System Quality Number and Stop Logic

Mugabe posted a few follow up questions regarding the flood of information Scott posted here last week. Given that the tape is gyrating sideways and everyone is still in the process of digesting what was presented I propose that we facilitate the comment section to discuss follow up question and provide clarifications.

So let’s start with Mugabe’s comment – I actually responded to him in much detail and then fat fingered a page reload. So I’ll just have to do it all over again:

1. Strategies are tied to market conditions. Nothing works all the time. How true is this? And do I really care, as I’m not looking for superformance? CAGR of 20% with shallow drawdowns is good enough for me. Scott is obviously looking for something much higher.
2. The whole question of stops, specifically that they are usually too wide once a trade really gets moving. The idea of multiple stops in a race with each other is interesting, too.

I have been coding strategies for years as you guys know and let me assure you that it’s absolutely essential to start with determining the market conditions that drive your assumptions. It’s rather obvious really – nobody develops dip buying strategies in the brutal sell off periods produced by secular bear markets. We all respond by what we experience and thus develop ideas around what is happening right now or in recent history. Which often also leads to the demise of many great systems – they work fine as long as markets conditions remain the same but quickly fall apart when the character of the market changes. I often talk about market weather and it’s key that you find ways to identify market characteristics that work well for you or your systems.

Let’s simply say for example that your belief is that buying every dip at the SMA(100) provides a good edge. If you trade this on a daily or an hourly chart you will quickly discover periods in which this will work very well and times in which it will fall apart and your edge goes down the drain. You may disagree and perhaps you are only interested in systems that trade all the time. Ivan for instance completely opposes market weather optimization – he is happy with taking every single entry. But he also has developed the trading maturity and years of experience that tell him that his edge will hold up over time. Let me assure you that he is probably one out of a million people out there – not only is Ivan extremely intelligent (my guess would be an IQ of 180+) but he also has developed a very resilient (almost robotic) mental fortitude. As much as you may like to believe otherwise – chances are that this is NOT YOU (sorry – but it’s not me either). So at minimum being active in favorable market conditions will make things a lot easier for you psychologically. It’s one thing to define and prove an edge – it’s another being able to actually stick with a system that follows it.

Technically speaking it will also lead to higher SQN scores. If you remember our System Quality Number formula:

SQN = root(n) * expectancy / stdev(R)

Given our example yielding an expectancy of 0.8 – you have a system that wins 30% of the time. When it wins it nets you 5R while losing trades lose 1R:

(0.3 * 5) – (0.7 * 1) = 1.5 – 0.7 = 0.8

That’s all fine and dandy but recall that expectancy does not account for opportunity and standard deviation (which addresses MFE and MAE). Perhaps this system only trades 25 times per year:

SQN = root(25) * 0.8 / 2.5 = 5 * 0.8 / 2.5 = 1.6 (which is tradeable)

Now by selecting certain market conditions you may be able to change your rule sets to be more opportunistic, thus allowing you 100 entries per year.

SQN = root(100) * 0.8 / 2.5 = 10 * 0.8 / 2.5 = 3.2 (which is excellent)

You also may be able to optimize your stop logic by starting out with a wide ISL and then narrow down as things go in your favor. This may be much more optimal than simply picking a static stop and losing a big good chunk of the maximum favorable excursion that certain market conditions offer you. I cannot over emphasize how important this is. If you know which market conditions work best for you then you will have a much better lock on your MFE and MAE (maximum adverse excursion). Let’s assume you only have 50 trades per year but you can narrow standard deviation down to 2.0:

SQN = root(70) * 0.8 / 2.0 = 8.3 * 0.8 / 2.0 = 3.32 (which is also excellent)

Had we not clearly defined our favorable market conditions and also trades 70 times but in all market conditions our SQN score would suffer quite a bit:

SQN = root(70) * 0.8 / 2.5 = 8.3 * 0.8 / 2.5 = 2.66

Now an SQN score of 2.66 is well worth pursuing but it most likely will have larger draw down periods and will not be as ‘fun’ to trade. There is something very rewarding about taking entries under very specialized conditions and see things take off when things shift into focus.

I strongly suggest that you fully absorb this rather trivial formula and that you fully internalize it. SQN as well as expectunity are very important aspects of trading which will open up an entire new world to you. Suddenly many of those formerly abstract moving pieces start making sense and the dynamics of system performance are actually starting to work in your favor. You will instantly realize how MAE and MFE affect your respective systems and how you may be able to optimize for it. If your expectancy drops then you will also know why. And if you don’t have enough opportunity you may be able to compensate by trading other markets instead of trading the same market more often.

Of course I cannot complete this post without also mentioning an important caveat that I have learned the hard way.

Premature optimization is the root of all evil.

It’s one thing to capture statistics and define market conditions – it’s another to immediately jump to conclusions and add one rule and exception after the other. Make sure that whatever you discover works as simply and as often as possible. More rules are rarely better than fewer rules. In other words – the answer does not lie in complexity, rather it often hides out in the open. As Scott already mentioned – the problem with especially intelligent traders (and I hope that speaks to you) is that we have a tendency to believe in complex solutions.

My IQ tested out somewhere in the upper middle at 149 – and thinking that I was a lot smarter than I really am I often fell into the trap of developing complex rules which in the end turned out to be nothing but form fitting. Start out with the bare essentials and then work yourself up from there. Instead of 10,000 lines of code start out with maybe 50 or 100 and test your basic assumption. Then grow from there, slowly adding one component after the next, thus making sure that all the pieces work well together. You still may wind up with something complicated years later but each small unit will have proven is merit.

Also, don’t be afraid of throwing pieces away if they have limited value – both code or rules. As human beings we quickly fall in the trap of the sunk cost effect – another of our cognitive biases. If you start out with a huge code base or with 100 rules you will not be able to determine where your problems are if things fall apart. However, if you start out small and add new rules piece by piece you have a much better chance of finding out what is suddenly breaking things for you.

In the coming days let’s go over various aspects of system development that speak to you and that have evoked more questions. Feel free to ask away and I’ll do my best to cover it to the best of my abilities (or ask Scott for backup if I have to). Also – don’t be shy to post charts here. Just because entries are only 20% of trading it doesn’t mean they are of no importance. But perhaps we can start shifting things toward hunting for edges and applying them to our favorite charts. So in the future I would love to see folks post a chart and say – I’ll trade this via CI rules, or via my own X or Y system – here are the basic specs. If something promises a consistent edge I would be more than happy to code the rules in NinjaTrade and make it available to you guys.

It’s not too late – learn how to consistently bank coin without news, drama, and all the misinformation. If you are interested in becoming a subscriber then don’t waste time and sign up here. The Zero indicator service also offers access to all Gold posts, so you actually get double the bang for your buck.

Cheers,