out of a long trade, one may be able to define a stop for a short trade, but automatically assuming the risk of this new trade skips over a robust evaluation of the risk/reward trade-off for the new trade. As a result, it’s like there is only one giant risk/reward trade-off decision taken of whether to operate against the system at all, instead of many, many independent decisions to take completely distinct trades. If there is some unrecognised flaw in operating against the system, then the single decision can be fatal.

    One other thing that struck me with regard to the point about performance – I was reading a book this morning that claims George Soros averaged around 30% returns per year over his career. We have to regard this as pretty good. I’m reading another book now about a guy who won a Robbins trading contest in Japan two years running, the first year with performance of 1000%. It’s something to aspire to, but even George is only doing 30% a year, but at least he sustained it over a time period of time.

    Like

    Reply
    1. ZenFXTrader

      Hi gaijin, thanks for the very constructive comment. You raise very interesting and valid points which I have thought about deeply and often.
      1. Risk-Reward: The fundamental flaw in thinking about risk/reward in trading (excluding other fields of work) is that it assumes we know what the reward is going to be, i.e. it assumes we KNOW how far the move will go. I’ve found that we can only know our risk, and we cannot estimate the reward with any significant degree of confidence. If one insists on having a risk/reward kind of thinking it leads us to exit a perfectly good trade when it reaches it’s expected target. Often that means we miss out on further gains. On the other hand, it also means that some trades return less than the amount risked. But my thinking is – a profit is a profit, even if it is only 10 pips. All I have to do is define and limit my risk to what I can tolerate; and let the reward be whatever it will be. RISK is the only thing we can and should worry about. If I can get a 10 pip profit with >99% confidence everyday, I’m willing to risk even a $million on those odds. So its a combination of level of risk + probability of that risk event occurring.
      If one thinks deeply about crowd psychology, one feature that stands out is that crowd behaviour moves in something like tides. It builds in one direction, reaches a maximum and then dissipates. This translates into markets as follows: price moves in one direction, builds momentum, reaches an extreme, tries to overcome for a little while and then moves in the opposite direction. It almost never moves strongly in one direction and then immediately strongly reverses (excluding exceptional circumstances like the Japan earthquake/tsunami). My system captures this behaviour by reversing when it becomes clear that further advance in one direction is blocked. Yes, prices often go into a range, but in that case my stop-loss keeps moving closer to the entry level and if it does break against me my loss is small and I immediately flip in the direction of the break.
      About performance – the ones that really make big money stay under the radar, few of us get to hear about them. I’m pretty sure Soros’ performance is actually much higher than what is publicly known. Performance is also a function of the amount of capital employed. I’m sure that even this system, rigorously followed, will not work as well if I’m dumping or buying > $50m / $100m in each trade. The mechanics of the trading will change when the amount is large and that necessarily will impact results.

      Like

      Reply

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s